COPYRIGHT AND CITATION CONSIDERATIONS FOR THIS THESIS/ DISSERTATION

o Attribution — You must give appropriate credit, provide a link to the license, and indicate if
changes were made. You may do so in any reasonable manner, but not in any way that
suggests the licensor endorses you or your use.

o NonCommercial — You may not use the material for commercial purposes.

o ShareAlike — If you remix, transform, or build upon the material, you must distribute your
contributions under the same license as the original.

How to cite this thesis

Surname, Initial(s). (2012). Title of the thesis or dissertation (Doctoral Thesis / Master’s
Dissertation). Johannesburg: University of Johannesburg. Available from:
http://hdl.handle.net/102000/0002 (Accessed: 22 August 2017).
AN INTEGRATED TOTAL QUALITY MANAGEMENT
MODEL FOR THE GHANAIAN CONSTRUCTION
INDUSTRY
A thesis presented

by

SAMUEL KWAME ANSAH

to

THE FACULTY OF ENGINEERING AND THE BUILT

ENVIRONMENT

in fulfilment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

in the subject of

ENGINEERING MANAGEMENT

UNIVERSITY OF JOHANNESBURG, JOHANNESBURG,

SOUTH AFRICA

PROMOTER: PROF. W.D. THWALA

CO-PROMOTER: PROF. C.O. AIGBAVBOA

OCTOBER 2018
 DEDICATION
This thesis is dedicated to my late parents (Mr. and Mrs. Owusu Ansah), and my elder brother
(Mr. Felix Yaw Acheampong), and most especially to my wife (Lucy Ansah). I thank them for
their support and love. Without their patience, understanding and encouragement this study
would not have been completed.

i
 DECLARATION
I, Ansah Kwame Samuel, declare that “An integrated Total Quality Management model for the
Ghanaian construction industry” is my own work and that all the sources that I have used or
quoted have been indicated and acknowledged by means of complete references. The thesis is
submitted in fulfilment of the requirements of the degree Doctor of Philosophy in Engineering
Management.

________________________ ________________________
Samuel Kwame Ansah Date

ii
 ACKNOWLEDGEMENTS

My gratitude to Almighty God and the Lord Jesus Christ for given me strength, grace and
guidance throughout the course of my study, and without whom accomplishing this task would
not have been possible. I would also like to recognize and acknowledge the following
individuals for having contributed in various ways to the successful completion of this research:

– Professor Wellington Didihukut Thwala for offering professional advice and motivation
throughout the course of my study,
– Professor Clinton Ohi-Williams Aigbavboa for his brotherly, professional advice and
motivation throughout the course of my study;
– Professor Lawrence Atepor (Vice Chancellor) and Dr. Zakari Mustapha (Dean, School of
Enginering) both of Cape Coast Technical University, for their support;
– My wife, Lucy Ansah, and children (Leticia Ansah, Gillian Ansah and Priscilla Ansah)
for their endurance;
– My elder brother, Mr. Felix Yaw Acheampong, for his financial support during the
course of my study;
– Delphi experts for their valuable time during the survey;
– Dr. Richard Devey, Head of Department at STATKON, for data input and analysis;
– Mr. Eric Awere for editing the literature review;
– Mr. Michael Manford for assisting me through the statistical analysis;
– My family and close friends at Cape Coast Technical University for their support and
prayers; and
– All teaching and non-teaching staff in the Building Technology and Civil Engineering
Departments of Cape Coast Technical University who have been instrumental during the
data collection.

iii
 ABSTRACT
This research project investigated and modelled Total Quality Management (TQM) for the
Ghanaian construction industry. The primary aim of the research was to model the extent to
which Leadership/Top Management features, Company Supplier Quality Management features,
Client Focus and Involvement features, Company Quality System Evaluation features, Company
Vision and Plan Statement features, Product Design Management features, Product Selection
Management features, Construction Process Management and Improvement features, and
Construction Employees’ Involvement and Motivation features predict TQM for the construction
industry, these factors being classified as the exogenous variables. Mixed-methods research
which involved both Qualitative and Quantitative approaches was adopted for the study.
Empirical data was collected through a Delphi study and a field questionnaire survey. Analysis
of results from the Delphi study was done with Microsoft Excel to output descriptive statistics. A
conceptual integrated TQM for the Ghanaian construction industry model was based on the
theory developed from literature review findings and the Delphi study. A questionnaire survey
was conducted among the top management working in the construction industry in Ghana. From
the 641 sample questionnaires, 536 questionnaires were returned which represents 83.62 per
cent. An exploratory factor analysis (EFA) was conducted on the initial eight-factor constructs
and their variables to determine their reliability for their inclusion in the confirmatory factor
analysis (CFA). Nine-factor constructs were realized after the EFA factor loading test. Further,
CFA was conducted on these nine-factor constructs using structural equation modelling (SEM)
software with Eqations (EQS) version 6.2 software programme to validate and determine their
reliability and inclusion in the final model. Findings from the literature on TQM studies revealed
the theory that TQM implementations and practices and the latent variables lead to TQM in the
construction industry. Findings from the Delphi study revealed that several factors
(Leadership/Top Management features, Company Supplier Quality Management features, Client
Focus and Involvement features, Company Quality System Evaluation features, Company Vision
and Plan Statement features, Product Selection and Design Management features, Construction
Process Management and Improvement features, and Construction Employees’ Involvement and
Motivation features) were considered to be the most important determinants of TQM in the
Ghanaian construction industry. Both findings revealed that TQM could be considered as an
eight-factor model defined by the influence of TQM practices and experts in construction

iv
industry. Findings from the field questionnaire survey revealed that the hypothesis on TQM
constructs had an influence on TQM in the construction industry and could not be rejected.
Hence, it was found that the SEM results on the model’s goodness-of-fit and statistical
significance of parameter estimates met the cut-off criteria for the hypothesized model’s fit to the
sample data. The study’s contribution to the body of knowledge is significant because it
addresses the lack of theoretical information about which factors are most significant in
predicting TQM. Also, the study developed a new holistically-integrated model for the prediction
of TQM for the Ghanaian construction industry. The current integrated model advances that
TQM for the Ghanaian construction industry is a nine-factor construct. Another noteworthy
contribution to the body of knowledge is in the methodology adopted. Hence, this study offers a
base for other researchers to use as a follow-up for future studies. The study recommends that
construction firms and all stakeholders involved in construction works should consider the
empirically tested constructs when planning and implementing TQM programmes as this will
enhance the quality of the implementation of TQM in construction industry.

Keywords: Total Quality Management, Construction Industry, Structural Equation Modelling,

Latent Variables, Top Management, Ghana

v
 TABLE OF CONTENTS
DEDICATION ................................................................................................................................. i

DECLARATION ............................................................................................................................ ii

ACKNOWLEDGEMENTS ........................................................................................................... iii

ABSTRACT ................................................................................................................................... iv

TABLE OF CONTENTS ............................................................................................................... vi

LIST OF TABLES ...................................................................................................................... xvii

LIST OF FIGURES .................................................................................................................... xxv

LIST OF APPENDICES ........................................................................................................... xxvii

LIST OF ABBREVIATIONS .................................................................................................. xxviii

CHAPTER ONE ............................................................................................................................. 1

1 INTRODUCTION .................................................................................................................. 1

1.1 BACKGROUND .............................................................................................................. 1

1.2 THE RESEARCH PROBLEM STATEMENT................................................................ 4

1.3 AIM OF THE STUDY ..................................................................................................... 6

1.4 RESEARCH MOTIVATION .......................................................................................... 7

1.5 SIGNIFICANCE OF THE STUDY ................................................................................. 8

1.6 THE RESEARCH ............................................................................................................ 9

1.6.1 Research Questions ................................................................................................... 9

1.6.2 Research Objectives .................................................................................................. 9

1.6.3 Research Methodology ........................................................................................... 10

1.7 CONTRIBUTION TO KNOWLEDGE ......................................................................... 14

vi
 1.7.1 Value of the Study .................................................................................................. 15

1.7.2 Delimitations ........................................................................................................... 15

1.8 ETHICAL CONSIDERATION ..................................................................................... 16

1.9 OVERVIEW OF CHAPTERS ....................................................................................... 16

1.10 CONCLUSION .............................................................................................................. 19

CHAPTER TWO .......................................................................................................................... 20

2 THEORETICAL AND CONCEPTUAL PERSPECTIVES OF TQM RESEARCH ........... 20

2.1 INTRODUCTION .......................................................................................................... 20

2.2 THEORY AND DEFINITION ...................................................................................... 20

2.2.1 What is Theory? ...................................................................................................... 20

2.2.2 Theories towards the Implementation of TQM ...................................................... 21

2.3 THEORY ADOPTED FOR THIS STUDY ................................................................... 24

2.3.1 System Theory ........................................................................................................ 24

2.3.2 Theory of Constraints ............................................................................................. 27

2.4 EMPIRICAL RESEARCH ON TQM RESEARCH ...................................................... 30

2.5 QUALITY AND FORMS OF QUALITY MANAGEMENT PRACTICES ................ 32

2.5.1 Quality..................................................................................................................... 32

2.5.2 Quality Control ....................................................................................................... 34

2.5.3 Quality Assurance ................................................................................................... 36

2.5.4 Quality Management ............................................................................................... 38

2.5.5 Total Quality Management ..................................................................................... 39

vii
 2.6 MODELS FOR TOTAL QUALITY MANAGEMENT IMPLEMENTATION ........... 40

2.6.1 Formal Quality Award Models ............................................................................... 40

2.6.2 Other Models for Total Quality Management ........................................................ 40

2.7 CONCLUSION .............................................................................................................. 44

CHAPTER THREE ...................................................................................................................... 45

3 GAPS IN TOTAL QUALITY MANAGEMENT RESEARCH .......................................... 45

3.1 INTRODUCTION .......................................................................................................... 45

3.2 OBSERVED GAPS IN THE TQM LITERATURE ...................................................... 45

3.2.1 Lack of Consensus of TQM Concept and Constructs ............................................. 46

3.3 CONCLUSION .............................................................................................................. 54

CHAPTER FOUR ......................................................................................................................... 55

4 THE CONSTRUCTION INDUSTRY AND TOTAL QUALITY MANAGEMENT ......... 55

4.1 INTRODUCTION .......................................................................................................... 55

4.2 OVERVIEW OF CONSTRUCTION INDUSTRY AND TQM .................................... 55

4.3 TYPES OF CONSTRUCTION ...................................................................................... 57

4.3.1 Building Construction ............................................................................................. 58

4.3.2 Highway Construction ............................................................................................ 58

4.3.3 Industrial Construction............................................................................................ 59

4.4 SPECIFIC CHARACTERISTICS OF CONSTRUCTION SECTOR ........................... 59

4.5 MANAGEMENT ORGANIZATION SYSTEM IN THE CONSTRUCTION

COMPANIES ........................................................................................................................... 60

4.6 CONCEPT OF TQM ...................................................................................................... 62

viii
 4.7 IMPLEMENTATION OF TQM IN THE CONSTRUCTION INDUSTRY ................. 63

4.8 CRITICAL SUCCESS FACTORS (CSF) OF TQM IMPLEMENTATION IN THE

CONSTRUCTION INDUSTRY .............................................................................................. 65

4.9 BENEFITS OF TQM IMPLEMENTATION IN CONSTRUCTION INDUSTRY ...... 67

4.10 CONCLUSION .............................................................................................................. 68

CHAPTER FIVE .......................................................................................................................... 69

5 TQM IMPLEMENTATION IN DEVELOPED COUNTRIES ............................................ 69

5.1 INTRODUCTION .......................................................................................................... 69

5.2 THE EVOLUTION OF TQM IN THE DEVELOPED COUNTRIES .......................... 69

5.3 QUALITY PIONEERS IN DEVELOPED COUNTRIES AND THEIR

CONTRIBUTIONS TOWARDS TOTAL QUALITY MANAGEMENT CONCEPT ......... 75

5.3.1 Contribution by Dr. Joseph Juran ........................................................................... 75

5.3.2 Contributions by Prof. Edwards Deming ................................................................ 76

5.3.3 Contributions by Philip B. Crosby .......................................................................... 78

5.3.4 Contributions by Dr. Kaoru Ishikawa ..................................................................... 79

5.3.5 Contributions by Armand V. Feigenbaum .............................................................. 80

5.3.6 Summary of Contributions by Quality Pioneers towards TQM Concept ............... 81

5.4 TQM IMPLEMENTATION IN CHINA ....................................................................... 85

5.4.1 Why China was Chosen for the Study .................................................................... 85

5.4.2 Evolution of TQM Implementation in China.......................................................... 85

5.5 LESSON LEARNT FROM DEVELOPED COUNTRIES ON TQM

IMPLEMENTATION ............................................................................................................... 87

ix
 5.6 CONCLUSION .............................................................................................................. 87

CHAPTER SIX ............................................................................................................................. 88

6 TOTAL QUALITY MANAGEMENT IMPLEMENTATION IN DEVELOPING

COUNTRIES ................................................................................................................................ 88

6.1 INTRODUCTION .......................................................................................................... 88

6.2 THE SITUATION OF NIGERIA CONSTRUCTION INDUSTRY ............................. 88

6.2.1 Construction Industry in Nigeria ............................................................................ 88

6.2.2 Implementation of TQM in the Nigerian Construction Industry ............................ 89

6.2.3 Lessons Learnt from Nigeria TQM......................................................................... 94

6.3 THE SOUTH AFRICAN CONSTRUCTION INDUSTRY .......................................... 95

6.3.1 South African Policies on Construction .................................................................. 97

6.3.2 Poor Contractor Performance and its Causes in South Africa .............................. 102

6.3.3 Problems pertaining to Poor Quality in the South African Construction Industry 103

6.3.4 Implementation of TQM in South Africa Construction Industry ......................... 105

6.3.5 Lessons Learnt from the South African Construction Industry ............................ 107

6.4 CONCLUSION ............................................................................................................ 108

CHAPTER SEVEN .................................................................................................................... 109

7 TOTAL QUALITY MANAGEMENT IMPLEMENTATION IN THE GHANAIAN

CONSTRUCTION INDUSTRY ................................................................................................ 109

7.1 INTRODUCTION ........................................................................................................ 109

7.2 GENERAL OVERVIEW OF GHANA ....................................................................... 109

7.2.1 Political Overview ................................................................................................ 112

x
 7.2.2 Economic Overview.............................................................................................. 113

7.2.3 Sociological Overview .......................................................................................... 116

7.2.4 Technological Overview ....................................................................................... 116

7.3 IMPLEMENTATION OF TOTAL QUALITY MANAGEMENT IN THE GHANAIAN

CONSTRUCTION INDUSTRY ............................................................................................ 118

7.3.1 Overview of Building Regulation and Building Codes for Regulating Construction
Project Activities in Ghana ................................................................................................. 118

7.4 LESSONS LEARNT FROM TOTAL QUALITY MANAGEMENT

IMPLEMENTATION IN THE GHANAIAN CONSTRUCTION INDUSTRY ................... 121

7.5 CONCLUSION ............................................................................................................ 122

CHAPTER EIGHT ..................................................................................................................... 123

8 RESEARCH METHODOLOGY........................................................................................ 123

8.1 INTRODUCTION ........................................................................................................ 123

8.2 PHILOSOPHICAL CONSIDERATION IN RESEARCH METHODOLOGY .......... 123

8.2.1 Ontological Considerations ................................................................................... 124

8.2.2 Epistemology ........................................................................................................ 125

8.3 QUANTITATIVE VERSUS QUALITATIVE RESEARCH METHODOLOGY ...... 126

8.3.1 Quantitative Methodology .................................................................................... 127

8.3.2 Qualitative Methodology ...................................................................................... 129

8.3.3 Combined Quantitative and Qualitative Methods................................................. 132

8.3.4 Mixed-Methods Approach .................................................................................... 133

8.3.5 Justification of the Mixed-Method Approach ....................................................... 137

xi
 8.4 RESEARCH DESIGN ................................................................................................. 138

8.4.1 Methods................................................................................................................. 144

8.4.2 Stage One: Literature Review ............................................................................... 144

8.4.3 Stage Two: Delphi Method ................................................................................... 146

8.4.4 Stage Three (Questionnaire Survey) ..................................................................... 175

8.4.5 Methods of Data Analysis of the Quantitative Survey ......................................... 198

8.4.6 Structural Equation Modelling Analysis ............................................................... 199

8.4.7 Ethical Considerations .......................................................................................... 208

8.5 CONCLUSION ............................................................................................................ 208

CHAPTER NINE ........................................................................................................................ 209

9 RESULTS FROM THE DELPHI STUDY......................................................................... 209

9.1 INTRODUCTION ........................................................................................................ 209

9.2 OBJECTIVES OF DELPHI STUDY ........................................................................... 209

9.3 EXPERT PANEL SELECTED FOR THE DELPHI STUDY ..................................... 210

9.4 FINDINGS FROM THE DELPHI STUDY ................................................................ 214

9.5 DISCUSSIONS FROM THE DELPHI STUDY RESULTS ....................................... 226

9.6 CONCLUSION ............................................................................................................ 232

CHAPTER TEN.......................................................................................................................... 233

10 THE CONCEPTUAL INTEGRATED TOTAL QUALITY MANAGEMENT MODEL . 233

10.1 INTRODUCTION ........................................................................................................ 233

10.2 SELECTION OF VARIABLES FOR TQM IMPLEMENTATION ........................... 233

xii
 10.2.1 Leadership and Top Management (LTM)............................................................. 235

10.2.2 Company Supplier Quality Management (CSQM)............................................... 237

10.2.3 Client Focus and Involvement (CFI) .................................................................... 239

10.2.4 Construction Process Management and Improvement (CPMI) ............................ 241

10.2.5 Construction Employee Involvement and Motivation (CEIM) ............................ 243

10.3 MODEL SPECIFICATION AND JUSTIFICATION ................................................. 253

10.4 STRUCTURAL COMPONENT OF THE MODEL .................................................... 255

10.5 MEASUREMENT COMPONENT OF THE MODEL DYNAMIC ........................... 256

10.6 CONCLUSION ............................................................................................................ 257

CHAPTER ELEVEN .................................................................................................................. 258

11 SURVEY ANALYSIS ........................................................................................................ 258

11.1 INTRODUCTION ........................................................................................................ 258

11.2 DESCRITIVE STATISTICS ....................................................................................... 259

11.2.1 Demographic Profile of the Respondents and Firms ............................................ 259

11.2.2 Level of TQM Implementation in the Ghanaian Construction Industry............... 263

11.2.3 Attributes Influencing the Implementation of TQM in the Ghanaian Construction

Industry ............................................................................................................................... 264

11.2.4 Project Outcomes that Motivate Construction Companies to Implement TQM... 272

11.2.5 Reasons for Non-Implementation (Non-Adoption) of TQM in Ghanaian

Construction Industry.......................................................................................................... 274

11.2.6 PERFORMANCE OF CONSTRUCTION COMPANIES IN GHANA .............. 275

11.3 INFERENTIAL STATISTICS ..................................................................................... 277

xiii
 11.3.1 Conceptual TQM Model Indicator Variables obtained after Exploratory Factor
Analysis............................................................................................................................... 277

11.3.2 Results of Exploratory Factor Analysis Data........................................................ 281

11.4 THE STRUCTURAL EQUATION MODELING (SEM) ........................................... 291

11.4.1 Structural Equation Modelling Analytic Strategy................................................. 292

11.4.2 Conceptual TQM Model Indicator Variables obtained after CFA ....................... 293

11.4.3 The Measurement Model Confirmatory Factor Analysis ..................................... 295

11.4.4 Confirmatory Factor Analysis of the Latent Construct......................................... 301

11.4.5 Structural Model – Testing of the Hypothesized SEM Model ............................. 381

11.4.6 Hypothesized Relation for the Structural Model .................................................. 384

11.4.7 Fit Statistics on the Structural Model.................................................................... 385

11.4.8 Model Explanation ................................................................................................ 398

11.4.9 Summary of SEM Model ...................................................................................... 402

11.5 CONCLUSION ............................................................................................................ 402

CHAPTER TWELVE ................................................................................................................. 404

12 DISCUSSION OF RESULTS ............................................................................................ 404

12.1 INTRODUCTION ........................................................................................................ 404

12.2 QUESTIONNAIRE SURVEY RESULTS .................................................................. 404

12.2.1 Leadership and Top Management Features’ Influence on TQM .......................... 406

12.2.2 Company Supplier Quality Management Features’ Influence on TQM ............... 408

12.2.3 Client Focus and Involvement Features Influence on TQM ................................. 409

12.2.4 Company Quality System Evaluation Features’ Influence on TQM .................... 410
xiv
 12.2.5 Company Vision and Plan Statement Features’ Influence on TQM..................... 411

12.2.6 Product Design Management Features’ Influence on TQM ................................. 413

12.2.7 Product Selection Management Features’ Influence on TQM ............................. 414

12.2.8 Construction Process Management and Improvement features influence on TQM ...
............................................................................................................................... 415

12.2.9 Construction Employees’ Involvement and Motivation Features’ Influence on

TQM ............................................................................................................................... 416

12.2.10 Other Questionnaire survey results ....................................................................... 418

12.3 MAJOR FINDINGS FROM DELPHI SURVEY AND QUESTIONNAIRE RESULTS .

...................................................................................................................................... 421

12.4 CONCLUSION ............................................................................................................ 425

CHAPTER THIRTEEN .............................................................................................................. 427

13 CONCLUSION ................................................................................................................... 427

13.1 INTRODUCTION ........................................................................................................ 427

13.1.1 Research Objective RO1 ....................................................................................... 427

13.1.2 Research Objective RO2 ....................................................................................... 429

13.1.3 Research Objective RO3 ....................................................................................... 430

13.1.4 Research Objective RO4 ....................................................................................... 432

13.1.5 Research Objective RO5 ....................................................................................... 432

13.1.6 Research Objective RO6 ....................................................................................... 433

13.2 CONTRIBUTION AND VALUE OF THE RESEARCH ........................................... 434

13.2.1 Theoretical Contribution ....................................................................................... 434

xv
 13.2.2 Methodological Contribution ................................................................................ 435

13.2.3 Practical Contribution and Value .......................................................................... 435

13.3 RECOMMENDATIONS ............................................................................................. 436

13.3.1 Methodological ..................................................................................................... 436

13.3.2 Theoretical ............................................................................................................ 437

13.3.3 Practical and Policy Implication ........................................................................... 438

13.4 LIMITATIONS ............................................................................................................ 438

13.5 RECOMMENDATIONS FOR FURTHER RESEARCH ........................................... 439

13.6 CONCLUSION ............................................................................................................ 439

REFERENCES ........................................................................................................................... 442

APPENDIX A ......................................................................................................................... 474

APPENDIX B ......................................................................................................................... 475

APPENDIX C ......................................................................................................................... 476

APPENDIX D ......................................................................................................................... 478

APPENDIX E ......................................................................................................................... 494

APPENDIX F.......................................................................................................................... 509

APPENDIX G ......................................................................................................................... 526

xvi
 LIST OF TABLES
Table 2.1: Changing View of Quality ........................................................................................... 33

Table 3.1: Framework Comparison .............................................................................................. 48

Table 5.1: Important Events in the Development of TQM ........................................................... 72

Table 5.2: Universal Processes for Managing Quality ................................................................. 76

Table 8.1: Research Procedure ................................................................................................... 142

Table 8.2: Key Delphi Questions and Phrasing for the Study .................................................... 157

Table 8.3: Influence or Likelihood Scale .................................................................................... 166

Table 8.4: Impact Scale............................................................................................................... 166

Table 8.5: Conceptual Model Indicator Variables ...................................................................... 181

Table 9.1: Residential location of experts................................................................................... 211

Table 9.2: Qualification of Panel of Experts .............................................................................. 212

Table 9.3: Panel of Experts’ Years of Experience ...................................................................... 212

Table 9.4: Panel of Experts’ Fields of Specialization ................................................................. 213

Table 9.5: Total Quality Management (TQM) in the Construction Industry Main Attributes ... 216

Table 9.6: Leadership and Top Management ............................................................................. 217

Table 9.7: Company Supplier Quality Management .................................................................. 217

Table 9.8: Client Focus and Involvement ................................................................................... 218

Table 9.9: Company Quality System Evaluation ........................................................................ 219

Table 9.10: Company Vision and Plan Statement ...................................................................... 219

Table 9.11: Product Selection and Design Management ............................................................ 220

Table 9.12: Construction Process Management and Improvement ............................................ 221

xvii
Table 9.13: Construction Employees’ Involvement and Motivation .......................................... 222

Table 9.14: Factors that Enable Construction Companies to Implement TQM ......................... 222

Table 9.15: Factors that Affect Construction Companies for not Implementing TQM in Ghana
..................................................................................................................................................... 223

Table 9.16: Organizational Factors that Affect Construction Companies in the Implementation
of TQM ....................................................................................................................................... 225

Table 9.17: Effects of Implementing TQM in the Construction Industry .................................. 226

Table 110.1: Conceptual Model Latent Features ........................................................................ 248

Table 11.1: Gender...................................................................................................................... 259

Table11.2: Age Category ............................................................................................................ 260

Table 11.3: Highest Qualification ............................................................................................... 260

Table 11.4: Work Experience in the Construction Industry ....................................................... 261

Table 11.5: Position Held in the Firm ......................................................................................... 261

Table 11.6: Tenure of Current Position ...................................................................................... 262

Table 11.7: Existence of Firm ..................................................................................................... 262

Table 11.8: Type of Current Construction Projects .................................................................... 263

Table 11.9: Classification of Firm .............................................................................................. 263

Table 11.10: TQM Implementation in the Ghanaian Construction Industry .............................. 264

Table 11.11: Leadership and Top Management features ............................................................ 265

Table 11.12: Company Supplier Quality Management Features ................................................ 266

Table 11.13: Client Focus and Involvement Features ................................................................ 267

Table 11.14: Company Quality System Evaluation Features ..................................................... 268

xviii
Table 11.15: Company Vision and Plan Statement Features ...................................................... 269

Table 11.16: Product Selection and Design Management Features ........................................... 270

Table 11.17: Construction Process Management and Improvement Features ............................ 271

Table 11.18: Construction Employees’ Involvement and Motivation Features ......................... 272

Table 11.19: Project Outcomes that Motivate Construction Companies to Implement TQM ... 273

Table 11.20: Reasons for Non-Implementation (Non-Adoption) of TQM................................. 274

Table 11.21: Performance of Construction Companies in Ghana .............................................. 276

Table 11.22: Conceptual Model Indicator Variables .................................................................. 277

Table 11.23: Factor one (F 1) ..................................................................................................... 282

Table 11.24: Factor two (F 2) ..................................................................................................... 283

Table 11.25: Factor three (F 3) ................................................................................................... 284

Table 11.26: Factor four (F4)...................................................................................................... 285

Table 11.27: Factor five (F 5) ..................................................................................................... 286

Table 11.28: Product Design Management (PDM) - Factor six (F6A) ...................................... 288

Table 11.29: Product Selection Management (PSM) - Factor six (F6B) .................................... 289

Table 11.30: Factor seven (F7) ................................................................................................... 290

Table 11.31: Factor eight (F8) .................................................................................................... 291

Table 11.32: Conceptual Model Indicator Variables .................................................................. 293

Table 11.33: Cut-off Criteria for Fit Indices ............................................................................... 296

Table 11.34: Univariate Statistics and Yuan, Lambert and Fouladi’s Normalized Multivariate
Estimates ..................................................................................................................................... 299

Table 11.35: Postulated Leadership and Top Management Features Model .............................. 303
xix
Table 11.36: Residual Covariance Matrix for Leadership and Top Management Model
(Unstandardized) ......................................................................................................................... 304

Table 11.37: Residual Covariance Matrix for Leadership and Top Management Model .......... 305

Table 11.38: Robust Fit Indexes for Leadership and Top Management Features Construct ..... 306

Table 11.39: Factor Loading and Z-statistics of Leadership and Top Management Measurement
..................................................................................................................................................... 308

Table 11.40: Reliability and Construct Validity of Leadership and Top Management Feature
Model .......................................................................................................................................... 309

Table 11.41: Postulated Company Supplier Quality Management Features Model ................... 312

Table 11.42: Residual Covariance Matrix for Company Supplier Quality Management Model
(Unstandardized) ......................................................................................................................... 312

Table 11.43: Residual Covariance Matrix for Company Supplier Quality Management Model
(Standardized) ............................................................................................................................. 313

Table 11.44: Robust Fit Indexes for Features Construct ............................................................ 315

Table 11.45: Factor Loading and Z-statistics of Company Supplier Quality Management
Measurement ............................................................................................................................... 316

Table 11.46: Reliability and Construct Validity of Company Supplier Quality Management
Feature Model ............................................................................................................................. 317

Table 11.47: Postulated Client Focus and Involvement Features Model ................................... 320

Table 11.48: Residual Covariance Matrix for Client Focus and Involvement Model
(Unstandardized) ......................................................................................................................... 320

Table 11.49: Residual Covariance Matrix for Client Focus and Involvement Model
(Standardized) ............................................................................................................................. 321

Table 11.50: Robust Fit Indexes for Client Focus and Involvement Features Construct ........... 323

xx
Table 11.51: Factor Loading and Z-statistics of Client Focus and Involvement Measurement . 324

Table 11.52: Reliability and Construct Validity of Client Focus and Involvement Feature Model
..................................................................................................................................................... 325

Table 11.53: Postulated Company Quality System Evaluation Features Model ........................ 327

Table 11.54: Residual Covariance Matrix for Company Quality System Evaluation Model
(Unstandardized) ......................................................................................................................... 328

Table 11.55: Residual Covariance Matrix for Company Quality System Evaluation Model
(Standardized) ............................................................................................................................. 329

Table 11.56: Robust fit Indexes for Company Quality System Evaluation Features Construct . 330

Table 11.57: Factor Loading and Z-statistics of Company Quality System Evaluation
Measurement ............................................................................................................................... 331

Table 11.58: Reliability and Construct Validity of Company Quality System Evaluation Feature
Model .......................................................................................................................................... 332

Table 11.59: Postulated Company Vision and Plan Statement Features Model......................... 335

Table 11.60: Residual Covariance Matrix for Company Vision and Plan Statement Model
(Unstandardized) ......................................................................................................................... 336

Table 11.61: Residual Covariance Matrix for Company Vision and Plan Statement Model
(Standardized) ............................................................................................................................. 336

Table 11.62: Robust Fit Indexes for Company Vision and Plan Statement Features Construct 338

Table 11.63: Factor Loading and Z-statistics of Company Vision and Plan Statement
Measurement ............................................................................................................................... 339

Table 11.64: Reliability and Construct Validity of Company Vision and Plan Statement Feature
Model .......................................................................................................................................... 340

Table 11.65: Postulated Product Design Management Features Model ..................................... 343

xxi
Table 11.66: Residual Covariance Matrix for Product Design Management Model
(Unstandardized) ......................................................................................................................... 343

Table 11.67: Residual Covariance Matrix for Product Design Management Model (Standardized)
..................................................................................................................................................... 344

Table 11.68: Robust fit indexes for Product Design Management Features Construct .............. 346

Table 11.69: Factor Loading and Z-statistics of Product Design Management Condition
Measurement ............................................................................................................................... 347

Table 11.70: Reliability and Construct Validity of Product Design Management Feature Model
..................................................................................................................................................... 348

Table 11.71: Postulated Product Selection Management Features Model ................................. 351

Table 11.72: Residual Covariance Matrix for Product Selection Management Model
(Unstandardized) ......................................................................................................................... 351

Table 11.73: Residual Covariance Matrix for Product Selection Management Model
(Standardized) ............................................................................................................................. 352

Table 11.74: Robust Fit Indexes for Product Selection and Design Management Features
Construct ..................................................................................................................................... 354

Table 11.75: Factor Loading and Z-statistics of Product Selection and Design Management
Condition Measurement .............................................................................................................. 355

Table 11.76: Reliability and Construct Validity of Product Selection Management Feature Model
..................................................................................................................................................... 356

Table 11.77: Postulated Construction Process Management and Improvement Features Model359

Table 11.78: Residual Covariance Matrix for Construction Process Management and
Improvement Model (Unstandardized) ....................................................................................... 359

Table 11.79: Residual Covariance Matrix for Construction Process Management and
Improvement Model (Standardized) ........................................................................................... 360

xxii
Table 11.80: Robust fit Indexes for Construction Process Management and Improvement
Features Construct ...................................................................................................................... 362

Table 11.81: Factor Loading and Z-statistics of Construction Process Management and
Improvement Measurement ........................................................................................................ 363

Table 11.82: Reliability and Construct Validity of Construction Process Management and
Improvement Feature Model....................................................................................................... 364

Table 11.83: Postulated Construction Employees Involvement and Motivation ....................... 367

Table 11.84: Residual Covariance Matrix for Construction Employees Involvement and
Motivation Model (Unstandardized)........................................................................................... 367

Table 11.85: Residual Covariance Matrix for Construction Employees Involvement and
Motivation Model (Standardized) ............................................................................................... 368

Table 11.86: Robust Fit Indexes for Construction Employees Involvement and Motivation
Features Construct ...................................................................................................................... 370

Table 11.87: Factor Loading and Z-statistics of Construction Employees Involvement and
Motivation Measurement ............................................................................................................ 371

Table 11.88: Reliability and Construct Validity of Construction Employees Involvement and
Motivation Condition Feature Model ......................................................................................... 372

Table 11.89: Postulated TQM Features Model ........................................................................... 375

Table 11.90: Residual Covariance Matrix for TQM Model (Unstandardized) .......................... 375

Table 11.91: Residual Covariance Matrix for TQM Model (Standardized) ............................... 376

Table 11.92: Robust Fit Indexes for TQM Features Construct .................................................. 378

Table 11.93: Factor Loading and Z-statistics of TQM Measurement ........................................ 379

Table 11.94: Reliability and Construct Validity of TQM Feature Model .................................. 380

Table 11.95: Reliability and Construct Validity of the Latent Variables ................................... 381

xxiii
Table 11.96: Robust Fit Indexes for Structural Model 2.0 ......................................................... 388

Table 11.97: Model 2.0 Factor Loadings and Z-statistics .......................................................... 392

Table 11.98: Model 2.0 Factor Loadings, Z-statistics, Variance Accounted for & Reliability and
Construct Validity ....................................................................................................................... 395

xxiv
 LIST OF FIGURES
Figure 2.1: TQM: Building blocks ............................................................................................... 41

Figure 2.2: TQMEF (TQM-Efficiency) Model ........................................................................... 42

Figure 2.3: Simple Model ............................................................................................................ 43

Figure 2.4: TQM Framework ....................................................................................................... 43

Figure 4.1: Construction Organisation System ............................................................................. 61

Figure 5.1: Evolution of Quality Management ............................................................................ 74

Figure 8.1: Visual Model of Mixed Methods Design ................................................................. 135

Figure 8.2: Steps in the research design process......................................................................... 139

Figure 8.3: Research Design Outline .......................................................................................... 144

Figure 8.4: The Delphi process .................................................................................................. 161

Figure 8.5: Outline of Delphi Process......................................................................................... 168

Figure 110.1: Integrated TQM Conceptualised Model (Model 1.0) ........................................... 256

Figure 11.1: Measurement Model of Leadership and Top Management .................................... 302

Figure 11.2: Measurement model of Company Supplier Quality Management ......................... 311

Figure 11.3: Measurement Model of Client Focus and Involvement ......................................... 319

Figure 11.4: Measurement Model of Company Quality System Evaluation .............................. 326

Figure 11.5: Measurement Model of Company Vision and Plan Statement .............................. 334

Figure 11.6: Measurement Model of Product Design Management ........................................... 342

Figure 11.7: Measurement Model of Product Selection Management ....................................... 350

Figure 11.8: Measurement Model of Construction Process Management and Improvement .... 358

Figure 11.9: Measurement Model of Construction Employees Involvement and Motivation ... 366
xxv
Figure 11.10: Measurement model of TQM ............................................................................... 374

Figure 11.11: Hypothesized Model of TQM (Model 1.0) .......................................................... 385

Figure 12.12: Model 2.0 – An Integrated TQM Model .............................................................. 400

Figure 11.13: Model 2.0 – An Integrated TQM Model Covariances Association ..................... 401

xxvi
 LIST OF APPENDICES
APPENDIX A: Invitation Letter to Participate in a Delphi Study
APPENDIX B: Request to Experts to Submit their Curriculum Vitae
APPENDIX C: Delphi Instructions for Round 1 and Questionnaire
APPENDIX D: Delphi Instructions for Round 2 and an Example of Completed
Questionnaire Showing Group Medians
APPENDIX E: Delphi Instructions for Round 3
APPENDIX F: Introduction Letter and Questionnaire for Industry Survey
APPENDIX G: Model 2.0 Covariance Matrix (Unstandardized)
APPENDIX H: Model 2.0 Covariance Matrix (Standardized)

xxvii
 LIST OF ABBREVIATIONS
AAA Aggrissive Affirmative Action
ADF African Development Fund
AIT Asian Institute of Technology
ANC African National Congress
ANOVA Analysis of Variance
BSc Bachelor of Science Degree
BTech Bachelor of Technology Degree
CDM Construction Design and Management
CEIM Construction Employees Involvement and Motivation
CFA Confirmatory Factor Analysis
CFI Client Focus and Involvement
CFI Comparative Factor Index
CI Construction Industry
CI Confidence Interval
CIDB Construction Industry Development Board
CII Construction Industry Indicators
CIOB Chartered Institute of Building
CIP Contractor Incubator Programme
CPMI Construction Process Management and Improvement
CQM Construction Quality Manager
CQSE Company Quality System Evaluation
CSFs Critical Success Factors
CSIR Council for Scientific and Industrial Research’s
CSQM Company Supplier Quality Management
CVPS Company Vision and Plan Statement
DF Degree of Freedom
DK Building and Civil Engineering Contractors
DPhil/PhD Doctor of Philosophy Degree
DPW Department of Public Works
DSc Doctor of Science Degree

xxviii
DTI Department of Trade and Industry
ECDP Emerging Contractor Development Programme
EFA Exploratory Factor Analysis
EFQM European Foundation for Quality Management
EPWP Expanded Public Works Programme
EQA European Quality Award
EQS Equations
ESDS Employment and Skills Development Services
EU European Union
FV Full Variable Model
GCI Ghanaian Construction Industry
GDP Gross Domestic Product
GFI Goodness-of-Fit Index
GH¢ Ghanaian Cedi
GLM General Linear Modelling
GSS Ghana Statistical Service
HND Higher National Diploma
ICT Information Communication Technology
IGM Interacting Group Method
IQD Inter-Quartile Deviation
ISO International Standard Organisation
JIT Just-In-Time
LM Lagrange Multiplier
LTM Leadership and Top Management
MAD Average (Mean) Absolute Deviation
MANOVA Multivariate Analysis of Variance
MBNQA Malcolm Baldrige National Quality Award
MCAR Missing Completely at Random
MSc Master of Science Degree
MTech Master of Technology Degree
N Sample Size

xxix
n.d. No Date
NDPW National Department of Public Works
NFI Normal Fit Index
NGT Nominal Group Technique
PDCA Plan, Do, Check, Act
PSDM Product Selection and Design Management
QC Quality Control
QA Quality Assurance
QI Quality Inspection
QMS Quality Management Systems
RDP Reconstruction and Development Programme
RML Robust Maximum Likelihood
RMSEA Root Mean Square Error of Approximation
S- Bχ2 Satorra-Bentler Scaled Chi-Square
SAPOA South African Property Owners’ Association
SEM Structural Equation Modelling
SETA Sector Education and Training Authority
SME Small and Medium Sized Enterprises
SON Standard Organization of Nigeria
SPC Statistical Process Control
SPSS Statistical Package for Social Sciences
SRMR Standardised Root Mean Square Residual
STATKON Statistical Consultation Service
Stats SA Statistics South Africa
TLI Turker Lewis Index
TOC Theory of Constraints
TQC Total Quality Control
TQM Total Quality Management
TQMEF Total Quality Management Efficient Model
UJ University of Johannesburg
UK United Kingdom

xxx
US$ United States Dollar
USA United States of America
WLS Weighted Least Squares

xxxi
 CHAPTER ONE
1 INTRODUCTION
1.1 BACKGROUND
The construction industry provides physical infrastructural and contributes to national socio-
economic development in every country (Ofori, 2012) as well as creating a great deal of
employment in both the public and private sector. It employs and utilizes up to 10% of the
working population, contributes between 5 and 10% of the Gross Domestic Product (GDP) in all
countries and is accountable for about half of the gross fixed capital formation (Ofori, 2012).
Hence, the construction industry is considered as an economic backbone and major contributor to
the GDP. On the other hand, Ofori (2012) emphasized that a period of low construction output
can adversely affect the expansion of the general economy. In Ghana, the construction sector
contribution to the GDP has shown an increasing trend from 8.5% to 11.8% from 2010 to 2013
respectively (Vibe Ghana, 2014) and this shows a sign of its growing importance in the
development of the nation. The Ghana Statistical Service (GSS) report (2012) showed that the
major contributors to the industry sector in Ghana are the construction sector, mining and
quarrying, and water and sewerage activities, which recorded a growth of 19.2%, 5.2% and 0.6%
respectively.

Even though the construction sector in Ghana has much strength, there are numerous problems
as indicated by Ofori (2012) that have persisted for a long period. Foster and Pushak (2011)
added that in spite of Ghana’s achievement with increasing access regarding infrastructural
services, the quality of service remains low. Nuamah, Manu and Manu (2013) supported this
assertion and opined that the level of construction in Ghana is usually poor, particularly, in the
public sector. Moreover, Buertey, Amoa and Adjei-Kumi (2011) asserted that about 95% of
projects carried out in Ghana experience schedule delays. Nicco-Annan (2006) conducted a
survey of the construction of office structures in Accra, Ghana, and discovered that the cost of
the structures overruns was between 60% and 180% without taking inflation into consideration.
There were also time overruns of between 12 and 24 months and in a few situations, the
structures were still not able to be used because of some significant deficiencies. As a result of
poor quality management, a great deal of time, money and resources are wasted on construction
projects. Understandably, clients in the construction sector, which includes the Ghanaian

1
construction sector, have become increasingly disappointed. What they see is irregularity and
under-performance. What they obtain is often poor quality, overpriced and delayed projects
(Ansah, 2011).

It is therefore very important that the construction sector in Ghana is improved to produce
quality works in order to augment the growth of the national economy. It is also important for the
construction companies in Ghana to provide consistent value and quality regarding the products
that they construct in order to remain competitive in today’s construction environment. One of
the ways of accomplishing this is by the application of TQM principles in the construction
industry. However, the construction industry in Ghana has been slow in the implementation of
TQM technique, probably owing to the reluctance to change old management techniques.
Rategan’s (1992) study indicated that a 90% enhancement rate in employee relations, customer
satisfaction, financial performance and operating procedures is realized due to the
implementation of TQM. Moreover, studies by Low and Jasmine (2004) on TQM confirmed a
reduction in cost and better employee satisfaction that eventually boost client satisfaction.

Arditi and Gunaydin (1997) argued that TQM is a philosophy that involves all parties in the
organization in the attempt to enhance performance. It spreads throughout every aspect of a
company and makes quality a strategic purpose. Harrington and Voehl (2012) also indicated that
TQM is a culture that strives for customer satisfaction through innovation and continuous
improvement in every segment of the organization. The definitions of TQM are numerous,
depending on the organizations’ perception of quality. The definition of TQM in the BS 7850 is
the company practices and management philosophy that seek to control the human and material
resources of an organization in the most efficient way to achieve the goals of the organization
(BSI, 1992). A report from the Asian Institute of Technology (AIT) indicates that TQM is a
philosophy that reinforces the culture to foster continuous organizational enhancement through
systematic, integrated, constant effort involving everything and everyone, focusing mainly on
total satisfaction of internal and external customers, where employees work collectively in teams
with process ownership, directed by a dedicated top management, which takes a proactive
involvement (Nukulchai, 2003). On the other hand, British Quality Association defines TQM as
a comprehensive business management philosophy focusing on entirely satisfying customer

2
requirements with a maximum of efficiency and effectiveness (Wessel & Burcher, 2004).
Another meaning of TQM which is utilized by USA Defence Department is that it is a
philosophy and a set of directing concepts that signify the basis of a continually improving
organization (Tingey, 1997).

Tang, Ahmed, Aoieoung and Poon (2005) indicated that the five core principles of TQM are
continuous improvement, customer focus, employee participation, teamwork, and process Focus.
These five principles are related to the five drivers of change suggested by the Egan (1998)
report. These are committed leadership, integrated processes and teams, focus on the customer,
commitment to people and quality driven agenda. Harris and McCaffer (2002), on the other
hand, posited that TQM consists of all activities that managers carry out to improve their quality
and strategy such as quality control, quality planning, quality improvement and quality
assurance. It includes continuous improvement, training and re-training of staff, customer
satisfaction, top management support, defect-free products at first attempt, elimination of
reworks, and cost-effectiveness, amongst others.

Several researchers have verified the benefits and usefulness of TQM in the construction
industry (Ahmed, 1993; Arditi & Gunaydin, 1997; Low & Jasmine, 2004; Harrington & Voehl,
2012). The benefits of TQM recognized by these researchers include improved employee job
satisfaction, higher customer satisfaction, reduction in construction costs, improved schedule
performance, improved relationships with subcontractors, reduced rework, improved safety,
higher productivity, lower employee turnover, speeding up construction work. Other benefits of
TQM recognized by these researchers include improved methods of working, better control over
the construction process, gaining competitive advantage, increase profitability, decreasing waste
and rework, more customers focused and better coordination of activities. Furthermore, their
studies confirmed that the implementation of TQM may decrease the severity and frequency of
schedule overruns.

Despite numerous benefits associated with TQM implementation in most developed countries,
the principles of TQM are not employed in the construction industry in Ghana and this has led to
high construction costs, low quality works and dissatisfied clients. In the quest of performance

3
quality in the construction industries and with a growing awareness of the construction
excellence, construction firms have no choice other than to ensure that clients obtain value for
money. Hence, it is important that TQM is embraced in the Ghanaian construction industry in
order to assist raise productivity and quality. Foster and Pushak (2011) suggested that Ghana
may benefit from a systematic framework for regulating the quality of public services. However,
such an objective demands that a continuous improvement process be instituted within the
company in order to provide quality management. The continuous improvement process is
referred to as TQM.

Lad and Beck (2009) emphasized that construction quality management such as TQM is the
solution to an effective project delivery. Steyn, Basson, Carruthers, Du Plessis, Prozesky-
Kuschke, Kruger, Van Eck, and Visser (2004:189) argued that quality never happens by itself: it
is always an outcome of careful research into the requirements of the deliverables that will meet
all the needs of the customers as well as the expectations of the stakeholders involved. Therefore
the application of a TQM system in the Ghanaian construction may require extensive research to
establish contributing and deliverable variables for its successful implementation. It should,
however, be noted that a great deal of research has been carried out throughout the world on
TQM. Despite extensive study on TQM in other countries, there is a scarcity of research on
TQM in the Ghanaian construction industry. As a follow-up to the previous studies, this research
was therefore undertaken to identify as much as possible the critical success factors that will help
to develop a TQM model to guide TQM implementation in the Ghanaian construction industry,
based on recognized TQM principles and existing frameworks. The study also discusses the
implication of its findings on TQM and makes recommendations to improve the existing
strategies of project management practices in Ghana.

1.2 THE RESEARCH PROBLEM STATEMENT

Performance of the Ghanaian construction industry is a major cause of concern amongst clients
and stakeholders in the industry (Kwaw, Yalley, Cobbinah & Opintan-Baah, 2011). The
background to this study has established that both private and public sector clients of the
construction industry continue to complain about the industry’s performance and its seemingly
inability to deliver projects on time, within budget and to expected quality standards. Also, it was

4
recognized that major clients are dissatisfied with consultants’ performance in co-ordinating
teams, in design and innovation, in providing a quick and reliable service and in providing value
for money. The World Bank report (2004) indicates that in many developing countries, major
investments are not generating the quality of the services demanded. Also, many of the
infrastructure projects in these countries run over budget and take far longer than necessary to
complete. Anvuur, Kumaraswamy and Male (2006) posit that both consultancy services and
works in Ghana take extremely long to reach financial closure and are subject to unnecessary
delay. This is because little attention has been given to TQM with regard to its implementation in
the Ghanaian construction industry.

It has also been revealed that there is increasing global competition in the construction sector.
Therefore, to remain competitive, Ghanaian construction firms have to focus their business
strategies on strategic advantages through the improvement of business excellence and
performance. Hence, in the pursuit of performance quality in the Ghanaian construction
industries, there is a necessity for an efficient management system such as TQM. It has been
emphasized that TQM practice is necessary for long-term survival in all business, including the
construction industry. Nonetheless, in most developing countries and taking into consideration
the Ghanaian construction industry, the principles of TQM are not employed owing to lack of
knowledge and misunderstanding of its implementation. This has been confirmed by Imbeah
(2012). Although real estate firms in Ghana are aware of the significance of quality, their
knowledge about TQM is inadequate. Hence, it is essential that managers of construction firms
have adequate knowledge of the management of time, quality, risk, and human resources,
amongst others, in a bid to ensure a successful project completion and to meet clients’ desires
and expectations (Adusa-Poku, 2014).

As stated earlier, to enhance the quality of works and to reduce the cost of construction, the
construction industry in Ghana needs proactive management methods (Western-oriented
techniques) such as TQM. Unfortunately, such techniques of project management are not
straightforward procedures that anyone can learn and implement. This is because of substantial
cross-cultural problems and lack of understanding among construction professionals in applying
the techniques in non-Western countries. Hence, the need for empirical evidence on reasons for

5
lack of understanding and failure to use quality management techniques such as TQM in the
Ghanaian construction industry and perhaps in other African countries with similar conditions,
particularly third world countries, cannot be overemphasized. It should, however, be admitted
that a great of research has been carried out throughout the world on TQM. Regardless of the
extensive research on TQM in other countries, there is a lack of study on the implementation of
TQM in the Ghanaian construction industry. This suggests that the application of a TQM system
in the Ghanaian construction industry may require extensive research to establish contributing
and deliverable variables for its successful implementation. This is important because quality
always depends on an outcome of careful research into the requirements of the deliverables that
will meet all the needs of the customers as well as expectations of the stakeholders involved.
Based on the empirical evidence, recognized TQM principles and existing frameworks
established in the study, a holistic TQM model can then be developed for the Ghanaian
construction industry.

The problem that was resolved in this research is as follows: Given that the past designs of TQM
established in the western world cannot be depended upon in developing nations, and the results
of what decides TQM ideas in construction industry in developing nations are not known from
the previously performed analyses as well as the lack of analysis into the overall effect of the
direct and natural effective involvement of TQM constructs, the achievement of more consistent
and high quality products that meet client/customer requirements in the construction industry is
unlikely. The above problems are addressed in this study.

1.3 AIM OF THE STUDY

The aim of this study was to develop a holistic integrated Total Quality Management model for
the Ghanaian construction industry. The study determined the critical factors that influence TQM
implementation in the construction industry. The identified critical factors and the recognized
TQM principles and existing frameworks guided the development of the integrated model for the
Ghanaian construction industry. This integrated model will help construction companies in
Ghana to offer more reliable quality and value to their customers/owners. Considering the
increasing competitiveness in the construction sector and the need to be competitive in today’s

6
market, it is imperative that Ghanaian construction companies establish appropriate quality
management techniques to reduce cost and increase the value or quality of their projects.
Therefore an integrated TQM model will be an appropriate quality management technique in this
regard. TQM can be a key element of a successful business but, if not properly implemented, can
also lead to failure. Hence, the main objective of an integrated TQM model is to accomplish
client objectives through continuous improvement of products and operations by full
participation and commitment of all experts who are part of that product process.

1.4 RESEARCH MOTIVATION

The image of most contractors in the Ghanaian construction industry has been dented owing to
their inability to fulfill their clients’ specifications. This has led to a loss of the level of clients’
satisfaction. It has therefore become relevant for contracting organizations in Ghana to establish
a quality management system that will help to bring projects to a satisfactory completion within
time, cost limitations and maintaining quality standards, with greater improvement in quality of
construction, construction processes, and the level of clients’ satisfaction.

TQM has been successfully employed in the Western economies but the question is whether this
technique is applicable in Ghana. The growing weight of empirical evidence from cross-cultural
management research suggests that Western management concepts may be wholly or partially
inapplicable and irrelevant in other cultures. In addition, the use of such tool and technique in
Africa with Ghana as a sub-region will not enhance project success if they run counter to cultural
and work values. For the very reason that values at work and in social settings are culturally
base. When dealing with human behaviour (i.e. managing) we must recognize the cultural
context. The recognition of economic rationality and efficiency, assumed as basis for many
project management tools and techniques does not reflect local realities.

Some quality gurus and project management writers such as Crosby (1979), Ishikawa (1985),
Deming (1986), Feigenbaum (1991), and Juran (Juran and Gryna, 1993) in the Western
Countries have developed certain propositions in the field of TQM, which have gained
significant acceptance throughout the world. Their insights provide a good understanding of the

7
TQM philosophy, principles, and practices which they belief are straight forward procedures that
anyone can learn and implement.

Contrary to the common belief that the Western-oriented techniques of project management such
as TQM are just simple techniques that anyone can learn and apply, there are considerable cross-
cultural problems and lack of understanding among construction experts in using the strategies in
non-Western nations. Hence, the major elements that this research needs to consider and build
into the proposed model are Company Vision and Plan Statement, Product Selection and Design
Management, and Company Quality System Evaluation. It is important to adopt a system that
offers a platform for these aspects to be incorporated and addressed. The motivation behind this
research therefore was to establish techniques that could be employed to fill the gap in the
Ghanaian construction industry.

1.5 SIGNIFICANCE OF THE STUDY

There is a substantial lack of understanding among construction professionals and individuals of
the quality systems and tools that could be implemented in the Ghanaian construction industry to
improve not only the end products, but also the processes and the overall system. It should
however be noted that there are a number of techniques, such as six sigma, just-in-time and TQM
that could be employed to fill that gap in the Ghanaian construction industry. These techniques
were used elsewhere to improve the processes and products, ultimately resulting in more
satisfied customers, which then led to increased productivity and profitability of the companies.
TQM as a holistic or totality approach to quality represents a better way of solving the issues that
are currently being face in the Ghanaian construction sector. Hence, the importance of this study
was to develop an integrated TQM model to guide the implementation and the use of quality
management concepts in the Ghanaian construction industry. Based on the proposed model,
building professionals and individuals in the Ghanaian construction industry will have a better
way of solving the problems that are related to construction quality. In addition, the data
collected from the field survey can be used as a basis for future research as well as in
longitudinal surveys.

8
1.6 THE RESEARCH
1.6.1 Research Questions
Based on the research problem statement and aim of the research, the following research
questions emerged:
RQ1. What factors determine TQM implementation in the construction industry?
RQ2. What are the gaps in theories and literature on TQM implementation in the
construction industry?
RQ3. What are the main and sub-attributes of TQM in the Ghanaian construction industry?
RQ4. What critical factors affect construction firms for not implementing TQM in Ghana?

RQ5. What constructs are required to reinforce the achievement of an integrated TQM model
for Ghanaian construction industry?
RO6. To what extent does the hypothesized integrated TQM model fit into the identified
model?

1.6.2 Research Objectives

In order to provide answers to the research questions and achieve the aim of the research, the
following objectives are set:
RO1. To establish the factors that determines TQM implementation in the construction
industry;
RO2. To investigate theories and literature on TQM implementation in the construction
industry and identify the gaps that needs to be considered;
RO3. To determine the main and sub-attributes attributes of TQM in the Ghanaian
construction industry;
RO4. To evaluate the critical factors that affect construction firms for not implementing
TQM in Ghana;
RO5. To develop an integrated TQM model to guide construction firms in managing
construction projects in Ghana; and

RO6. To determine the validity of the integrated TQM model through a comparative analysis
of the Delphi and literature outcome with the field questionnaire survey outcome.

9
1.6.3 Research Methodology
Different methods of investigation were used in the research, such as quantitative and qualitative
research designs, usually referred to as mixed-method design. This approach was used to answer
research questions and meet the research objectives, as well as developing an integrated TQM
model that applied to the study areas. The qualitative method used was made up of structured
and semi-structured (using an interview guide) interview. This was possible through the use of
the Delphi technique. The findings from this section helped to refine the survey tool (structural
questionnaire) for the study. The Delphi technique was used to resolve conflicting issues on
TQM in the study areas through consensus. With regard to the quantitative aspect of the study, a
face-to-face administered questionnaire survey was conducted among Building and Civil
Engineering Construction companies in Ghana. Data gathered via the questionnaire survey was
analyzed using Structural Equation Modelling (SEM) software Version 6.2, which was used to
assess the factor structure of the constructs. The conceptual variables were tested as a prior using
the SEM of the questionnaire survey results. The SEM process was undertaken as Confirmatory
Factor Analysis (CFA) of the prior model for an integrated TQM for the Ghanaian construction
industry.

The research objectives were achieved through the following methods: Research objectives RO1
and RO2 were achieved through conducting a literature review on the subject in question.
Research objectives RO3 and RO4 were achieved by conducting a Delphi survey. Research
objective RO5 was achieved by conducting a field questionnaire survey.

Detail techniques used in the study were given under the research methodology section:

i. Quantitative Technique

Quantitative research is about quantifying the relationships between variables and finally
measuring them. Statistical models were constructed to explain the observed variables. Certain
characteristics (variables) were searched for and the study endeavoured to demonstrate
something significant about how they were allocated within a certain population. The nature of
the research determined the variables in which the researcher is interested. A variable needed to
be measured for the purpose of quantitative analysis. Data was collected concerning many

10
variables through a questionnaire survey. The variables in which the researcher was interested
were dependent or independent.

ii. Qualitative Technique

The qualitative research technique allows the topics being researched to offer much ‘richer’
solutions to questions put to them by the researcher and can offer useful ideas which might have
been missed by any other technique. Not only does it offer useful information to certain research
questions in its own right, but there is a strong case for using it to complement quantitative
research methods. For example if the area of interest has not been previously investigated, then
qualitative research may be a vital forerunner. In conducting any quantitative research; for
example, it is impossible to carry out a meaningful structured questionnaire survey if the
important are not known. At the other extreme, qualitative research may also help to understand
the findings of quantitative research.

iii. Structural Equation Modelling (SEM)

Structural equation modelling (SEM) represents an extension of General Linear Modelling

(GLM) techniques, such as the Analysis of Variance (ANOVA) and multiple regression analysis.
SEM was used to study the connections among latent constructs that are indicated by multiple
measures. Sampling size is usually large (N >200) and is somewhat dependent on model
complexity, the estimation method to be used, and the distributional characteristics of observed
variables (Kline, 2005). The evaluation of two models from the Delphi survey and literature
enabled the development of the TQM model for the Ghanaian construction industry.

iv. Mixed-Method Research

Mixed-techniques research includes the use of more than one approach to or method of design,
data collection or data analysis within a single programme of research, with the incorporation of
the different approaches or techniques happening during the programme of research, and not just
at its finishing point (Johnson, Onwuegbuzie, Turner & 2007:120). Mixed-methods research is
an investigation technique with philosophical presumptions as well as techniques of inquiry. As a
technique, it includes philosophical presumptions that guide the route of the gathering and
analysis of data and the mixture of qualitative and quantitative techniques in many stages in the

11
study process. As a method, it concentrates on gathering, examining, and combining both
quantitative and qualitative data in a single study or series of studies. Its main assumption is that
the use of quantitative and qualitative techniques, in mixture, provides a better understanding of
research problems than either strategy alone (Creswell & Clark, 2007). The quantitative data in a
typical mixed-method includes closed-ended questions such as those discovered about mind-set,
behaviour, or performance instruments. The collection of this kind of data involves using
closed-ended guidelines, against which the researcher ‘checks’ the observed behaviour.
Sometimes quantitative information is discovered in records, such as attendance records or
census records. The analysis of the quantitative data consists of statistically examining ratings
gathered on instruments and checklists to test hypotheses or to answer research questions
(Creswell, 2003; 2010).

In contrast, typical qualitative data in a mixed methodology study consists of open-ended

information that the researcher gathers through interviews with participants. The general open-
ended questions to be used during interviews allow the participants to supply answers in their
own words. These can be thematically analysed and converted into qualitative data, which can
also be transcribed in quantitative data, such as the case of Delphi Studies where frequencies of
measures of central tendencies are used to draw consensus. Also, qualitative data may be
collected by observing participants or sites of research, gathering documents from a private or
public source, or collecting audiovisual materials, such as videotapes or artifacts. The analysis of
the qualitative data (words or text or images) typically follows the path of aggregating the words
or images into categories of information and presenting the diversity of ideas gathered during
data collection (Creswell, 2010).

The open- ended versus closed-ended nature of the data differentiates between the two types
better than the sources of the data. The combination of both approaches (qualitative and
quantitative) can offset the weaknesses of either approach when used itself. For instance, mixed
method research provides more comprehensive evidence for studying a research problem than
either quantitative or qualitative research alone. Also, researchers are given permission to use all
of the tools of data collection available, rather than being restricted to the types of data collection
typically associated with qualitative or quantitative research (Creswell, 2003). Further, Mixed
Method research helps answer questions that cannot be answered by qualitative or quantitative

12
approaches alone. It is also very practical because the researcher is free to use all methods
possible to address a research problem. It can be used to increase the generalisability of the
research result, which in this present day and age is a major consideration. It can also provide
stronger evidence for a conclusion through convergence and verification of findings.

The review of literature on TQM provides the background to the study. Various sources were
reviewed including books, articles in accredited journals, published and unpublished works such
as dissertations, and web-based publications on the specified field of study. Two methods were
used to collect the empirical data. These methods include the Delphi and field method or
Questionnaire survey method. Detailed descriptions of both these methods are presented in the
methodology chapter of the thesis. According to the Delphi technique, data was obtained through
the use of structured questionnaires. Experts were asked to complete the questionnaires and reach
consensus on the rated likelihoods and impact of various TQM factors. The process involved a
three-round iterative process with the main aim of getting experts to reach consensus on the
questions raised in the questionnaires. Experts were encouraged to give reasons for their
dissenting views. The Delphi technique was used to explore the level to which the core attributes
or sub-factors effect or affect TQM in the construction industry.

Data that was collected in the questionnaire survey was the evidence of factors of TQM concept,
namely leadership and top management, company supplier quality management, client focus and
involvement, company quality system evaluation, company vision and plan statement, product
selection and design management, construction process management and improvement, and
construction employees’ involvement and motivation. In addition, data regarding TQM at
project level was also required. The TQM was determined from perceptions of respondents. This
included all top management working in construction firms in Ghana. The top management
considered were those working on on-going projects. The data was obtained by means of
questionnaires.

i. Data Sources

In the Delphi study, data regarding the ratings of the impact of factors on TQM was obtained
from the expert panel. On the other hand, data from the questionnaire survey was obtained from

13
top management working either for construction clients, designers or contracting organisations.
The data related to on-going construction projects.

ii. Data Analysis

Data obtained from the Delphi was analysed with Microsoft EXCEL spreadsheet software. The
output from the analysis was a set of descriptive statistics such as means, median, standard
deviations and derivatives of these statistics. SEM was utilized using the EQS software packages
to analyse data obtained from the field questionnaire survey. Outputs from the analysis were
univariate and multivariate descriptive statistics as well as measures of goodness-of-fit of the
hypothesised model. Other outputs included measures of statistical significance of parameter

estimates.

1.7 CONTRIBUTION TO KNOWLEDGE

Generally, literature provided knowledge on TQM but most often these research studies were
carried out in environments that were totally different from those of the developing countries and
Ghanaian conditions. Hence, this research identified and evaluated the critical factors and issues
that affect TQM implementation in the Ghanaian construction industry. After reviewing the
existing TQM literature and critically analyzing the research results, it became clear that this
study was the only one that consistently investigates the consequences and factors affecting
TQM execution in Ghanaian construction companies.

In this research, new knowledge was generated from existing TQM knowledge integrated with
specific characteristics of Ghanaian construction firms. This study has therefore delivered in-
depth understanding of TQM implementation in the construction industry. Thus, new knowledge
related to TQM implementation in construction firms has been derived. Specifically, the
following contributions to knowledge were achieved:
• Extensive review of TQM literature from both developed and developing countries;
• Main factors and principles that underpinned TQM within the construction industry were
determined:

14
 • Critical factors and issues that impact TQM execution in Ghanaian construction firms
were determined; and
• An integrated TQM model to guide construction firms in managing construction projects
in Ghana was developed.

1.7.1 Value of the Study

The integrated TQM model which has been developed by this study will help construction
companies in Ghana to offer more reliable quality and value to their customers or owners. It will
also represent a more effective way of solving the problems that are related with construction
quality and that are currently being face in the Ghanaian construction sector. An integrated TQM
is a process of getting rid of poor quality from production rather than getting rid of poor quality
products. It is a perspective that includes everyone in a company in constant efforts to improve
quality and achieve customers’ satisfaction. Continuous enhancement is the perspective that
seeks to make continuous developments to the process of transforming information into results.
Several researchers have revealed that the implementation of TQM concepts can allow
companies to compete globally (Womack, Jackson & Roos, 1990; American Quality Base and
Ernst & Young, 1991;Easton, 1993; Handfield, 1993; Hendricks & Singhal, 1996, 1997). Hence,
with increasing global competition in the construction industry, the Ghanaian construction firms
which are expected to participate on this global scale need to adopt the integrated TQM model
which has been developed by this study to enable them to generate quality works, to be more
effective, and to increase their competitive advantage.

1.7.2 Delimitations
This study concentrates on the TQM in the context of the Ghanaian construction industry. The
reason for the selection of this area for the research is the role that the construction industry
playss in the provision of infrastructure and development of the Ghanaian economy. The
research concentrated on areas where quality in the construction industry needs to be applied
such as workmanship, processes and materials because poor quality results in any of these areas
can potentially lead to significant losses throughout the industry. Basically, the study is
delimited to TQM in the Ghanaian construction industry with emphasis on the principles and
practices of TQM, construction professionals’ and top management’s level of understanding and
15
usage of TQM principles, as well as critical factors and issues that affect implementation of
TQM. Finally, an integrated TQM model was developed to guide construction firms in managing
construction projects in Ghana.

1.8 ETHICAL CONSIDERATION

Ethical issues were a key consideration in undertaking this research. The principle of voluntary
participation was upheld. This requires that people would not be coerced into participating in
research. Furthermore, participants were only involved in the research when they had given
informed consent. Great effort was made to help protect the privacy of research participants by
ensuring confidentiality by not making available identifying information to anyone who is not
directly involved in the study. Confidentiality was further enhanced by keeping participants
anonymous throughout the study. The study will also be subject to independent review by the
promoters to help protect all participants and the researcher against potential legal effects of
failing to address important moral issues or to uphold integrity, honesty and quality assurance.

1.9 OVERVIEW OF CHAPTERS

The compilation of the entire research project was organised as follows:

Chapter 1 – Introduction

The introductory chapter provides the background to the subject. It also communicates the
importance of the research, problem statement, and guiding questions that were investigated. The
delimitation, rationale/motivation of the study, purpose and objectives of the study were also
presented in this section.

Chapter 2 – Theoretical and Conceptual Perspectives of TQM Research

This chapter reviews theories and concepts applied in this research regarding quality
management practices and improvement measurement.

Chapter 3 – Gaps in the TQM Research

This chapter addresses the gaps observed in TQM research.

16
Chapter 4 – The Construction Industry and TQM

This chapter presents an overview of the construction industry and TQM in general. The chapter
discusses types of construction undertaken by the construction industry and the working
environment in the construction industry. This is significant because the working environment in
the construction industry is crucial to its success. This chapter also describes how TQM is
implemented in the construction industry and its associated issues such as benefits and critical
success factors (CSFs) affecting its implementation.

Chapter 5 – TQM Implementation in Developed Countries

This chapter discusses TQM implementation in developed countries. The chapter specifically
examines TQM implementation in China. The chapter also discusses the evolution of TQM and
its implementation in developed countries. An evaluation of policies and government
intervention with regard to TQM implementation is also presented. In addition, the views of
quality pioneers in developed countries and their contribution towards the TQM concepts are
highlighted in this chapter.

Chapter 6 – TQM Implementation in Developing Countries

This chapter discusses TQM implementation in developing countries. The chapter examines the
construction industry in Nigeria and South Africa and how the TQM concept has been accepted
in these countries. An evaluation of policies and government intervention with regard to TQM
implementation is also presented. Furthermore, the philosophical basis for TQM implementation
in Nigeria and South Africa is discussed. Also explored in this chapter are the challenges facing
TQM implementation in the Nigerian and South African construction industries.

Chapter 7 – TQM Implementation in the Ghanaian Construction Industry

This chapter discusses TQM implementation in Ghanaian construction industry. The chapter
looks into the construction industry in Ghana and how the TQM concept has been accepted by
Ghana as a country. An evaluation of policies and government interventions with regard to
quality management implementation is also presented. In addition, the philosophical basis for

17
TQM implementation in Ghana is discussed. Also explored in this chapter are the challenges
facing TQM implementation in the Ghanaian construction industry.

Chapter 8 – Research Methodology

This chapter presents an outline of the way in which the research was designed and conducted. It
describes the research tools and their design, the method for the data collection, the research
technique used, the data treatment, the population, and the sampling design. It also indicates how
the results are interpreted.

Chapter 9 – Results from Delphi Study

This chapter presents the results of the data analysis from the Delphi study.

Chapter 10 – The Conceptual Integrated TQM Model

This chapter presents the conceptual integrated TQM model of the study.

Chapter 11 – Survey Analysis

The chapter on survey analysis presents results of the empirical study. They are presented as
statistical measures in literature, tables, charts and graphs. This chapter also presents the
development of a TQM model for a holistic approach to construction quality management based
on the outcomes of the study.

Chapter 12 – Discussion of Survey Results

This chapter presents discussions and an analysis of the findings presented in Chapter 11.
Discussions are relative to the guiding questions. The chapter also presents test results and the
analysis from the model test surveys.

Chapter 13- Summary of Findings, Conclusion and Recommendations

The summary presents summaries of all results from the literature, as well as the empirical study
and the model validation surveys. Conclusions are made on the entire study. Based on the
conclusions drawn from the study, recommendations are also presented in this chapter.

18
1.10 CONCLUSION
The first step in conducting the study is to acquire an understanding of the research study.
Chapter 1 therefore introduced the subject of the research study and gave insight into the
structure, the background and significance thereof. An outline of the technique adopted for the
research was also provided. Furthermore, it relayed information on how the research report was
presented. The perceived contribution to the body of knowledge has been provided as well. The
motivation for the study was that TQM is a major challenge to the construction industry. It was
noted that construction industry in Ghana is still lagging behind in terms of TQM. Therefore, it is
necessary to develop an integrated TQM model to guide the construction contractors and
construction professionals/managers in their daily routine in the management of construction
projects.

19
 CHAPTER TWO
2 THEORETICAL AND CONCEPTUAL PERSPECTIVES OF TQM RESEARCH

2.1 INTRODUCTION
This chapter discusses the review of theories and concepts applied in this research and followed
with quality management practices and improvement measurement. To understand the theory
behind quality management, it is necessary to be familiar with the process and how the
interacting elements within the TQM theory work. This is explained in detail in this chapter.

2.2 THEORY AND DEFINITION

2.2.1 What is Theory?
Several definitions of theory exist within and across disciplines as indicated by Hamilton (1997).
Moreover, Kinloch (1977), Silva (1977) and Dublin (1978) opined that there are many relevant
terms such as conceptual framework, axiom, assertion, prediction, postulate, forecast,
proposition, system, theoretical design and typology. These terms are occasionally used as
synonyms for theory but have particular and often different definitions. The link of theory to
research has been commonly argued and several definitions of theory are abundant. There is a
universal accord that theory is important to consistently arrange and synthesise information, find
out connections among variables, and guide the finding of new information to move research
forward (Touliatos & Compton, 1988; Creswell, 2003). Without theory-based research, a
discipline chaotically moves in every direction in the absence of an objective (Mitchell & Jolley,
1992). The link of theory to research techniques has been similar to the use of a roadmap when
driving. “Without it, all roads look the same, you can never tell for sure where you are, you
spend a lot of time making incorrect turns and losing your way, and when you get to where you
are going, you may not know it” (Guy, Edgley, Arafat & Allen, 1987:64). This shows that
advances in any field of study are unlikely without an understanding of theory and conceptual
frameworks. Therefore, scientists can communicate their findings without any difficulty with
sets of agreed upon concepts. In order to inform the model constructs for the present study on an
integrated TQM model, previous theoretical frameworks are reviewed.

20
Lennon and Burns (2000:221) asserted that a theory “…allows us to describe and forecast
behaviour, it also dictates which and whose behaviour are deserving of study and which should
be omitted from the study”. Theory development can take place through deductions or induction.
In deduction, the researcher moves from theory to fact or from the general to the specific (Guy et
al., 1987). Both qualitative and quantitative research techniques depend on theory. According to
Creswell (2003:125), deductive theory is often used in quantitative research. “The researcher
advances a theory, gathers information to test it, and reflects on the confirmation or
disconfirmation of the theory by the outcomes.” The theory becomes a framework for the whole
research.

2.2.2 Theories towards the Implementation of TQM

2.2.2.1 Deming's Theory
The theoretical significance of the Deming approach to TQM concerns the creation of an
organizational system that encourages learning and cooperation for facilitating the execution of
process management practices, which, in turn, results in continuous enhancement of procedures,
products, and services as well as employee fulfilment, both of which are necessary to client care,
and eventually, to company success (Anderson, Rungtusanatham, & Schroeder, 1994). Edwards
Deming motivated a thorough strategy to problem solving and promoted the commonly known
Plan, Do, Check, Act (PDCA) cycle. The PDCA cycle is also known as the Deming cycle,
although it was designed by a colleague of Deming, Dr Shewhart. The Deming cycle is a
worldwide enhancement technique, the idea being to regularly enhance, and thereby decrease the
difference between the needs of the customers and the performance of the process. The cycle is
about studying and continuous enhancement, studying what works and what does not in a
systematic way; and the cycle repeats; after one cycle has been completed, another is started.
Deming also placed significance and liability on management at the individual and company
level, knowing management to be liable for 94% of quality problems. Deming therefore insisted
that management’s responsibility is to build excellent systems that promote high quality
management. He proposes that the reason for applying quality management techniques is to help
companies stay in business and that quality improvement has to be led by management. His
fourteen-point plan is a complete philosophy of management that can be applied to small or large

21
organisations in the public, private or service sectors. He believed that the adoption of, and
action on, the fourteen points were indications that management intended to stay in business
(Anderson et al., 1994 in Tam, 2000).

2.2.2.2 Crosby's Theory

Philip B. Crosby is well-known for the theories of “Quality is Free” and “Zero Defects” (Crosby,
1979). He emphasized avoidance rather than after-the-event examination, doing things right the
first time, and zero problems. Crosby asserted that errors are due to two reasons: Lack of facts
and lack of concentration. Education and training can remove the first cause and an individual
commitment to quality (zero defects) and concentration to detail will alleviate the second. He is
another person acknowledged with beginning the TQM movement. He made the point, much
similar to that of Deming, that if you spend money on quality, it is money that is well spent.
Crosby’s concept relies on four absolutes of quality management as follows (Crosby, 1979):
• Quality is defined as to sticking requirements,
• Prevention is the best way to guarantee quality,
• Zero defects (mistakes) is the performance standard for quality, and
• Quality is measured by the cost of non-conformity.

2.2.2.3 Joseph Juran's Theory

Joseph M. Juran is responsible for what has become well-known as the "Quality Trilogy". The
quality trilogy is made up of quality planning, quality control and quality improvement. High
quality management requires quality actions to be planned, improved and managed. The
procedure accomplishes control at one level of quality performance, and then plans are made to
improve the performance on a project-by-project basis. This activity eventually achieves
breakthrough to an improved level, which is again controlled, to prevent any deterioration. He
stresses that if a quality enhancement project is to be successful, then all quality enhancement
actions must be well organized and managed. Juran also believed quality is related to client
satisfaction or dissatisfaction with the product, and emphasised the need for ongoing quality
enhancement through a succession of small enhancement projects carried out throughout the
organisation. He concentrated not just on the end customer, but on other external and internal
customers. Each person along the chain, from product designer to final user, is a supplier and a

22
customer. In addition, the person will be a process, carrying out some transformation or activity
(Juran &.Gryna, 1993).

2.2.2.4 Ishikawa's Theory

Originator of the last theory, Kaoru Ishikawa is often known for his namesake diagram, but he
also developed a theory of how companies should handle their quality improvement tasks.
Ishikawa takes a look at quality from a human viewpoint. He points out that there are seven basic
tools for quality improvement, namely:

• Pareto Analysis - Pareto analysis enables to recognize the big issues in a process;

• Cause and Effect Diagrams - Cause and effect diagrams help to get to the main cause of
problems;
• Stratification - Stratification examines how the information that has been gathered fits
together;
• Check Sheets - Check sheets look at how often a problem happens;
• Histograms - Histograms observe difference;
• Scatter Charts - Scatter charts illustrate connections between different types of factors;
and
• Process Control Charts - A control chart helps to determine what variations to focus
upon (Ishikawa, 1985).
He considered these seven tools should be known commonly, if not by everyone, in an
organisation and used to analyse issues and develop improvements. Used together they form an
effective kit. One of the most well-known of these is the Ishikawa (or fishbone or cause-and-
effect) diagram. Like other tools, it assists groups in quality improvements. It organises the
minimal and major causes leading to one effect (or problem), describes the issue, and recognizes
possible and potential causes by figuring out the possible ones. The diagram consistently
symbolizes and analyses the real causes behind a problem or effect. It organises the major and
minor contributing causes leading to one effect (or problem), defines the problem, and identifies
possible and potential causes by figuring out the possible ones. It also helps groups to be
systematic in the creation of ideas and to confirm that it has stated the direction of causation
properly. The diagrammatic format helps when presenting results to others (Ishikawa, 1985).

23
2.3 THEORY ADOPTED FOR THIS STUDY
This study adopted Deming’s theory of profound knowledge which is a management philosophy
grounded in systems theory. System theory is based on the principle that each organization is
composed of a system of interrelated processes and people which make up system’s components.
The success of all workers within the system is dependent on management’s capability to
orchestrate the delicate balance of each component for optimization of the entire system (Bowen,
2010).The system of profound knowledge is based on system appreciation to understand the
company's processes and systems, variation knowledge to understand the occurrence of variation
and their causes, knowledge theory to understand quality programs and psychology knowledge
to understand human nature. In Deming’s fourteen points, he proposed that among other points,
management commitment, positive corporate culture, employee’s education and training and
proper communication system is paramount in implementation of TQM. He further noted that if
a company focuses on costs, the costs rise while quality deteriorates (Kenya Institute of
Management, 2009). This is consistent with the theory of constraints discussed by Zadry and
Yosuf (2006). The Theory of Constraints (TOC) which is a set of concepts, principles and tools
that can be used to improve management of systems and maximize performance by identifying
the most restrictive limiting factor that constraints the system’s performance and managing it. It
focuses on improving performance rather than reducing costs.

This study is anchored on these two theories in that: it takes all the organizations’ systems to
have a successful implementation of TQM and the organization performance is highly dependent
on its ability to continuously improve on management of its systems.

2.3.1 System Theory

Systems theory is not a new concept. The notion of thinking about things in wholes rather than
parts, have been discussed by early philosophers. Even then, some philosophers have varying
perspectives on ways to view the world. For example, Aristotle pondered about the notion of
wholeness, whereas Descartes supported the notion of breaking things down into smaller parts.
Aristotle and Descartes viewed the living human body in different ways. Aristotle argued that the
whole body is much more than the sum of its parts, while Descartes viewed the body as separate
from the mind (Mazzochi, 2008). According to Aristotle, all things have several parts, and that

24
the whole is different from its parts. Aristotle also talked about connections between bodies or
elements that creates unity and one-ness (Cordon, 2013). This is evident in the following passage
from Metaphysics, written by Aristotle: In the case of all things which have several parts and in
which the totality is not, as it were, a mere heap, but the whole is something beside the parts,
there is a cause; for even in bodies contact is the cause of unity in some cases, and in others
viscosity or some other such quality. Aristotle argued that the whole living body is much more
than the sum of its body parts (Cordon, 2013).

Conversely, according to Descartes, the mind is separate from the body; and that they are two
separate entities. Descartes was the first to introduce reductionism to western thinking
(Mazzochi, 2008). This reductionist approach to science, informed the work of Newton, and the
development of Newtonian epistemology. This worldview states that complexities in the world
can be resolved by analyzing and reducing phenomena to their simplest components (Mazzochi,
2008). Contrary to Aristotle’s view, reductionism is a worldview that asserts that a complex
system is nothing but the sum of its parts, and that it can be described by describing its individual
constituents (Cordon, 2013). In the reductionist approach, individual factors within systems can
be analyzed to get a better understanding of the larger whole. As stated previously, system theory
is based on the principle that each organization is composed of a system of interrelated processes
and people which make up system’s components. Systems theories can be used as a framework
to solve many complicated problems and system issues. Systems theories are also helpful in
understanding organic systems, or systems involving living beings. As organic systems are
always interacting with each other and with their environment, the system is always changing.
Systems theories can help us better understand how changes in nature occur, for example in
evolution. Systems theories can also be used to help us better understand how humans interact
with each other, and with their environment, and the intricacies that exists within their systems.

According to Cordon (2013), a system is defined as “a regularly interacting or interdependent

group of items forming a unified whole”, and as “a group of devices or artificial objects or an
organization forming a network especially for distributing something or serving a common
purpose (Cordon, 2013). Meadows (2009) also defined system as “a set of things–people, cells,
molecules, or whatever – interconnected in such a way that they produce their own pattern of

25
behaviour over time” (Meadows, 2009). These definitions are consistent with other existing
definitions of a system, as they share four common elements: (1) having a group of objects,
molecules, or forces; (2) the relationships and interactions between the groups within their
environment; (3) how these groups make up a larger whole; and (4) the function or purpose of
the elements within the group, that affects the function or purpose of the group as a whole.
Systems may be open or closed, simple or complex. A complex system is one that includes many
other micro-systems, or a network of systems, thus forming a much larger and complex system.
When a group of people are interacting together in an environment, much more complicated
systems develop. These systems include socio-cultural systems, and symbolic systems. In
addition, information systems, as part of this complicated system, also become much more
complicated as new structures emerge, new relationships and social systems are formed,
hierarchy and rules develop, and intangible elements, such as culture and beliefs, are created.
Despite the complexities, all of these elements are interconnected and play a role with within a
larger system (Meadows, 2009).

When looking at systems, it is important to look at the smaller components of the system, within
the context of the larger system. Fundamentally, a system is made up of components that are
interdependent of each other. By looking at the components in isolation, and without looking at
the larger system as a whole, the whole system could collapse. The interdependency of each of
the component means that minute changes to any of the component could result in a domino
effect, thus ultimately changing the system as a whole. This is also true of large and complicated
systems that are composed of smaller systems. Changes to the smaller systems cannot be made
without taking into consideration its potential effect on the larger system.

According to Meadows (2009), a system can lose its ‘system-ness’ when the multiple inter-
relations that held it together no longer function and dissipates (Meadows, 2009). Looking at the
biological structure of a man, if the interconnected nerves stopped functioning, then the transfer
of information and signals to the other organ systems will also not occur. For example, if the
nerves that innervate the respiratory system and the cardiac system stopped sending signals to the
heart and lungs, the heart and lungs will cease to function, ultimately causing death. Again, using
the human body as an example, if the nervous system is only partially functional, then other

26
interrelated systems will find ways to compensate for this partial loss, to make the larger system,
which is the body, to continue its function. To elaborate on this further, if a small part of the
nervous system is affected, such as damage to the 5th thoracic vertebrae of the spinal cord (T5
spinal cord injury), then the flow of information to parts of the body that are innervated by the
nerves in that area will also be affected (Cordon, 2013). As the flow of information from the
nervous system to other systems, such as the musculoskeletal and excretory system is
incomplete, full function of the body will also be incomplete. Depending on the severity of the
injury, individuals with this type of injury usually have complete bladder incontinence, and
paralysis of the lower body and legs (Cordon, 2013). Although the upper body will try to
compensate for the loss of function, other elements, such as a wheelchair, will need to be
introduced to the system for the body to maintain homeostasis. In this context, homeostasis
means preserving the person’s ability to mobilize. Ultimately, the musculoskeletal system in the
lower body will continue to deteriorate, and the wheelchair will take its place within this new
system. The result is a new functioning system, that is adapted, but not quite the same as the old
system. One of the greatest attributes of a system is its ability to change and adapt (Meadows,
2009). In construction organizations, which involve people, processes and structures, there are
multiple types of systems that are involved. Each of the systems is inter-related with one another.
Leaders need to be systems thinkers in order to facilitate sustainable change in their
organizations.

2.3.2 Theory of Constraints

The theory of constraints (TOC) is a concept that emphasizes the role of constraints in limiting
the performance of an organization. TOC drives managers to attack constraints in order to reach
their primary goal. It focuses management’s attention on the factors that impede system
performance. It also provides an action framework that combines the activities of managers
around a few highly visible system elements. Central to the TOC philosophy was that any
organization (or system) has a constraint (or small number of constraints) which dominate the
entire system. The secret to success lies with managing these constraints, and the system as it
interacts with these constraints, to get the best out of the whole system.

The key to success can be found with handling these constraints and the ssystem as it
communicates with these constraints to get the best out of the whole system. TOC has wide
27
applications in production companies, but it can also be used successfully to increase efficiency
in places outside of production, such as the construction industry. TOC can be used together with
other management techniques such as TQM and just-in-time (JIT) to provide an extensive,
connected set of methods that highlight continuous enhancement in all areas of operation
(Goldratt, 1995).

There are numerous key principles underlying TOC. A few of these key principles are worth
stressing because of their importance for implementing the management approach in
organizations (Goldratt, 1995). The principles include the following:

• Processes/organizations are chains. This is critical to TOC. If processes and organizations

function are regarded as chains or flows, the weakest links can be found and reinforced.
The linkages in question can be between the different steps or activities in a process or
between different organizations within a supply chain.
• Local versus system optima. Because of differences and interdependence, the best
possible performance of a system as a whole is not the same as the sum of all the local
optima. (Local optima are calculated measures for functional areas within an
organization).
• Cause and effect. All systems operate in an environment of cause and effect. One event
causes another to happen. This cause-and effect relationship can be very complex,
especially in complex systems. Capturing the essence of cause and effect within the
system and identifying measurements that emulate these relationships are the keys to
optimizing system performance.
• Physical versus policy constraints. Most of the constraints experienced in systems
develop from policies, not physical things. Physical constraints, such as the number of
nurses in a hospital or the number of production machines in a factory, can be logically
recognized and worked with. Policy constraints (e.g. behaviour patterns, attitudes, lack of
information and assumptions) are possibly more destructive than physical constraints, yet
are much more difficult to recognize and deal with. The perception that generating in
large batches is optimal is an example of a policy constraint that can make applying TOC
or related advanced manufacturing approaches challenging.

28
 • Total system impact. All organizations are systems comprised of interdependent
activities, each with its own level and type of variability. In order to optimize
performance, management needs to understand and focus on the total system impact of an
event or decision, not just on its local or immediate effects (Goldratt, 1995).

Revealing the main fundamental principles of TOC, the following five steps that create a
framework for TOC implementation and utilization emerged (Goldratt, 1995):
• Step 1. Identify the system’s constraints. The first phase is to recognize the constraint in
the system that limits throughput or enhancement toward the objective.
• Step 2. Decide how to exploit the constraint(s). Make a decision on a plan for the main
constraint that best facilitates the system’s objective. This needs taking advantage of the
existing capacity at the constraint, which is often wasted by making and selling the
incorrect products, and by inappropriate policies and procedures for scheduling and
controlling the constraint.
• Step 3. Subordinate everything else to the above decisions. Modify or handle the
system’s policies, processes, and/or other resources to support the above decisions.
Management guides its initiatives toward enhancing the efficiency of the constraining
task or activity and any other task or activity that directly impacts the constraining task
or activity.
• Step 4. Elevate the constraint(s). Add potential or otherwise change the position of the
unique resources as the prominent main constraint. In this phase, extra potential is
acquired that will increase (elevate) the overall outcome of the constraining task or
activity. This varies from step 2 in that the added output comes from additional
purchased capacity, such as buying a second machine, tool, or applying new technology.
• Step 5. Return to step 1. Do not let inertia become the new constraint - go back to step
1, but do not allow past decisions made in steps 2 to 4 to become constraints. As a result
of the focusing process, the improvement of the original constraining task or activity may
cause a different task to become a constraining task or activity. Inertia could blind
management to additional steps necessary to improve the system’s output now limited by
a new constraint (Goldratt, 1995).

29
The five concentrating actions allow management to remain targeted on what is vital in an
organization - the system’s constraint(s). Why is the constraint the most essential target?
Obviously, it is the pacesetter for the entire system. No matter how fast the other elements can do
their job, the system cannot produce at a rate quicker than its slowest element.

2.4 EMPIRICAL RESEARCH ON TQM RESEARCH

Globally, a considerable number of studies have been carried out in the area of TQM
implementation. It was apparent that different scientists and researchers implemented different
TQM definitions and frameworks, depending on their own understanding of TQM and research
goals. As a result, there is little consensus on the meaning of TQM and what it comprises. TQM
can be defined as a set of techniques and procedures used to remove or decrease variation from a
production process or service-delivery system in order to enhance performance, quality and
reliability (Steingard & Fitzgibbons, 1993). It incorporates fundamental management techniques,
existing enhancement efforts, and the technical tools (US Department of Defense, 1988) under a
disciplined approach focused on continuous improvement. Kanji and Asher (1996) posited that
TQM is a consistent procedure of enhancement for individuals, groups of individuals, and whole
firms. It has a set of four principles (delight the client, management by fact, people-based
management, and continuous improvement) and eight primary concepts (customer fulfilment,
internal customers are real, all work is process, measurement, teamwork, individuals make
quality, continuous improvement cycle, and prevention). TQM can as well be described as the
application of quality principles for the incorporation of all functions and processes within the
firm (Ross, 1993). On the other hand, TQM can be defined as a management approach for an
organization, focused on quality (ISO 8402, 1994), depending on the involvement of all its
members and seeking long-term success through customer satisfaction and benefits to all
members of the organization and to society.

Flynn, Schroeder and Sakakibara (1994) described TQM as an incorporated way of achieving
and retaining high quality outcome, concentrating on the maintenance and continuous
enhancement of procedures and problem prevention at all stages and in all functions of the firm
to get to know or surpass client objectives. Ho and Fung (1994), however, argued that TQM is a
way of managing to enhance the efficiency, versatility, and competition of a business as a whole.

30
It is also a technique of eliminating waste by including everyone in helping the way things are
done. Vuppalapati, Ahire and Gupta (1995), on the other hand, posit that TQM is an integrative
idea of management for consistently helping the quality of products and processes to accomplish
client satisfaction. Hackman and Wageman (1995) consistently analyzed the three great quality
gurus’ (Deming, Juran, and Ishikawa) suggestions about TQM. According to their analysis
outcome, the following five interventions are the core of TQM, namely precise recognition and
measurement of client wants and needs; creation of supplier partnership; use of efficient teams to
recognize and fix quality problems; use of scientific methods to observe efficiency and recognize
points of high leverage for efficiency improvement; and use of process management heuristics to
improve team efficiency.

Choi and Eboch (1998) considered the TQM paradox using management of process quality,
human resources control, strategic quality planning, and information and analysis as the
constructs of TQM implementation. Black and Porter (1996), on other hand, recognized ten
significant factors of TQM as being clients, people and management, supplier partnership,
teamwork structure for improvement, communication of improvement information, client
satisfaction orientation, external interface management, strategic quality management,
operational quality planning, quality improvement measurement systems, and corporate quality
culture. Research conducted by Powell (1995) also identified the following components as TQM
framework: Executive dedication, adopting philosophy, nearer to customers, nearer to suppliers,
benchmarking, open organization, training, employee empowerment, zero-defects mentality,
flexible manufacturing, process improvement, and measurement. Ho and Fung (1994)
recognized ten TQM elements as commitment, leadership total customer satisfaction, continuous
improvement, total involvement, training and education, ownership, reward and recognition,
error prevention, and cooperation and teamwork. On the other hand, Waldman (1994) recognized
eight key TQM components as top management commitment to place quality as a top priority, a
broad definition of quality as meeting customers’ expectations, TQM values and vision, the
development of a quality culture, involvement and empowerment of all organizational members
in cooperative efforts to achieve quality improvements, an orientation toward managing-by-fact,
the dedication to continuously improve employees’ capabilities and work processes through

31
training and benchmarking, and attempts to get external suppliers and customers involved in
TQM efforts.

Even though a considerable number of studies have been carried out in the field of TQM
implementation, no globally accepted TQM definition or elements currently are available. In
fact, researchers have different thoughts about the TQM concept and elements. Nevertheless,
most agree with the fact that TQM is a philosophy or approach to management focusing on
continuous improvement, customer focus, systematic process management, supplier partnership,
and teamwork. The achievement of such a management philosophy requires a set of practices.

2.5 QUALITY AND FORMS OF QUALITY MANAGEMENT PRACTICES

2.5.1 Quality
Quality definition is dynamic since the clients differ based on their views. Some common
explanations of quality are conformance to requirements, fitness for use, and value for price paid.
Tang et al. (2005) described quality as a state that satisfies the legal, aesthetic and functional
requirement of a product or project by clients. Duncan, Thorpe and Sunmer (1990) refer quality
to standards and the ways and means by which those standards are achieved, maintained and
improved upon. According to the International Organization for Standardization (1994: 19),
quality is “…the totality of characteristics of an entity that bear on its ability to meet stated or
implied needs.”

The term ‘quality’ is often misused, especially when associated with prestigious products.
Although quality is applicable to those products, it does not necessarily refer to their prestigious
attributes, but merely to the fitness of purpose to the customer’s requirements. Harris and
McCaffer (1995: 364) emphasize that quality is meeting client requirements. Looking at quality
from the fitness for purpose perspective, Duncan et al (1990: 15) state that before we can say
whether fitness for purpose has been achieved, we need to know what exactly the purpose is and
how fitness in terms of factors such as performance, duration, reliability, and accuracy, among
others, is to be defined and measured.

32
 2.5.1.1 Quality Past vs. Present
Quality has existed from the earliest decades as behaviour, so we can refer to quality as anything
prepared or manufactured with a degree of excellence or as the worth of a product or service. For
more than 25 years there have been many changes towards improved quality. As shown in Table
(2.1), each civilisation has implemented a part of the concept of quality, but this knowledge
started in the USA, and particularly in the manufacturing sectors.

Table 2.1: Changing View of Quality

Past Present
Quality is the responsibility of blue collar Quality is everyone's responsibility,
workers and direct labour employees including white-collar workers, the indirect
working on the product
labour force and the overhead staff
Quality defects should be hidden from the Defects should be highlighted and brought
customers and management to the surface for corrective action
Quality problems lead to blame, faulty Quality problems lead to cooperative
justifications and excuses solutions
Corrections-to quality problems should be Documentation is essential for "lessons
accompanied with minimum learnt" so that mistakes are not repeated
documentation
Increased quality will increase project costs Improved quality saves money and
increases business
Quality is internally focused Quality is customer focused

Quality will not occur without close People want to produce quality products
supervision of people
Quality occurs during project execution Quality occurs at project initiation and must
be planned for within the project
Source: Kerzner (2003)

33
2.5.2 Quality Control
Quality control entails checking that all the various levels of the procedure of serving the client
have been carried out perfectly and any problems identified have been fixed. According to Zairi
(1991: 37), quality management can be defined as “…operational methods and actions aimed
both at tracking a procedure and removing causes of unsatisfactory performance of relevant
levels of the standard loop (quality spiral) in order to result in economic effectiveness.” Zairi
(1991: 37) further defines quality control as “…the use of methods (mainly statistical) to obtain,
sustain and try to improve on quality requirements of products and services.” On the other hand,
quality control is defined fundamentally as the actions and techniques used to obtain and sustain
the standard of a product, procedure, or service (Oakland and Marosszeky, 2006). It includes a
tracking activity but is also concerned with finding and removing causes quality problems so that
the requirements of the client are continually met. To simplify this definition one can say quality
control refers to a set of activities and methods used to complete quality requirements, by
registering, analysing and writing reports about all information related to quality so that this
information is the basis when making decisions related to quality.

Quality control is significant for the following reasons:

• It helps to follow the path of specification required;

• It helps in the design of the product required;

• It helps to inspect the product during manufacture or production to determine whether it

is in conformance to the customer’s specification; and

• It helps to monitor the use of products and gives feedback if there is a necessity for
improvement (Zairi, 1991).
There are three types of quality control which include the following:
• Irregular control: This is the type of control applied when a customer complains about the
product.

• Routine control: This type of control is regular control taken at different stages of
production/construction.

34
 • Scientific control: This is control through measurement and it is analysed using statistical
sampling theory (Zairi, 1991: 38).

These three types of quality control are all relatively effective but irregular quality control tends
to make a customer lose confidence in the quality of goods produced. Routine and scientific
controls are very good approaches to quality control because the product is still in the production
process when they are carried out. They ensure the quality of the finished product conforms to
the specified requirements in the long run.

Burke (2007: 260) also defines quality control as “…the procedure companies go through to
validate that the item has reached the essential situation as determined by the specifications,
build-method and the contract.” Quality control means monitoring whether specific project
results conform to appropriate quality standard and identifying causes of unsatisfactory
outcomes. The method of testing should be outlined in the project quality plan. This could
involve checklists, inspections, reviews, verification, and validation against standards and
requirements. The project quality plan should also give a definition of deviation and state how to
approach deviation. Harris and McCaffer (1995: 362) note that quality control introduced
inspection to stages in the development of goods and services to ensure that they are carried out
to specified requirements. Inspection is the process of checking or confirming that what is
produced is what is required. They also indicate that quality control is done on a sampling basis
dictated by statistical methods. An example in the construction industry is the making of
concrete and sampling cube tests.

Quality control is the earliest and most basic form of quality management, primarily concerned
with defect detection. Quality control is a useful tool that helps to detect defects early enough
during execution and provides for correction to ensure the product meets the quality
specification. Harris and McCaffer (1995: 364) pointed out that the major objectives of quality
control can be defined as follows:
• To ensure the completed work meets the specification;

• To reduce customers’ or clients’ complaints;

• To improve the reliability of products or work produced;

35
 • To increase customers’ or clients’ confidence; and

• To reduce production costs.

There are five stages of approach to quality control. These are the following:
• Set the quality standard or quality of design required by the customer;

• Plan to achieve the required quality;

• Manufacture right first time;

• Correct any quality deficiencies, i.e. defective work; and

• Provide for long-term quality control and planning (Harris & McCaffer, 1995: 364).

As noted above, authors have different ideas when it comes to quality control, but their major
goal in quality control is to derive a method of detecting and documenting defects and to devise
means of correction before the product is completed. This has particular significance to the
present research in that projects are unique so that there is no possibility of ‘returning’ a
defective final product. Moreover, the products of construction have high value (cost) and are
expected to have long duration so that investment in quality at all steps of the process are
rewarded by reduced maintenance, longer life of the infrastructure and reduced risk of failure,
with the possible negative consequences to public health and safety.

2.5.3 Quality Assurance

Several authors (Duncan et al., 1990; Zairi, 1991; Oliver, 1992; Harris & McCaffer, 1995) have
different approaches to the understanding of quality assurance, which is reflected in the diversity
of their definitions. Duncan et al. (1990: 16) define quality assurance as a structured procedure
put in place to help prevent, manage and control so that the products being produced satisfy
requirements, meet up with time objectives and are cost effective. Harris and McCaffer (1995:
366) state that quality assurance emphasizes problem prevention, unlike quality control which
focuses on problem detection once the item is produced or constructed. Quality assurance
concentrates on the production or construction management methods and procedural approaches
to ensure that quality is built into the production system.

36
Oliver (1992: 18) refers to BS4778: Part 1 to define top quality guarantee as “…all those
organized and methodical activities necessary to provide sufficient assurance that a service or
product will fulfill given requirements for quality”. From this context, quality assurance
documents build confidence to both purchaser or client (customer) and the management. The
client has the assurance that the product will meet his or her quality requirements while
management is assured that the product meets up with their own requirements, that of the client
or customer and those of society. The group of activities aimed at providing confidence to a
purchaser is called external quality assurance, while internal quality assurance comprises the
activities aimed at providing confidence to the management of an organization. Zairi (1991: 40)
argues that quality assurance means that quality control is carried out in a systematic way. Zairi
(1991: 40) further explains that quality management uses what is referred to as a “death
certificate approach”, which means that it rejects inspection as the answer to quality problems
and encourages the implementation of procedures in order to comply with standards. He also
argues that to make sure that products and services are in compliance with set standards, quality
assurance relies on the use of statistical process control techniques.

According to Burke (2007: 255), quality assurance is defined as “...a systematic process of
defining, planning, implementing and reviewing the management processes within a company in
order to provide adequate confidence that the product will be consistently manufactured to the
required condition”. The Project Management Body of Knowledge defines quality assurance as
the planned and systematic activities implemented within the quality system to provide
confidence that the project will satisfy the relevant quality standards (Project Management
Institute, 2000). It is evident that quality assurance emanated from and is dependent on quality
control. Quality assurance builds confidence in the client and the supplier that steps have been
taken to ensure that the specified requirements needed by the client have been followed by the
supplier during production or execution.

37
2.5.4 Quality Management
As defined by Zairi (1991: 41), quality management is “…that aspect of the overall management
functions that determines and implements the quality policy and as such, is the responsibility of
the top management.” Quality management refers to all activities of the overall management
function that determine the quality policy, objectives and responsibilities, and the
implementation of these by means such as quality planning, quality control, quality assurance
and quality improvement within the quality system (ISO 9000 Handbook, 1994: 19). The ISO
Handbook also states that quality management is not separate from general management. When
used effectively, quality management should be an integral part of an organisation’s overall
management approach. Abdul-Rahman (1995: 23) states that the management of quality in
construction is related to time and cost management and vice versa. He also notes that a poorly
managed project may lead to extra cost and time extensions and that a poor time- and cost-
controlled project can affect conformance with a client’s requirements, a crucial aspect of project
quality management. Moreover, poor quality management causes re-work in construction
projects when the quality of a project does not meet the required specification, satisfy the needs
of the customer or the outcome of the project is not fit for the purpose it is needed for. The
effects of re-work on projects are time and cost overruns.

According to Love and Li (2000: 479), quality failures have become an epidemic of the
procurement process in construction and invariably lead to time and cost overruns in projects.
The procurement system adopted for projects can be a major determinant in the achievement of
quality in construction projects. Love and Li (2000: 479) argue that in order to improve the
performance of projects, it is necessary to identify the causes and costs of re-work. Love and Li
(2000: 479) recommended that construction companies and consultant firms (particularly design
consultants) implement quality management practices as well as placing emphasis on
coordinating project documentation during the design development process so that the amount of
rework in projects can be reduced or even eliminated. Re-work can be reduced from the
development stage of a project when the design and planning for the project are in progress.
Love and Li (2000: 489) state that if the construction industry is to improve its performance, all
organizations involved in the project supply chain should implement quality management
practices. It is evident that there is a need for the management of the executing company to be

38
involved with the suppliers’ organisation to plan on how quality is to be achieved in all projects
they execute.

2.5.5 Total Quality Management

Total Quality Management (TQM), as the name implies, is the management of all aspects of
quality (Zairi, 1991: 41). TQM has several definitions based on the organisations’ understanding
of quality. BS 7850 (BSI, 1992) describes TQM as the management philosophy and company
practices that aim to utilize the material and human resources of a company in the most effective
manner to attain the goals of the company. The British Quality Association defines TQM as
“…an all-embracing business management philosophy concentrating on completely satisfying
client requirements with a maximum of efficiency and effectiveness” (Wessel & Burcher, 2004).
On the other hand, the Department of Defence in the USA defines TQM as a philosophy and a
set of guiding rules that represent the foundation of a continually improving organization
(Tingey, 1997).

Zairi (1991: 42) argues that TQM can be defined by several parameters, such as leadership,
attitudes, systems, continuous improvement and customer supply chains. He explains that
leadership is possibly the most essential element in the TQM philosophy, as expressed by quality
management gurus such as Deming, Juran and Crosby. He stated further that a company’s
ambitions and desire to succeed are a reflection of the company’s leadership which is
implemented through a series of actions and ideas. TQM is not about attaining certain standards
of competitiveness or introducing new technologies. It is about changing attitudes and behaviour
towards doing business where parameters are set by the client or agreed with the client.

Of all the definitions, a simple definition of TQM is a meeting of internal and external customer
requirements, and the main difference between quality and TQM is that the quality term usually
focuses on a temporary process. An example is in construction: in order to obtain a proper
strength for concrete according to the specifications, it should be cured by water for three to four
days until the required quality is achieved. This is a temporary process, while TQM is a long-
term process and adopts a strategic dimension to guide each production, financial, marketing and
administrative plan in the direction which supports the strategic dimension.

39
2.6 MODELS FOR TOTAL QUALITY MANAGEMENT IMPLEMENTATION
2.6.1 Formal Quality Award Models
Globally, there are numerous quality prizes and awards, such as the Deming Prize in Japan
(1996), the European Quality Award in Europe (1994), and the Malcolm Baldrige National
Quality Award in the USA (1999). The primary aim of these awards is to increase awareness of
TQM because of its significant input to higher levels of competitiveness. Other aims include to
encourage understanding of the requirements for the achievement of quality excellence and
successful implementation of TQM and also to encourage companies regarding the introduction
of a continuous improvement process. Each award model is based on a supposed model of TQM.
The award models do not focus exclusively on service, product excellence or traditional quality
management methods, but consider a variety of management activities that influence the quality
of the final offerings. They give a useful audit framework against which companies can assess
their TQM implementation practices. Even though each award has its individual distinctive
categories and focus, there are some common areas. Each award model has two parts: One is
TQM implementation (that is, the enablers); the other is the overall business outcomes. All three
award models highlight the significance of leadership, human resources management, employee
participation, employee education and training, process management, strategy and policy,
information, supplier quality management, and customer focus. These models give an
understanding of the practical way of applying TQM, as well as a firm foundation for this
research, and give the researcher a better understanding of the TQM concept.

2.6.2 Other Models for Total Quality Management

Zairi (1991:49) in Ayandibu (2010) describes a model that shows that TQM depends on building
blocks which influence the strength of the organization. The building blocks are summarized
below:
i. The foundation: These are necessary for a successful TQM programme. Improvement,
introduction of change, flexibility and adaptability should be the basics of the
organization. People should be properly nurtured, provided with the right tools, right
working environment, and given flexibility to take part in the continuous
improvement by contributing to their own tasks and solving problems.

40
 ii. The pillars: These are the various ways and processes (quality systems) by which human
inputs are conveyed to output to benefit the end user. These include procedure,
documentation, recording and analysis, workplace design, and technological
innovation, amongst others. The strength of the whole organization is dependent on
the strength of each pillar, thus management should increase the strength of each
pillar and add more where applicable.
iii. The top: Just like the roof of a building, this part should be weather resistant because it
covers the whole organization. This part represents the senior management and they
should have vision for the future of their organization when planning for quality. The
figure below shows the proposed model by Zairi (1991:49) looking at the three levels
of TQM; the top, the pillars and the foundation.

− Quality Planning
− Leadership
− Vision for World Class
Competitiveness

Work Place design: Layout,

Standards like BS 5750

Methods, Ergonomics,
Management Control
User-Supplier Chain

Process Flexibility
Systems like JIT

Safety

− Continuous Improvement
− Added Value Management Activity
− Employee Involvement

Figure 2.1: TQM: Building blocks (Source: Zairi 1991: 49 in Ayandibu 2010:25)

41
There are also some models which have also been suggested recently. One is the TQM-efficient
model (TQMEF) suggested by Subhash and Narag (2007), and targeted at suggesting a TQM
model for Indian organizations, as revealed in Figure 2.2. Critical components of TQM that have
appeared in this model are processes and efficiency. Without sufficient concentration on
efficiency, there will be no product enhancement and all TQM effort will be lost. The TQM
managers must keep an eye on cost and waste reduction, resources planning and usage and,
above all, safety to have a positive effect on society and the growth of the organization. The key
issue is concentrating on client care and making operations effective.

Processes & Customer

Efficiency Focused
Performance
TQM

People Team Building

Management & Business
Partner
Development
Figure 2.2: TQMEF (TQM-Efficiency) Model (Source: Subhash & Narag, 2007)

The TQM framework developed by Adusa-Poku (2014) is also a recently proposed TQM
framework for the Ghanaian construction industry. This framework relies upon a set of primary
principles and aspects that is the base for developing the key performance specifications within
the quality framework. The set of fundamental factors forming the building blocks of the
suggested TQM framework in harmony with the study are Process Management, Continuous
Improvement, Employees’ Satisfaction/Empowerment, Supplier Chain Management, Customer
Focus, Management/Leadership, and Training. The suggested framework by Adusa-Poku (2014)
is dependent on the European Foundation for Quality Management (EFQM). EFQM is a quality
42
model depending on nine (9) outcomes; five (5) enablers and four (4) components which are
outcomes. A simple model illustrating the correlation between enablers and outcomes is shown
in Figure 2.3 and it illustrates that when people acquire proper process management techniques
then performance improves. It was noted that process management is at the centre of all
performances. Excellent process management leads to better performance.

Figure 2.3: Simple Model (Source: Adusa-Poku, 2014)

The proposed TQM framework by Adusa-Poku (2014) is dependent on the above simple model
and EFQM and is as shown in Figure 2.4.

Figure 2.4: TQM Framework (Source: Adusa-Poku, 2014)

The suggested framework by Adusa-Poku (2014) relies upon six criteria. Three of them are the
triangular pillars on which TQM originates its support and are fundamental requirements needed

43
for a successful execution of TQM Framework. The three fundamental requirements are Process
Management, Leadership/Management, and Customer Focus. Process Management is at the top
of the triangle, showing that without process management the framework will be unsuccessful.
Customer Focus and Leadership/Management are the Human Resources who perform their role
effectively in handling all necessary processes. When one of these criteria fails, TQM will also
fail. The connectors to the main pillars for an effective and efficient TQM implementation are
Continuous Improvement, Training, and Supplier Management.

2.7 CONCLUSION
The chapter reviewed the theories and concepts applied in this research. How the interacting
elements within the TQM theory work were explained in this chapter. Quality management
principles, practices and improvement measurement were also discussed. The chapter further
considered the propositions of quality gurus (Deming, Juran, Crosby, Feigenbaum, and
Ishikawa) on concept of TQM. The study adopted Deming’s theory of profound knowledge
which is a management philosophy grounded in systems theory in order to help to understand
and develop a framework for TQM. Systems theory is based on the principle that each
organization is composed of a system of interrelated processes and people which make up the
system’s components. The three quality award models identified also provide a framework for
identifying a range of intangible and tangible processes that influence the firm’s TQM
implementation and the end results. The next chapter addresses the gaps identified in TQM
research.

44
 CHAPTER THREE
3 GAPS IN TOTAL QUALITY MANAGEMENT RESEARCH

3.1 INTRODUCTION
This chapter addresses the gaps identified in TQM research. These gaps have not been evaluated
as all-inclusive constructs in the previous models, and they form the additional new constructs
for the current study’s conceptual framework. The identified gaps and new constructs are Vision
and Plan Statement, Product Selection and Design Management, and Company Quality System
Evaluation. The identified gaps and how to achieve them in TQM implementation are discussed
in this section.

3.2 OBSERVED GAPS IN THE TQM LITERATURE

The extensive review of TQM literature proposes that TQM includes an enormous variety of
topics and perspectives. Different writers have had diverging opinions on the concept of TQM
and its basic elements or constructs since its inception in 1920. Nawaz and Ikram (2013) agreed
that though TQM can be applied in the construction sector, there is a lack of consensus on the
implementation process and lack of understanding of the critical success factors (CSFs) as well.
Hence, it should be noted that for a successful implementation of TQM, it is very important that
construction companies understand the TQM constructs and its critical success factors (CSFs). It
should also be noted that an enormous of research has been carried out in the field of TQM and
its implementation. Sila and Ebrahimpour (2002) reviewed 347 article papers on TQM from
1989 to 2000 and established twenty-five TQM factors which are extensively used by
researchers to measure TQM implementation. Their research discovered eight common cores of
the factors viz: leadership and top management commitment, employee training, customer focus
and satisfaction, teamwork, employee involvement, continuous improvement and innovation,
and quality information and performance.

It should be acknowledged that quality pioneers such as Crosby (1979), Ishikawa (1985),
Deming (1986), Feigenbaum (1991), and Juran (Juran and Gryna, 1993) in advanced countries,
have also given some suggestions relating to the TQM field, which have achieved considerable
recognition throughout the globe. The details of their works give a better appreciation of the
45
TQM principles, philosophy, and practices. It was found in the previous section that quality
gurus have dissimilar opinions about TQM, though some resemblances can be found. It was also
found in the previous section that the three quality award models also provide a framework for
firms’ TQM implementation. These models provide a solid foundation for this study, and give
the researcher a better understanding of the concept of TQM and its constructs. But it was
obvious that there are some gaps in their conceptual framework that have failed to capture the
factors affecting TQM implementation in Ghana and other developing countries as well as the
TQM studies in general.

Using the conceptual frameworks of Saraph, Benson and Schroeder (1989); Flynn, Schroeder
and Sakakibara (1994); Ahire, Golhar and Waller (1996) and those of quality award models i.e
the Deming Prize Quality Award model for Japanese firms, the European Quality Award Model
for European firms, and the Malcolm Baldrige National Quality Award model for American
firms, it becomes clear that most of the research studies relating to TQM implementation were
done in developed countries. On the other hand, studies on TQM implementation conducted in
developing countries and in Ghana in particular indicated factors affecting TQM implementation
but these studies have not adequately provided an overview of the concept compared to those
conducted in developed countries. The next section of the study therefore identifies the gaps in
the TQM implementation theoretical frameworks and attempts to address them. The
consideration of the identified gaps is based on the notion that a TQM model cannot be achieved
without an understanding of TQM concepts and its constructs. This is because the TQM model is
not a simple, single-track factor assessment but a combination of numerous variables. Hence, a
more robust holistically integrated model of TQM implementation is developed in the current
model by considering all the critical variables identified in this study.

3.2.1 Lack of Consensus of TQM Concept and Constructs

Even though several research studies have been carried out in the TQM field, no generally
accepted TQM concepts and constructs currently exist. In fact, researchers have diverse views
about TQM concept and constructs. The concept is still an issue of discussion (Easton & Jarrell,
1998), and still an unclear and confusing concept (Dean & Bowen, 1994). Hence, TQM appears
to mean different things to different people (Hackman & Wageman, 1995). The Asian Institute

46
of Technology (AIT) defines TQM as a philosophy that reinforces the culture to promote
continuous organisational development through orderly, integrated, consistent attempt including
everything and everyone, concentrating mainly on full fulfillment of external and internal
customers, where workers work together in groups with process ownership, guided by a
dedicated and proactive top management (Nukulchai, 2003). On the other hand, the British
Quality Association defines TQM as an all-embracing business management philosophy
focusing on completely satisfying customer requirements with greatest effectiveness and
efficiency (Wessel & Burcher, 2004). Also the Department of Defence in the USA defines TQM
as a philosophy and a set of guiding principles that represent the foundation of a continually
improving organization (Tingey, 1997). All the definitions address TQM as a philosophy but
from diverse points of view. Hence, a common word in TQM definition and concept is a
philosophy.

Zairi (1991: 42) argues that TQM can be defined by several parameters, such as leadership,
attitudes, systems, continuous improvement and customer supply chains. He explains that
leadership is possibly the most essential element in the TQM philosophy. A company’s
ambitions and desire to succeed are a reflection of the company’s leadership which is
implemented through a series of actions and initiatives. These definitions clearly indicate that
TQM has come to mean different things to different people. This study therefore identifies
constructs that are brought together to form a TQM definition and also promote a better
understanding of the TQM concept in the construction industry.

The articles written by Saraph et al. (1989), Flynn et al. (1994), and Ahire et al. (1996) were the
three commonly recommended articles in the field of TQM implementation. Ahire et al. (1996)
strongly suggested that a combination of the three frameworks identified in the three articles
should be considered for future study on TQM. This study follows that recommendation and
attempts to integrate their TQM constructs as much as possible as well as incorporating the
constructs identified in framework developed by Adusa-Poku (2014) in Ghana. Table 3.1
indicates the TQM constructs in this study and the other four frameworks i.e. Saraph et al.
(1989), Flynn et al. (1994), Ahire et al. (1996), and Adusa-Poku (2014) frameworks. The two
constructs “Product quality” and “Supplier performance” in the Ahire et al. framework were not
incorporated in this framework since they represented TQM outcomes. “Role of quality
47
department” in the framework of Saraph et al. was not included in this framework since every
department in any organization would be involved in quality management. “Benchmarking” and
“Internal quality information usage” in the Ahire et al. framework is related to the construct of
“Evaluation” in this study. “Process control” and “Cleanliness and organization” in the Flynn et
al. framework are comparatively the same as the construct of “Construction process management
and improvement” used in this study. Also “Process Management” and “Continuous
Improvement” in the Adusa-Poku framework are comparatively the same as the construct of
“Construction process management and improvement” adopted in this study. “Customer
involvement” in the Flynn et al. framework and “Customer focus” in the Ahire et al. and Adusa-
Poku frameworks are merged in this study. All the motivational related constructs (Employee
education and training, Employee empowerment, Employee involvement, Continuous
improvement, Employee relations, Feedback, Teamwork) identified in the four frameworks i.e.
those of Saraph et al. (1989), Flynn et al. (1994), Ahire et al. (1996), and Adusa-Poku (2014) are
considered as construction employees’ involvement and motivation constructs in this study. This
study added three more constructs, “Company vision and plan statement” “Product selection and
design management” and “Company quality system evaluation”. Hence, this TQM concept
covered a broader scope of TQM in comparison with the four frameworks. TQM could therefore
be defined in this study as follows:
A management philosophy for continuously improving overall performance of organization
based on leadership and top management, company supplier quality management, client focus
and involvement, company quality system evaluation, company vision and plan statement,
product selection and design management, construction process management and improvement,
and construction employee involvement and motivation.

Table 3.1 lists the eight (8) TQM elements in this study and the TQM elements identified in the
other studies framework.

Table 3.1: Framework Comparison

This framework i: Leadership and top management; ii: Company supplier quality
management; iii: Client focus and involvement; iv: Company
quality system Evaluation; v: Company vision and plan statement;

48
 vi: Product selection and design management; vii: Construction
process management and improvement; viii: Construction
employee involvement and motivation.

Saraph, Benson and Schroeder’s framework

i: Role of divisional top management and quality policy; ii: Role of
quality department; iii: Training; iv: Product/service design; v:
Supplier quality management; vi: Process management/operating;
vii: Quality data and reporting; 8: Employee relations.
Flynn, Schroeder and Sakakibara’s framework
i: Quality leadership; ii: Quality improvement rewards; iii: Process
control; iv: Feedback; v: Cleanliness and organization; vi: New
product quality; vii: Interfunctional design process; viii: Selection
for teamwork potential; ix: Teamwork; x: Supplier relationship; xi:
Customer involvement.
Ahire, Golhar and Waller’s framework
i: Top management commitment; ii: Customer focus; iii: Supplier
quality management; iv: Design quality management; v:
Benchmarking; vi: SPC usage; vii: Internal quality information
usage; viii: Employee empowerment; ix: Employee involvement;
x: Employee training; xi: Product quality; xii: Supplier
performance.
Adusa-Poku’s framework i: Process management; ii: Leadership commitment; iii: Customer
focus; iv: Continuous improvement; v: Training and development;
vi: Supplier management
Thus, TQM in this study involves eight elements. Hence, to implement TQM in construction
industry is simply to implement these elements, which happens through a set of practices such as
using certain quality techniques and tools. There are practices that can support the
implementation of each of the eight elements identified in this study. The conceptual definitions
and the practices that support the implementation of the three constructs identified in this study
are presented in the following sub-section.

49
 3.2.1.1 Gap One: Company Vision and Plan Statement (CVPS)
The company vision and plan statement appear in two aspects: The vision statement and plan
statement which are explained as follows:

A vision statement explains how a company wants to be seen in its chosen business. In this
regard, it explains standards, principles, and values. In addition, a vision statement is the
advertisement of the intention to change. As such, it drives the company ahead and acts against
complacency. All workers should be able to understand how they can contribute to the vision. A
statement of values and behaviour is an influential motivating force that can be used to impel a
process of change forward (Kanji & Asher, 1993). The purpose of a vision statement is to
communicate the firm’s values, principles, aspirations and purpose so that workers can make
decisions that are consistent with and supportive of these objectives (Meredith & Shafer, 1999).
An effective vision statement tends to be written using language that can motivate workers to
high stages of performance and further, to foster their commitment. Hence, a company should
have a long-term vision statement. A diversity of workers should be involved in the development
of the vision statement and quality policy, which in return, should be well conveyed to workers
at different stages to stimulate commitment. A quality policy is the general direction and
intention of an organization with regard to quality, as formally articulated by top management
(ISO 8402, 1994). Likewise, a quality policy explains how a company wants to be seen
concerning its quality. Therefore a quality policy is a quality ‘vision statement’. In fact, a vision
statement generally flows down to mission statements that detail short-term firm goals or
departmental aims. In order to understand a vision statement, a company must make plan
statements that support the realization of its vision (Mann, 1992).

A plan statement, on the other hand, is a formalization of what is intended to occur at some time
in the future. A plan cannot guarantee that an event will actually happen; it is a statement of
intention that “will happen” (Slack, Chambers, Harland, Harrison & Johnston, 1995). In a firm,
there are several kinds of plans, including a strategic business performance plan, quality goal
plan, and quality improvement plan. A strategic business performance plan can be divided into
long- and short-term business performance plans that include, for example, market share, profits,
annual sales, exports, and sales growth. A quality goal plan can include conformity rate, defect

50
rate, internal failure costs, external failure costs, performance, reliability, and durability. A
quality improvement plan aims for quality enhancement, which is an action taken throughout the
business to increase the effectiveness and efficiency of activities and processes in order to
provide added benefits to both the business and its customers (ISO 8402, 1994).

3.2.1.2 Gap Two: Product Selection and Design Management (PSDM)

Product design translates customer anticipations for functional requirements into specific
engineering and quality characteristics, which can be called specifications. Juran and Gryna
(1993) assert that sound product design can add to the improvement of product quality to be
better than that of competitors, escalating a company’s competitive advantage in the
marketplace. Product design begins with market research. Hence, it is an essential practice for
design engineers to have some marketing experience and knowledge, making it easier for them
to appreciate customer desires, expectations, and future requirements. In this regard, product
design will be more market-oriented (Feigenbaum, 1991; Juran & Gryna, 1993).

In order to have efficient product design, design engineers are required to have some shop floor
experience such as processing technology, understanding of performance of production
equipment, skill for operating production equipment, and production processes. Such knowledge
can contribute to robust product design. Thus, there will be fewer problems during the process of
production (Feigenbaum, 1991; Juran & Gryna, 1993; Slack et al., 1995). Before production, a
new product design should be carefully reviewed in order to avoid problems during production.
Design review is documented and a comprehensive and systematic examination of a design is
done to assess its capability to fulfil the requirements for quality, identify problems, if any, and
propose the development of solutions (ISO 8402, 1994). Customer requirements and
expectations also should be carefully taken into consideration during the process of product
design. It is imperative that the design team should get detailed information from the field. Field
failure data and customer complaints should be adequately detailed to offer a means of analyzing
the causes so that appropriate corrective action can be taken towards improving product design
(Feigenbaum, 1991; Juran & Gryna, 1993). Different sections in a company should take part in
new product design. Deming (1986) informed that such design teams comprised of people from
sections such as engineering, design, production, and sales can contribute to the improvement of

51
product design and design for the future. According to Juran and Gryna (1991), the involvement
of different sections in product design can guarantee fewer problems during the process of
production as well as after products have been delivered to customers.

Price is still an essential factor influencing the competitive capability of products in the
marketplace (Meredith & Shafer, 1999). Consequently, cost should be paid enough attention
during the process of product design. It is important that reducing production cost does not
sacrifice product performance. Value engineering is a technique for assessing the design of a
product to guarantee that the important functions are provided at minimal overall cost (Juran &
Gryna, 1993).

For traditional products, the product design process is not difficult and can be realized by
experienced design engineers without using any unique techniques. For contemporary products,
certain unique methods and techniques should be used to attain successful product design (Juran
& Gryna, 1993). According to the reseacher’s previous research (Zhang, 2000) experimental
design is an extensively used tool in product design. Its application has greatly reduced the time
and expense needed to develop the new product, significantly improved the performance of the
new product, and led to the success of new product design.

Quality function deployment is also a significant and efficient method in product design (Slack et
al., 1995). It is mainly concerned with the correlation between customer desires and new product
attributes which can support the establishment of a market advantage (Slack et al., 1995). This
technique consists of a series of interconnecting matrixes that translates customer needs into
product and process characteristics (Juran & Gryna, 1993).

3.2.1.3 Gap Three: Company Quality System Evaluation (CQSE)

The concept of evaluation can be defined as a systematic examination of the extent to which an
entity is able of fulfilling specified requirements (ISO 8402, 1994). Juran and Gryna (1993)
affirmed that a formal evaluation of quality provides a starting point by offering an
understanding of the size of the quality issue and the areas demanding attention. Evaluation can
recognize the difference between actual performance and the goal. Evaluating the situation in a

52
company’s quality management practices offers an important base for the company to enhance
its quality management practices. Such evaluation information should be communicated to
workers in order to encourage workers to make things better. Hackman and Wageman (1995)
suggested that evaluation of variability is a change principle. Uncontrolled variance in processes
or outcomes is the main cause of quality problems and must be evaluated and controlled by those
who perform the firm’s frontline work. Only when the root causes of variability have been
identified are workers in a position to take corrective steps to enhance work processes.

It should be noted that a company functions in a turbulent and dynamic environment. In order to
maintain competitive advantages in this environment, the company should continuously evaluate
its various business strategies. A business strategy is a set of objectives, plans, and policies for
the company to compete successfully in its markets (Meredith & Shafer, 1999). In effect, the
business strategy specifies what the company’s competitive advantage will be and how this
advantage will be attained and sustained. Based on such evaluation activities, the company can
amend its business strategy in order to keep it dynamic (Mann, 1992).

A quality audit can be used for quality systems, processes, products, and services. One function
of a quality audit is to evaluate the need for improvement or corrective action (ISO 8402, 1994).
A quality audit is systematic and independent examination to establish whether quality activities
and associated results comply with planned arrangements, and whether these arrangements are
implemented effectively and are appropriate to accomplish goals.

Benchmarking is also an effective tool to use as a continuous process of evaluating a company’s

products, services, and processes against those of its strongest competitors or of companies well-
known as world-class or industry leaders. A benchmark is a point of reference by which
performance is judged or measured. Competitive benchmarking is the continuous process of
measuring products, services, and practices against those of the strongest competitors or leading
companies (DuBrin, 1995). Slack et al. (1995) stated that there are numerous types of
benchmarking such as external, internal, competitive, non-competitive, performance, and
practice. Benchmarking is capable of judging how sound an operation is performing, and can be
seen as one approach to setting reasonable performance standards. It is a useful tool for guiding

53
the establishment of quality improvement goals, evaluating various activities within the
company, and assessing customer requirements (Hackman & Wageman, 1995).

In order to encourage workers to pay attention to quality, quality-related data should be used for
evaluating employee performance. Quality-related indices should be combined with general
employee performance standards. Quality-related data should also be used to evaluate the
performance of employees at different levels and the performance of the whole company, and
should be exhibited on the shop floor in order to enable workers to understand what happens
regarding quality. It should be noted that the main objective of evaluation is improvement, not
criticism. In order to have an effective evaluation, a quality information system is essential as it
is an organized method of collecting, storing, analyzing, and reporting information on quality to
assist decision-makers at all levels (Mann, 1992; Juran & Gryna, 1993).

3.3 CONCLUSION
Different writers have had conflicting opinions on the concept of TQM since its inception. It was
also obvious that there are some gaps in their conceptual framework that have failed to capture
the factors affecting TQM implementation in Ghana and other developing countries, and the
TQM studies in general. The chapter therefore addressed the gaps observed in previous TQM
research frameworks which were not evaluated as all-inclusive constructs in the previous
models. The identified gaps and new constructs are Vision and Plan Statement, Product Selection
and Design Management, and Company Quality System Evaluation. These identified gaps are
discussed in relation to how to achieve them in TQM. In addressing the gaps, several works were
drawn upon but the work of Saraph et al. (1989), Flynn et al. (1994), Ahire et al. (1996), Adusa-
Poku (2014) as well as the Deming theory were specifically used. The three qualities award
models also give an insight into the practical way of implementing TQM, as well as providing a
solid foundation for this study, and a better understanding of the concept of TQM.

54
 CHAPTER FOUR
4 THE CONSTRUCTION INDUSTRY AND TOTAL QUALITY MANAGEMENT

4.1 INTRODUCTION
This chapter presents an overview of the construction industry and TQM implementation in
general. The chapter discusses types of construction undertaken by the construction industry and
the working environment in the construction industry. This is significant because the working
environment in the construction industry is crucial to its success. This chapter will also
elucidates on how TQM is implemented in the construction industry as well as its associated
issues such as benefits and critical success factors (CSFs) affecting its implementation.

4.2 OVERVIEW OF CONSTRUCTION INDUSTRY AND TQM

The construction industry is described as a collection of industries (Kwakye, 1998). This could
be the best description possible because any completed building is composed of materials and
equipment produced from other industries. Hence, understanding the nature of the construction
industry and how it works is an important part of managing and developing the construction
process.

Hinze (2001) posits that the construction of various types of facilities often represents the
culmination of the efforts of several designers. Construction includes all immobile structures,
decks, bridges, sewage treatment plants, and factories. Most of the reshaping of the earth’s
surface can be attributed to the construction industry. Hinze also indicated that the failure rate for
construction firms is relatively high, with business failures in construction accounting for
approximately 12% of all business failures. It has also been estimated that 20% of all
construction-related businesses eventually fail. Failures result from many factors, including over-
extension of resources, subcontractor default, inadequate labour, acts of God, managerial
inexperience, and other economic causes.

The construction industry is one of the industries that are prone to problems, and most of these
problems are serious and need robust and appropriate approach to overcome or at least to reduce
55
their consequences. The construction industry has several issues that need a concerted effort
from all parties involved in this sector to address them. Moreover, the construction industry is
classified as a one of the riskiest sectors because of the fragmentation of supply chains and the
short-term relationships between main contractors on the one side and sub-contractors and
suppliers on the other side. Therefore, through this literature review, the researcher has
endeavoured to answer the question as to whether TQM is the right approach to solve these
problems. To address this, many issues will come under discussion in the following paragraphs.

The construction industry occupies a huge economic segment for any country, and has a
significant effect on the efficiency and development of other industries. It is characterized by the
participation of many parties and the existence of a proper management system becomes
paramount to manage and direct those parties. The construction project usually consists of three
primary participants: the client, the consultant or designer, and the contractor. The process of
most construction projects is similar and starts with the client and the consultant’s office where
the client’s requirements are transferred into drawings and specifications (the design stage). The
project is usually placed out for bidding to contractors: each contractor attempts to provide better
prices than their competitors, and usually (though not always) the lowest price will win the
project. There is a contractual agreement that will be established between the client and the
contractor under the umbrella of the consultant’s office and during this stage, the consultant
works as a supervisor for the project to ensure that all work is implemented according to the
project drawings and specifications. The contractors may be classified as general contractors or
subcontractors (specialty contractors). General contractors are responsible for executing most
major construction projects in all ranges of construction activities. On the other hand,
subcontractors limit their activities to one or more construction specialty activity, such as water-
proofing, electrical work, heating, or excavation.

Hinze (2001) emphasized that the construction industry is characterized by custom-built projects,
whereas standardized methods (mass production) are common in manufacturing. In most
manufacturing or service sectors, all activities are implemented in-house, in other words, in a
closed and shaded area, while a construction project is usually implemented in an open area with
a large number of people, much equipment and materials, unpredictable weather, the attitudes of

56
different people and a large number of activities within a specific and limited time. Therefore,
many researchers are working to find out how to apply a quality system which could offer a
solution for this environment, such as Burati and Kalidindi, (1991) whose research into the
application of TQM in the construction industry has been ubiquitous in the last decade.

Implementing TQM in construction companies means a comprehensive change in every aspect

of the construction process. The process of change is difficult for two main reasons: firstly,
historically the construction industry has been reluctant to implement change. Secondly, a longer
time is needed than in other sectors. Sommerville and Sulaiman (1997) analysed the
implementation of TQM in construction companies and found that most construction managers
tend to lack a long-term strategy and systemic views of construction management, and have a
relatively conservative attitude towards managerial changes.

As mentioned earlier, one of the definitions of quality is meeting requirements, and the question
which arises in the construction industry is that of whose requirements have to be met. The
answer is fourfold: the requirements of designer, supervisor, owner, and constructor. According
to Culp, Smith and Abbott (1993), there is a difference between quality in fact and quality in
perception. This means that a product can be of high quality (quality in fact) and yet it may not
meet a customer's needs and vice versa.

Since the pace of construction industry development is very fast, and in order to reduce redoing
work, the needs for change become increasingly important. According to Love, Li, Irani and Holt
(2000), the industry's problems will remain until all organizations involved in the procurement
of construction begin to take responsibility for initiating changes within their own organizations.
Such change can be initiated through the effective implementation of TQM (Nesan & Holt,
1999).

4.3 TYPES OF CONSTRUCTION

There are several ways the construction projects can be categorized. Hinze (2001) stated that
there are four broad categories, namely housing construction, non-residential building
construction, engineering construction, and industrial construction. Many authors such as

57
Hudson, Sears, and Keoki (2000) have classified the common main types of construction as
building construction, highway construction and industrial construction.

4.3.1 Building Construction

The construction of a building involves several subdivisions and the majority of these
subdivisions consist of small renovations, which are privately owned, with the owner of the
property generally acting as labourer, paymaster, and design team for the entire project. The
important point is that it does not matter what kind of building project it is as they all have some
common elements of design, financial and legal considerations, whether it is a public project or a
privately owned. Construction projects, if not managed or executed properly, may experience
cost overruns, structural collapse and/or litigation. People involved in construction may mitigate
these issues by making detailed plans and maintaining a schedule to ensure a positive outcome
(Hudson, et al., 2000 in Al-Musleh, 2010).

For the construction of large buildings, the client normally hires consultants and ensures his or
her own team of workers and advisors deal with the overall process. This is done to ensure that
the project runs smoothly and the required standards and specifications are met. The role of
advisors can be those of mortgage bankers, project financiers, accountants, lawyers, insurance
brokers, architects, designers or engineers (Hudson, et al., 2000 in Al-Musleh, 2010).

4.3.2 Highway Construction

The development of infrastructure is dependent on our transportation system. Bridges, roads, rail,
airports, harbours, mining, arcades, plazas, high rises and canal system are all included in this
category of heavy or highway construction. Usually these projects are publicly owned, but they
can be privately owned if it is a case of the development part of the government development
scheme. As discussed above in the previous section on building construction, heavy or highway
construction involves design, financial, and legal considerations. However, these projects are not
usually for profit, but undertaken for the public interest. It is essential that heavy/highway
construction teams should deliver work of high quality as this type of construction has a direct
effect on the environment and the infrastructure itself (Hudson, et al., 2000 in Al-Musleh, 2010).

58
4.3.3 Industrial Construction
Industrial construction is a very important component of the construction industry. These
constructions are usually owned by industrial corporations, which are without question for-profit
organisations. These corporations and organisations are in fields such as medicine, petroleum,
chemical, power generation, and manufacturing. Highly specialized skills and expertise are
needed in this type of construction, as it requires special installation of equipment and
machinery. As with the other types of construction explained above, industrial construction has
legal and contingency requirements (Hudson, et al., 2000 in Al-Musleh, 2010).

4.4 SPECIFIC CHARACTERISTICS OF CONSTRUCTION SECTOR

The European Commission Communications (1997) in (Al-Musleh, 2010) stated that the
construction sector is highly regulated, with the following specific characteristics which
differentiate it from other industrial sectors:
• It is a heterogeneous and fragmented sector which depends on a large number of very
different professions, including diversity of technology, customers, projects, and market
sectors.

• Logistical and transport aspects are very important. Construction is one of the most
geographically dispersed sectors with marked regional differences.

• The final product is one of the few non-transportable industrial products, adaptable to a
variety of uses and representing one of the most durable of human artifacts. It forms the
physical infrastructure for living and working, for production and transportation and for
essential services. Half of construction projects relate to renovation.

• Most construction projects are prototypes.

• Investments in machinery, tools and other elements have to be depreciated over a shorter
period than is usual for other industrial sectors.

• The entry-level for new contractors is relatively low because the need for operational
capital is small.

59
 • It is closely linked to the economic cycle, and, being generally conducted outside, is
affected by seasonal climatic variations;

• The sector is very labour intensive, with high mobility of the workforce and growing
skills needed as construction technology becomes more sophisticated. The duration of
contracts is often linked to the length of the site construction phase.

• Accident rates tend to be high.

• Finally, the sector generates an enormous quantity of construction waste and demolition
material (European Commission Communications, 1997 in Al-Musleh, 2010).

The above paragraphs reveal that the nature of the construction industry is unique or has
different characteristics compared to other industries such as the manufacturing industry.

The large number of workers and the diversity of daily activities are critical characteristics of the
construction industry. Therefore, it is essential for adequate management systems to manage
construction company workers and activities and guide their processes

4.5 MANAGEMENT ORGANIZATION SYSTEM IN THE CONSTRUCTION

COMPANIES
Newcombe, Fellows, Langford and Urry (2001) proposed the management organization system
in a construction company as seen in Figure (4.1). The system includes the following:
i Strategic system: The strategic system performs the task of deciding and managing the long-
term direction of the construction companies. The strategic management of the business receives
inputs in the form of market intelligence, assessment of the firm’s current capabilities and
internal and external stakeholder’s attitudes.

ii Social system: The social system’s sole input is people of various types and levels. Through
the processes of motivation, group formation, leadership and commitment, the system seeks to
achieve an output of satisfied, committed and involved personnel.

60
iii Information system: The information system provides the life blood running through the
arteries of the construction company.
iv Management system: The management system is shown in Figure (4.1) as central to the whole
organisational system. It occurs at three levels in the construction organisation, namely strategic,
administrative and operational, each with distinct functions.

Strategic Information

System System

Management

Organisational
Social System
System

Figure 4.1: Construction Organisation System

Source: Newcombe et al. (2001)

The working environment in the construction industry is crucial to its success. Consequently, to
be successful within the construction industry, there are five M's that must be available for any
construction company to survive: Management, Manpower, Money, Machineries, and Materials.
On the other hand, a construction company could lose its reputation if five D's emerge: Delay,
Defective, Dirty, Dispersed, and Discontinuous. The interesting factor is the relations which exist
between the five M's and the five D's because any default in one of the M's could lead to five D's
(Newcombe et al., 2001).

61
4.6 CONCEPT OF TQM

The concept of TQM started in the manufacturing industry, giving the impression that TQM
cannot be applied in any other industry but the manufacturing industry. Regardless of the fact
that construction is different from the manufacturing sector in terms of its uniqueness in
production, studies according to Ahmed (1993), Arditi and Gunaydin (1997), Low and Jasmine
(2004), and Harrington and Voehl (2012) have all confirmed the benefits and applicability of
TQM in the construction industry.

TQM is a system focusing on customer satisfaction through the concept of continuous

improvement. This concept emerged after the 1980s with the purpose of developing and
expanding quality management strategy by adding more aspects related to quality. Most of the
literature indicates that the interest in the TQM concept at the level of production began in the
USA as only theories and but these were implemented in practice in Japan after World War II in
order to improve the quality of industrial production consistently and comprehensively (Alfred &
Mike, 1986).

The TQM concept is one of the modern management concepts which have helped to increase the
competitiveness among organizations. This has resulted from the level of customer awareness
which helps them to select a product or service of high quality and at a reasonable price.
Electronic Business Magazine reported in (1992) that 91 % of 70 companies using TQM had
indicated that their quality had improved when compared with that of their competitors (Talha,
2004). In general, TQM is a way for managers to improve the effectiveness, flexibility, and
competitiveness of a business as a whole.

TQM is considered by many researchers as an important approach in quality and business

performance improvement, and therefore other industries are forcing the construction industry to
adopt TQM. Pheng and Ann Teo (2004) discussed the implementation of TQM in construction
firms, and concluded in their study that TQM has been recognised as a successful management
philosophy in the manufacturing and service industries, so can likewise be embraced in the
construction industry to help raise quality and productivity.

62
The US and European industries begun to understand that poor quality costs companies
significant amounts in sales revenues nationally, and improving the quality of goods and
services would help to improve productivity, lower costs and increase profitability. The US and
Europe have woken up little late, after the competitiveness of Japanese manufacturing in the
early 80's. There is no doubt that most Japanese products are better quality, and lower in cost
than US and European products, which may be a result of Japanese manufacturers understanding
the TQM concepts earlier than others.

4.7 IMPLEMENTATION OF TQM IN THE CONSTRUCTION INDUSTRY

The implementation of TQM in construction has been the subject of many arguments. Several
studies have examined the implementation of TQM in construction and the general observation
is that the construction industry, which includes the Ghanaian construction industry, is slow in
implementing the TQM concept despite its identified potential benefits (Ansah, Aigbavboa &
Thwala, 2015). Haupt and Whiteman (2003) stated that this is one reason why construction
remains behind regarding the implementation of TQM. Agha (2011) opines that the reason for
the late arrival of the construction industry to the TQM approach is that construction
professionals are not familiar with its principles and techniques. However, Boaden and Dale
(1992) note that the relative immaturity of the construction industry in adopting TQM may be an
advantage if companies learn from the mistakes and the best practices of other organizations.
Ramachandran (2010) observed that the construction industry is more complex compared to
other industries and that the use of the TQM approach can be useful in analyzing and remedying
defects as soon as they occur.

Since TQM is a long-term endeavour, some of the firms in the construction industry may not
want to engage in an effort that may not provide any yield till after several years. On the other
hand, contractors often perceive TQM as an extra cost which they may not be willing to incur.
Tutesigensi and Pleim (2008) identified reasons for lack of implementation of quality plans by
small and medium construction firms to include lack of knowledge, the perception that
customers may not need such plans, and lack of resources. Gunning and McCallion (2007)
enumerated obstacles to TQM as seen by contracting firms and they include lack of commitment
from management, lack of communication in organisations, cultural attitudes, and lack of
63
training. Hassin, Tookey and Vidalakis (2007) recommend that training and education are key
factors in the implementation of TQM. Other factors include customer satisfaction, employee
participation and quality policy. Love et al. (2000) made a case for a cultural and behavioural
shift in the mindset of practitioners, academics and professional institutions if the construction
industry is to improve its performance and competitiveness. This view was reiterated by
Mahmood, Yusoff and Mohammed (2008) and Ramachandran (2010), stating that the
implementation of TQM requires a culture change and change in management behaviour.
Organisations would need to devise a strategy to bring all parties on board their quality policies
and plans so there can be better participation and cooperation on all levels. Two aspects
addressed in literature are the need for a radical culture change and the commitment of top
management in the implementation of TQM.

Pheng and Teo (2004) investigated how TQM can be applied more actively in the construction
industry in Japan. They noted that TQM can be embraced in the construction industry to help
raise quality and productivity and concluded that for the successful implementation of TQM,
organizations needed to develop a culture change and a change in status quo. It is easier to teach
new employees about TQM strategy than to try to teach old employees who are set in their ways.

McIntyre and Kirschenman (2000) surveyed the acceptance of TQM in upper Midwestern United
States and they found that the majority of their respondents (72.5%) who employed TQM
practices and benefits articulated include higher customer satisfaction and improved schedule
performance. They recommended an increased effort to be put into the education and training
requirements regarding the implementation and application areas of TQM and a more systematic
approach to the collection and analysis of data concerning the overall TQM process. A survey by
Koh and Low (2010), found that out of eight identified elements of TQM, customer
management, process management and top management leadership were implemented at a
higher level by construction companies.

Agha (2011) studied the suitable applications of TQM in different phases of construction
projects and concluded that though the industry is late in adopting TQM, it is a suitable tool for
improving business quality, increasing customer satisfaction and saving time. Sannni and

64
Windapo (2008) evaluated contractors’ quality control practices on construction sites in Nigeria
and they found that over 80% of contractors did comply with some form of quality control plan
and recommended that contractors be evaluated based on their compliance or otherwise with
quality plans.

TQM philosophy is applicable on various platforms in the construction industry, from the office
management to site operations, whether pre-construction phase or during construction, as a
necessary avenue for securing customer satisfaction and improving productivity while enhancing
organizations’ profitability.

4.8 CRITICAL SUCCESS FACTORS (CSF) OF TQM IMPLEMENTATION IN THE

CONSTRUCTION INDUSTRY
To successfully implement TQM, it is important to identify the factors required for the
implementation process. Saraph et al. (1989) defined CSFs as critical areas of managerial
planning and action that must be practised to achieve effective quality management in a business
unit. These factors may be constructs with latent variables which cannot be measured directly,
but can still be assessed indirectly from their manifestation. In a pioneering study, Saraph et al.
(1989) developed a quality management instrument, identifying eight (8) critical success factors
of TQM, namely the role of divisional top management and quality policy, role of quality
department, training, product/service design, supplier quality management, process
management/operating, quality data and reporting, and employee relations. Their study had
considerable influence on later studies, and subsequent research has resulted in the development
of different frameworks and constructs based on varying perceptions and objectives (Zhang,
2000). Although these frameworks or models have different TQM approaches, they all place the
emphasis on leadership, strategic planning, customer and market focus, human resources focus,
process management, continuous improvement, supplier management and business results in one
way or another (Conca, Llopis & Tari, 2003).

Constructs or elements of CSFs identified in frameworks for TQM point to two categories of
factors: soft and hard dimensions of TQM (Kanji., 1995; Powel., 1995; Dow, Samson & Ford,
1999; Oakland., 2000). Hard components of TQM concentrate on the tools and techniques,

65
systems and the supplementary measurement and control of the work process, ensuring
conformance to performance standards and the reduction of variability. On the other hand, soft
components relate to areas of behavioural concerns such as increasing customer orientation,
employee management, organizational and quality culture. These dimensions are interrelated and
together are very important for the successful implementation of TQM.

A great deal of research has been conducted in the field of TQM and its implementation. The
study by Sila and Ebrahimpour (2002) reviewing 347 articles on TQM from 1989 to 2000
identified 76 studies that employed factor analysis to extract factors for the successful
implementation of TQM. Out of these, they compiled 25 TQM constructs which are widely used
by researchers to measure TQM implementation. Their study revealed seven (7) common cores
of the factors, namely customer focus and satisfaction, employee training, leadership and top
management commitment, teamwork, employee involvement, continuous improvement and
innovation, and quality information. Literature also reveals that different countries have adopted
similar TQM factors as criteria for quality awards under different titles (Metri, 2005). However,
the criteria for all these quality awards are derived from four basic frameworks: the Malcolm
Baldrige National Quality Award (MBNQA), the European Quality Award (EQA) now called
European Foundation for Quality Management, the (EFQM) Excellence Award, and the Deming
Prize.

Jha and Iyer (2006) address the determination of the critical factors affecting quality
performance in construction projects. A preliminary survey was based on Indian construction
projects. The CSFs identified were project manager's competence, top management's support,
monitoring and feedback by project participants, interaction among project participants, and
owners' competence. On other hand, conflict among project participants, a hostile socio-
economic environment, harsh climatic conditions, PM's ignorance and lack of knowledge, faulty
project conceptualization, and aggressive competition during tendering are the factors that most
adversely affected the quality performance of construction projects.

66
Another study of TQM’s CSFs was conducted by Ramirez and Loney (1993) and covered US
companies from the manufacturing and service sectors. The results of this study showed that the
most critical factors are Management commitment, Customer satisfaction, Culture change,
Education, Participation management, Strategic planning, Goal clarity, Error prevention, Top
management steering committee.

Metri (2005) emphasises other critical factors when he analysed the critical success factors of the
fourteen most prominent TQM frameworks. Based on this, he proposed ten CSFs of TQM for the
construction industry. The following ten CSFs have emerged from his analysis: Top management
commitment, Quality culture, Strategic quality management, Design quality management,
Process management, Supplier quality management, Education and training, Empowerment and
involvement, Information and analysis, and Customer satisfaction.

4.9 BENEFITS OF TQM IMPLEMENTATION IN CONSTRUCTION INDUSTRY

The potential benefits offered by TQM techniques are varied and the consensus from various
studies is that it has been successfully applied in other industries and can also be beneficial in the
construction industry. Hassin et al. (2007) studied TQM implementation in the electrical
generation industry and found that its adoption would be of pre-eminent importance to the
industry in Libya. In considering the application of TQM to environmental construction, Kiwus
and Williams (2001) concluded that TQM techniques may reduce the frequency and severity of
schedule overruns, higher customer satisfaction, reduction in construction costs, improved
employee job satisfaction, and improved schedule performance.

Other benefits for implementing TQM include improved relationships with subcontractors,
reduced rework, improved safety, higher productivity, lower employee turnover, speeding up
construction work, improved methods of working, better control over the construction process,
gaining competitive advantage, increase profitability, decreasing waste and rework, better
coordination of activities and being more customer focused (Love et al., 2000; McIntyre &
Kirschenman, 2000; Chini & Valdez, 2003; Love et al., 2004; Hassin et al., 2006; Al-Momani,
2007; Khadour & Darkwa, 2008).

67
4.10 CONCLUSION
The chapter presented an overview of the construction industry and TQM implementation in
general. The chapter also discussed types of construction undertaken by the construction industry
and the working environment in the construction industry. This was significant because the
working environment in the construction industry is crucial to its success. The large number of
workers and the diversity of daily activities are critical characteristics of the construction
industry. Therefore, there is a need for adequate management systems to manage construction
company workers and activities and guide their processes. The chapter therefore clarified how
TQM is implemented in the construction industry and its associated issues such as benefits and
CSFs affecting its implementation.

68
 CHAPTER FIVE
5 TQM IMPLEMENTATION IN DEVELOPED COUNTRIES

5.1 INTRODUCTION
This chapter discusses TQM implementation in developed countries. The chapter specifically
looks into TQM implementation in China. The chapter also discusses the evolution of TQM. An
evaluation of policies and government intervention with regard to TQM implementation is also
presented. Finally, the views of quality pioneers in developed countries and their contribution
towards TQM concept is highlighted in this chapter.

5.2 THE EVOLUTION OF TQM IN THE DEVELOPED COUNTRIES

The beginning of TQM is traced back to the 1920s when statistical methods were employed in
analysing product quality control. Twenty years later, Americans such as Deming, Juran, Crosby
and Feigenbaum improved on this concept by expanding its focus from product quality to quality
in all issues of an organisation. Martinez-Lorente, Dewhurst and Dale (1998) mentioned that the
concept of TQM was developed through the huge evolution in quality which took place in the
US. The concept emerged after the 1980s with the purpose of developing and expanding quality
management strategy by adding more aspects related to quality.

Most of the literature indicates that the interest in the TQM concept at the level of production
which began in USA was as only theories and it was implemented in practice in Japan after
World War II in order to improve the quality of industrial production consistently and
comprehensively. One of the major reasons the Japanese have been so successful in business is
their ability to take a concept or idea from another culture and improve on it in a uniquely
Japanese fashion (Alfred & Mike, 1986). The growing intensity of global competition, especially
from Japan, led the US to follow Japanese strategy. This happened when Hewlett-Packard
criticized US chips manufacturers for poor product quality compared with their Japanese
competitors.

The proponents of quality started with quality control (QC) that deals with the inspection of
works. QC is where products are sampled and inspected for errors or defects. Later on, Quality
69
Assurance (QA) that is about prevention of defects was introduced. The principles for QA are
‘Fit for Purpose’ and ‘Right first Time’. QA is a program covering activities necessary to provide
quality in the work to meet the project requirements (Arditi & Gunaydin, 1997). Currently the
standard for ensuring quality work is by the concept of TQM. It is a philosophy that involves
every organization in the industry in the effort to improve performance. It permeates every
aspect of a company and makes quality a strategic objective (Arditi & Gunaydin, 1997). It
integrates fundamental management techniques, existing improvement efforts, and technical
tools under a disciplined approach focused on continuous improvement. TQM is as a culture that
strives for customer satisfaction through continuous improvement and innovation in all sectors of
the business (Harrington & Voehl, 2012).

According to Zairi (2007), TQM was “coined” in the first instance in the US military. Defenders
of TQM diffused the concept through business consultancies in Western countries, and its
influence spread from manufacturing to the construction and service industries and finally to
education (Ishikawa, 1985). Feigenbaum (1993) in Hossain et al. (2010) identified the following
six phases in the evolution of quality:
• Operator Quality Control
Operation quality control existed before the 19th century. During that period, each worker or
group of workers was responsible for the manufacture of products. In this way, the quality of a
product could be better controlled because the workers focused on their work only.
• Foreman Quality Control
Mass production was the main reason for the emergence of the foreman quality control, which
was developed between 1900 and 1918. Mass production is characterised by the division of
labour, specialisation of skills and standardisation. The classical management approach known as
the scientific management method was initiated by Fredrick Taylor and became very popular.
The foreman grouped the workers performing similar tasks under his/her supervision and he/she
took full responsibility for the quality of all works done by the group.

• Inspection Quality Control

The phase of inspection or testing quality control was developed during the World War I. By
World War II, the manufacturing system had become more complex and large numbers of
workers were reporting to each foreman who could quite easily have lost control of the work. As
70
a result, it was necessary to engage full-time quality inspectors. Quality inspection was adopted
to separate non-conforming parts, and so the term ‘quality’ meant inspection, and usually quality
was inspected during the production process itself.

In this era of mass production, all finished products were examined for defects to ensure quality.
Kanji and Asher (1993) stated that quality management started with simple inspection-based
systems in which the product is inspected and compared with specified requirements to check its
conformity. This means quality measurement at that time focused on the inspection process by
eliminating bad products, and it depended on random inspection.
• Statistical Quality Control
During the period between 1937 and 1960, statistical quality control reached its peak. There
became an increased need for quality inspection in the late mass production era because the
volume and variety of components increased dramatically. Owing to the huge costs involved in
quality inspection, Taylor's scientific management approach became inappropriate. Statistical
quality control concentrated on statistical tools and made the quality inspection department more
efficient, contributing most in sampling inspection rather than in complete inspection.
• Total Quality Control (TQC)
The TQC phase started in the 1960. Feigenbaum was the first expert who used the TQC term. He
considered that "…control must start with the design of the product and end when the product
has been placed in the hands of customers who remain satisfied". There are many elements such
as supplier development relationships, people empowerment and teamwork that are considered
as a part of the TQM concept, but are not included in TQC.
• TQM
TQM evolved in the 1980s and began to have a major impact on management and engineering
approaches to long-term success through customer satisfaction. It is based on the participation of
all members of an organisation in improving processes, products, service and the culture in
which they work. Garvin (1984) outlined the evolution of TQM as the outcome of four major
eras of development. He described the evolutionary process where quality has moved from an
initial stage of inspecting, sorting and correcting standards to an era of developing quality
manuals and controlling process performance. The third stage concerns comprehensive manuals
including areas of an organisation other than production, and the use of standard techniques such

71
as statistical process control (SPC). Martinez-Lorente et al. (1998) summarized the chronology
of TQM development as in Table (5.1).

Table 5.1: Important Events in the Development of TQM

Year Events
1924-1932 Hawthorn’s studies demonstrated the importance of the social and
psychological climate in work.

1924 Shewhart developed statistical process control (SPC).

1926 Bell Telephone began to apply statistical process control methods.

1940s US army pushed the use of sampling methods during World War II.

1950s Many attempts at work improvement undertaken (e.g. job enrichment, work re-
design, participative management, quality of work life, worker involvement).

1950 Deming's first visit to Japan.

1951 Creation of Deming Application Prize in Japan. First edition of Juran's Quality
Control Handbook published.

1954 Juran's first visit to Japan. Maslow's theories about human needs.

1960 Liberalisation of economy in Japan with pressure to improve quality to compete

with foreign companies. McGregor's X and Y theories.

1961 First edition of Feigenbaum's Total Quality Control published.

1962 The idea of quality circles appeared in the first issue of the Japanese Journal
Quality Control for the Foreman.

1970s The pressure of Japanese companies began to be felt in US companies.

1972 QFD was developed at Mitsubishi's Kobe shipyard.

1973 After the 1973 oil crisis the JIT system was adopted by a vast number of
Japanese companies. A small number of US and European companies began to
apply this system in the 1980s.

1974 Quality circles began to be widely introduced in the USA; the first quality circle
programme was launched in Lockheed in 1974 and in the UK Rolls-Royce
introduced the concept in 1979.

72
1979 First edition of Crosby's Quality is Free published. Xerox Corp. started to apply
benchmarking concept to processes. Publication of the BS5750 quality
management series.

1980 An NBC television documentary about the "Japanese miracle" proposed

Deming as a key element in this miracle.

1981 Ouchi's Z theory

1982 First edition of Deming's Quality, Productivity and Competitive Position

published.

1983 Quality on the Line, published by Garvin in Harvard Business Review, analysed
the differences between Japanese and US companies, showing some of the
reasons for the better performance of the Japanese. A paper about Taguchi's
design of experiments was published in Harvard Business Review.

1985 Navel System Command named its Japanese-style management approach

"TQM".

1986 First edition of Deming's Out of the Crisis published. It became a best seller.

1987 First edition of ISO 9000 quality management system series. Publication of the
Malcolm Baldrige National Quality Award.
Source: Martinez-Lorente, Dewhurst and Dale (1998)

73
 Total Quality
• Customer Orientation
• Companywide Improvement
Quality Assurance
• System Oriented
• System Accreditation
• Supplier Assessment
Quality Control
• Product Oriented
• Defect Detection
• Post Production Inspection

1940s & 1950s 1960s & 1970s 1980s & 1990s

Figure 5.1: Evolution of Quality Management (Source: Himanshu, 1999)

74
5.3 QUALITY PIONEERS IN DEVELOPED COUNTRIES AND THEIR
CONTRIBUTIONS TOWARDS TOTAL QUALITY MANAGEMENT
CONCEPT
5.3.1 Contribution by Dr. Joseph Juran
Dr. Joseph Juran originally worked in the quality programme at Western Electric. He
became well known in 1951 after the publication of his book Quality Control Handbook.
Juran (1988) introduced the management dimensions of planning, organizing, and
controlling and focused on the responsibility of management to achieve quality and the
need for setting goals. Juran defined quality as fitness for use rather than simply
conformance to specifications. He was a pioneer in teaching the Japanese on quality
management improvement and believed in top management commitment, support and
involvement in the quality effort. Juran’s ten steps to quality management are the following:
i.) Build awareness of opportunity to improve, ii.) Set goals for improvement; iii.) Organize
to reach goals; iv.) Provide training; v.) Carry out projects to solve problems; vi.) Report
progress; vii.) Give recognition; viii.) Communicate results; ix.) Keep score; and x.)
Maintain momentum by making annual improvement part of the regular systems and
processes of the company.

Juran (1973) considered quality management as three basic processes (Juran Trilogy),
namely quality control, quality improvement, and quality planning. In his view, the
approach to managing for quality consists of three steps: i) The sporadic problem is
detected and acted upon by the process of quality control; ii) the chronic problem requires a
different process, namely quality improvement; and iii) such chronic problems are traceable
to an inadequate quality planning process. Juran defined a universal sequence of activities
for the three quality processes which is listed in Table 5.2.

Juran defined four broad categories of quality costs which can be used to evaluate the firm’s
costs related to quality. Such information is valuable to quality improvement. The four
quality costs are listed as follows:
• Internal failure costs (scrap, rework, failure analysis), associated with defects found
prior to transfer of the product to the customer;

75
 • External failure costs (warranty charges, complaint adjustment, returned material,
allowances), associated with defects found after product has been shipped to the
customer;
• Appraisal costs (incoming, in-process, and final inspection and testing, product
quality audits, maintaining accuracy of testing equipment), incurred in determining
the degree of conformance to quality requirements; and
• Prevention costs (quality planning, new product review, quality audits, supplier
quality evaluation, training), incurred in keeping failure and appraisal costs to a
minimum (Zhang, 2000).

Table 5.2: Universal Processes for Managing Quality

Quality Planning Quality Control Quality Improvement

Establish quality goals Choose control subjects Prove the need
Identify customers Choose units of measure Identify projects

Discover customer needs Set goals Organize project teams

Develop product features Create a sensor Diagnose the causes
Develop process features Measure actual performance Provide remedies, prove
remedies are effective
Establish process controls, Interpret the difference Deal with resistance to
transfer to operations change
Take action on the Control to hold the gains
difference

Source: Zhang (2000)

5.3.2 Contributions by Prof. Edwards Deming

Prof. Edwards Deming is often referred to as the ‘father of quality control. He was a
Statistics Professor at New York University in the 1940s. Deming (1986) stressed the
responsibilities of top management to take the lead in changing processes and systems. He
also insisted management’s responsibility is to build good systems that promote good
76
quality management (Deming, 1986). He emphasized that employees cannot produce
quality goods that exceed what the system was capable of producing. He therefore proposed
14 points as the principles of TQM or steps for implementing TQM to help companies
improve on quality and productivity. These points are listed below:

• Create constancy of purpose for improving products and services.

• Adopt the new philosophy.

• Cease dependence on inspection to achieve quality.

• End the practice of awarding business on price alone; instead, minimize total cost by
working with a single supplier.

• Improve constantly and forever every process for planning, production and service.

• Institute training on the job.

• Adopt and institute leadership.

• Drive out fear.

• Break down barriers between staff areas.

• Eliminate slogans, exhortations and targets for the workforce.

• Eliminate numerical quotas for the workforce and numerical goals for management.

• Remove barriers that rob people of pride of workmanship, and eliminate the annual
rating or merit system.

• Institute a vigorous programme of education and self-improvement for everyone.

• Put everybody in the company to work accomplishing the transformation.

77
5.3.3 Contributions by Philip B. Crosby
Philip B. Crosby is another recognized western guru in the area of TQM. He developed the
phrase, “Do it right the first time” and the notion of zero defects, arguing that no amount of
defects should be considered acceptable. Crosby (1979) coined the term “zero defects” and
stated there is absolutely no reason for having errors or defects in any product or service.
He said the cost of quality is understated in comparison with the cost of improving quality.
Cosby concluded that the cost of quality included all things that are not included in getting
it right the first time.

Crosby (1979) identified a number of important principles and practices for a successful
quality improvement programme which include, for example, management participation,
management responsibility for quality, employee recognition, education, reduction of the
cost of quality (prevention costs, appraisal costs, and failure costs), emphasis on prevention
rather than after-the-event inspection, doing things right the first time, and zero defects.
Crosby claimed that mistakes are caused by two reasons: Lack of knowledge and lack of
attention. Education and training can eliminate the first cause and a personal commitment to
excellence (zero defects) and attention to detail will cure the second (Crosby, 1979).

Crosby offered a 14-step programme that can guide firms in pursuing quality improvement.
These steps are listed as follows:
(i) Management commitment: To make it clear where management stands on quality.
(ii) Quality improvement team: To run the quality improvement programme.
(iii) Quality measurement: To provide a display of current and potential nonconformance
problems in a manner that permits objective evaluation and corrective action.
(iv) Cost of quality: To define the ingredients of the cost of quality, and explain its use as a
management tool.
(v) Quality awareness: To provide a method of raising the personal concern felt by all
personnel in the company toward the conformance of the product or service and the quality
reputation of the company.

78
(vi) Corrective action: To provide a systematic method of resolving forever the problems
that are identical through previous action steps.
(vii) Zero defects planning: To investigate the various activities that must be conducted in
preparation for formally launching the Zero Defects programme.
(viii) Supervisor training: To define the type of training those supervisors need in order to
actively carry out their part of the quality improvement programme.
(ix) Zero defects day: To create an event that will make all employees realize, through a
personal experience, that there has been a change.
(x) Goal setting: To turn pledges and commitment into actions by encouraging individuals
to establish improvement goals for themselves and their groups.
(xi) Error causal removal: To give the individual employee a method of communicating to
management the situation that makes it difficult for the employee to meet the pledge to
improve.
(xii) Recognition: To appreciate those who participate.
(xiii) Quality councils: To bring together the professional quality people for planned
communication on a regular basis.
(ixv) Do it over again: To emphasize that the quality improvement programme never ends
(Crosby, 1979)..

5.3.4 Contributions by Dr. Kaoru Ishikawa

Dr. Kaoru Ishikawa was the first Japanese quality guru to emphasize the importance of the
‘internal customer’, the next person in the production process. He was also one of the first
to stress the importance of total company quality control, rather than just focusing on
products and services. Ishikawa (1985) promoted the use of teams or quality circles and the
seven tools of quality in performance improvement. He developed the Cause-Effect
diagram known as Fishbone Diagram.

Ishikawa defined quality as the development, design, production and service of products
that are most economical, most useful, and always satisfactory to the customer. He
emphasized the importance of training, and use of cause-effect diagrams for problem
solving, and quality circles as a way to achieve continuous improvement (Ishikawa, 1985).

79
Ishikawa (1985) argued that quality management extends beyond the product and
encompasses after-sales service, the quality of management, the quality of individuals and
the firm itself. He claimed that the success of a firm is highly dependent on treating quality
improvement as a never-ending quest. A commitment to continuous improvement can
ensure that people will never stop learning. He advocated employee participation as the key
to the successful implementation of TQM. Quality circles, he believed, are an important
vehicle to achieve this. Like all other gurus he emphasized the importance of education,
stating that quality begins and ends with it.

Ishikawa’s concept of TQM contains the following six fundamental principles:

• Quality first; not short-term profits first;
• Customer orientation; not producer orientation;
• Customer breaking down the barrier of sectionalism;
• Using facts and data to make presentations-utilization of statistical methods;
• Respect for humanity as a management philosophy, full participatory management;
and
• Cross-functional management (Ishikawa, 1985).

5.3.5 Contributions by Armand V. Feigenbaum

Feigenbaum (1991) defined TQM as an effective system for integrating the quality
development, quality maintenance, and quality improvement efforts of the various groups in
a firm so as to enable marketing, engineering, production, and service at the most
economical levels which allow for full customer satisfaction. He claimed that effective
quality management consists of four main stages, described as follows:

• Setting quality standards;

• Appraising conformance to these standards;
• Acting when standards are not met;
• Planning for improvement in these standards (Feigenbaum, 1991).

80
Feigenbaum considered top management commitment, employee participation, supplier
quality management, information system, evaluation, communication, the use of quality
costs, and the use of statistical technology to be essential components of TQM. He argued
that employees should be rewarded for their quality improvement suggestions since quality
is everybody’s job. He stated that effective employee training and education should focus
on the following three main aspects, namely quality attitudes, quality knowledge, and
quality skills (Feigenbaum, 1991).

5.3.6 Summary of Contributions by Quality Pioneers towards TQM Concept

After the approaches to TQM of the five quality gurus from developed countries have been
reviewed, it has become evident that each has his own distinctive approach. Nevertheless,
the principles and practices of TQM proposed by these quality gurus provide the author
with a better understanding of the concept of TQM in developed countries. Although their
approaches to TQM are not totally the same, they do share some common points which are
summarized as follows:
(i) It is management’s responsibility to provide commitment, leadership, empowerment,
encouragement, and the appropriate support to technical and human processes. It is top
management’s responsibility to determine the environment and framework of operations
within a firm. It is imperative that management should foster the participation of the
employees in quality improvement, and develop a quality culture by changing perception
and attitudes toward quality.
(ii) The strategy, policy, and firm-wide evaluation activities are emphasized.
(iii) The importance of employee education and training is emphasized in changing
employees’ beliefs, behaviour, and attitudes, thereby enhancing employees’ abilities in
carrying out their duties.
(iv) Employees should be recognized and rewarded for their quality improvement efforts.
(v) It is important to control the processes and improve quality system and product design.
The emphasis is on prevention of product defects, not inspection after the event.
(vi) Quality is a systematic firm-wide activity from suppliers to customers. All functional
activities, such as marketing, design, engineering, purchasing, manufacturing, inspection,

81
shipping, accounting, installation and service, should be involved in quality improvement
efforts.

According to Zairi (1991), W. E. Deming was the first western scientist invited to give
seminars to Japanese workers and managers on effective quality management. His main
interest was in the application of statistical techniques. He came up with the Deming Cycle:
Plan, Do, Check, Action (PDCA). Deming proposed that the purpose of using quality
management techniques is to help companies stay in business and that quality improvement
has to be led by management. Deming came up with fourteen points for total transformation
that are based on a company-wide quality improvement philosophy (Zairi, 1991).

Deming’s fourteen points highlight the need for a change in approach in terms of quality
management from the aims (point 1), to thoughts (point 2), to action (points 3 and 4),
improving (point 5), incorporating new approach (points 6,7,13 and 14) and eliminating
acts that negatively affect quality (points 8 to 12). Building quality into the product in the
first place (point 3) has significant implications on the operational costs of projects. There
will be a reduction in the cost of maintenance if the product is produced to the required
quality (Zairi, 1991).

Deming’s approach to quality management emphasizes that the purpose of using quality
management techniques is to help companies stay in business. This can be achieved when
there is quality improvement which leads to higher productivity and this lowers costs. The
prices will be lower because of lower production costs and this invariably attracts more
customers to the product with good quality and reduced price – getting value for their
money. When there is an increase in the number of customers for a product at a reasonable
profit for the producer, the producing company stays in business (Zairi, 1991).

Deming’s approach to TQM is quite different from management by objectives because he

argues that the management has two main areas of responsibilities, as mentioned earlier,
and quality management is a duty of everyone in the organization. Meanwhile, the
management by objectives concept requires all managers to participate in strategic planning

82
processes and implement performance systems to help the organization back on track.
Managerial focus of management by objectives is the final result and not the set of activities
that produce the result while Deming’s approach encourages the definition of a detailed
roadmap for implementation (Zhang, 2000).

Deming’s quality philosophy highlights the continuous improvement of the process for
which management is responsible. They also state that there are three types of quality that
management must understand. These include:
i. Quality of design/redesign;

ii. Quality of conformance; and

iii. Quality of performance

According to Zairi (1991: 21), Juran contributed as much to total quality as Deming did.
Juran was invited to speak to Japanese senior managers in 1954 on the importance of
planning, organising and managing quality programmes (Zairi, 1991: 21; Ishikawa, 1885:
19). Juran’s approach to quality control and management was twofold:
• Company’s mission; and

• Senior managers’ role.

Juran proposed that quality has to be controlled at each stage of the process and should be
aimed at controlling the following:
• Sporadic problems/avoidable costs (defects and product failure, scrapped materials,
labour wasted usage for re-work, repair, dealing with customer complaints); and

• Unavoidable costs dealing with chronic problems (prevention and control).

Sporadic problems can be solved easily using quality control techniques such as standard
statistical techniques, charts and diagrams; while unavoidable costs require introduction of
a new culture which is intended to change attitudes and increase companywide knowledge.
Planning, implementing and controlling quality according to the mission of a business
83
determines how lasting a business will be. Juran also argued that managerial processes are
necessary for the structured implementation of a total quality programme: planning, control
and improvement. He argued that the planning process is crucial for improvement to
become a continuous activity (Zairi, 1991: 23). Planning therefore has to be conducted with
a long-term view rather than on a project-by-project basis.

According to Zairi (1991: 23) and Wadsworth, Stephens and Godfrey (1986: 19), Crosby’s
quality drive is prevention and he argued that quality is free. The costs are only related to
the various obstacles for the first time. According to Crosby, the major objectives of
organizations implementing total quality should be Zero Defect. He proposed that
acceptable quality levels should be forbidden because they compromise the commitment
towards the achievement of Zero Defect. He also identified the following two major
problems which are causes of poor quality in the industry:
• Those which are due to employees poor awareness and knowledge; and

• Others which are due to carelessness and lack of attention.

The first problem can be easily identified, measured and solved but the second needs a
long-term management effort in changing culture and attitude. Crosby argued that if a
company is serious about achieving Zero Defect, they have to be serious about prevention.
He also proposed four steps for managers and fourteen steps to improve quality. Crosby’s
fourteen steps differ from Deming’s because they facilitate the introduction of continuous
improvement. Crosby’s approach to total quality is to change the culture and attitudes
within the organizations to implement continuous improvement. This approach is more
management oriented since it does not refer to the control of quality predominantly by the
use of various statistical techniques (Zairi, 1991).

These ‘pioneers’ of quality have their different approaches to quality. Deming’s approach is
more one of transformation, based on a company-wide quality improvement philosophy;
Juran’s approach is the company’s mission in relation to the customer’s specification and
the role of senior managers in providing leadership and resources in awareness of

84
developing systems for quality; and Crosby’s approach produces guidelines for managers
and gives room for corrective actions despite the fact that he preached zero defect.

Zairi (1991: 32) concluded that there is no doubt that TQM ideas and concepts would be
developed in the future to facilitate meeting the requirements of a business market which is
always changing. Despite the constant change in the construction industry, the TQM
approach allows flexibility in the achievement of quality on projects.

5.4 TQM IMPLEMENTATION IN CHINA

This section presents brief review results on TQM implementation in China.

5.4.1 Why China was Chosen for the Study

Over the past years, China has become a focus of interest for Western organizations and
management researchers, along with the awareness of the important role it has played in the
global economy. Accordingly, more and more researchers have been involved in
conducting research projects in relation to China in various fields, such as culture (e.g.
Adler & Ziglio, 1996), organizational studies (e.g. Shenkar & Von Glinow, 1994),
technology transfer (e.g. Tackaberry, 1998), marketing (e.g. Fock & Woo, 1998) and TQM
(e.g. Zhang, 2000). According to Zhang (2000) the existing Chinese literatures related to
the TQM showed that, a number of articles were published dealing with institutional
policies. Such articles mainly discussed what kinds of governmental quality policies should
be drawn up in order to encourage Chinese firms to emphasize quality management,
implement TQM, and improve product quality. Secondly, some articles reported the
specific experiences of implementing TQM in firms. These articles generally discussed the
benefits, importance, and methods of implementing TQM in firms.

5.4.2 Evolution of TQM Implementation in China

According to Zhang (2000), TQM has been introduced in China from 1978 onwards. In
order to encourage firms in implementing TQM, great efforts have been made by the
Chinese government. As a result, an increasing number of firms implemented TQM. In a
survey conducted by Yu et al. (1998), 96% of responding firms reported that they had done

85
so. Therefore, it can be said that the rate of TQM implementation is very high in Chinese
manufacturing firms.

In 1978, the Beijing Internal Combustion Engine Factory and the Qinghe Woollen Mill
started to implement TQM through cooperating with their foreign partners. In the
meantime, a number of experts and scholars began to disseminate the knowledge of TQM
to firms. In September 1978, some firms began to introduce QC circle activities. On 24
August 1979, the excellent QC circles’ reports were presented in Beijing. On 31 August
1979, the China Quality Control Association was established. One of its duties was to
cooperate with relevant governmental agencies to promote TQM in the country. Because of
the significant effects of TQM implementation in the Beijing Internal Combustion Engine
Factory, it was concluded that TQM could be effectively implemented in China. Thus, the
experiences of TQM implementation were then popularized and disseminated to other
firms. In order to encourage firms in implementing TQM, the State Economic Commission
issued the Provisional Regulations on TQM Implementation in Industrial Firms in March
1980. The provisional regulations, which integrated TQM theory with Chinese national
specific conditions, stipulated the significance, role, and implementation method of TQM.
In order to help firms implement TQM, a large number of training courses and seminars
were organized by the quality control associations at various levels. The statistical data in
1985 showed that 38,000 firms implemented TQM while 500,000 QC circles were
established in various firms. Today, there are tens of thousands of firms that have
implemented TQM, which has been implemented not only in state-owned firms but also in
collective and township firms, and not only in industrial firms but also in service firms.

The major TQM implementation practices in Chinese manufacturing firms could be

summarized as using various kinds of quality management tools such as the QC seven tools
and statistical process control in practice, implementing QC circles activities, analyzing and
identifying quality-related costs, emphasizing quality inspection, establishing quality bodies
(e.g. TQM implementation offices) in manufacturing firms, conducting quality audits, and
strengthening process control and improvement, product design, and after sales services.

86
5.5 LESSON LEARNT FROM DEVELOPED COUNTRIES ON TQM
IMPLEMENTATION
The historical development of quality as a management concept demonstrates that its
evolution did not occur abruptly in sudden changes in management philosophy, but
gradually through stable consistent improvement. This reflects a series of management
innovations that were created during the twentieth century. Therefore, the TQM movement
was not formulated as a separate philosophy, but derived from previously established
scientific management concepts.

5.6 CONCLUSION
The chapter discussed TQM implementation in developed countries. Among the issues
discussed in the chapter is the evolution of TQM in developed countries. The historical
development of quality as a management concept demonstrates that its evolution did not
occur abruptly in sudden changes in management philosophy, but gradually through stable
consistent improvement. This reflects a series of management innovations that were created
during the twentieth century. Therefore, the TQM movement was not formulated as a
separate philosophy, but was derived from previously established scientific management
concepts.

The views of quality pioneers in developed countries and their contributions towards the
TQM concept were also highlighted in the chapter. After the approaches to TQM of the five
quality gurus from developed countries have been reviewed, it has become evident that
each has his own distinctive approach. Although their approaches to TQM are not totally
the same, they do share some common points.

87
 CHAPTER SIX
6 TOTAL QUALITY MANAGEMENT IMPLEMENTATION IN DEVELOPING
COUNTRIES
6.1 INTRODUCTION
This chapter discusses TQM implementation in developing countries. The chapter looks
into the construction industry in Nigeria and South Africa and how the TQM concept has
been accepted in these countries. An evaluation of policies and government intervention
with regard to TQM implementation is also presented. In addition, the philosophical basis
for TQM implementation in Nigeria and South Africa is discussed. Also explored in this
chapter are the challenges facing TQM implementation in the Nigerian and South Africa
construction industries.

6.2 THE SITUATION OF NIGERIA CONSTRUCTION INDUSTRY

6.2.1 Construction Industry in Nigeria
The Nigerian construction industry is beset with many problems such as uncompleted
project resulting from delay and cash flow problems leading to time and cost overruns, poor
quality (defects and rework) and low levels of productivity, especially in projects handled
by indigenous construction firms (Saad & Dahiru, 2015). Above all, the Nigerian
construction industry has long been associated with poor quality end products and the
industry has not changed over the years (Mohammed & Rotimi, 2002). Furhermore, the
industry has made little progress in reducing the cost of reworks and defects prevalent in the
sector. As clients have also become more aware of the quality standards required in the
structure commissioned, the public at large believe that the industry fails to offer value for
money, and projects take too long to complete and are too prone to failure (Bamisile, 2004).

Concern over the global decline in the construction quality prompted Kubal (1994) to cite a
report of the USACE Blue Ribbon (1993) that explicit solicitude has been reported about
the decline in construction quality in the past decade in addition to concerns regarding the
decrease in customer satisfaction in the construction industry, despite the programmes
developed to improve the process and the products of construction. However, such concern

88
in Nigeria goes beyond customer satisfaction. In the Nigerian context, the issue of quality
and standard has been the subject of concern in the country’s construction industry in recent
times following the frequent collapse of building structures around the nation (Abiodun &
Afangadem, 2007). A report of the Lagos State Physical and Development Authority (2006)
indicated that out of 61 reported cases of building collapse, 13% were directly attributed to
faulty designs, while 53% of the recorded deaths were attributed to faulty designs.

According to Uduak (2006), the construction firms in Nigeria can be classified as follows:
i. Large firms: These are firms whose annual turnover range from 500 million naira and
above and have over fifty workers on their payroll. They undertake the erecting of a wide
range of buildings from housing estate to offices, hospitals and factories. They also
undertake both national and regional works. These firms occupy less than 3% of the
contracting firms in Nigeria.

ii. Medium sized firms: These represent hose whose annual turnover is between 100 and
500 million naira and have between 20 to 50 people on their payroll. The regional firms
undertake regional work such as real estate schemes and they constitute about 23% of the
total number of contracting firms in Nigeria.

iii. Small sized firms: These are firms whose annual turnover is less than 10 million naira to
100 million naira and have fewer than 20 people on their pay roll. They undertake small
works, industry minor repairs and maintenance works and they constitute almost 74% of
contracting firms in Nigeria.

6.2.2 Implementation of TQM in the Nigerian Construction Industry

There are several project quality management techniques, standards, policies, systems,
procedures, audits, manuals, planning, records, and surveillance systems but they are
seldom applied during project execution in Nigeria. This leads to poor project quality, high
maintenance/running costs, reduced building performance and a high probability of
occupants’ discomfort.

Quality is conformance to standards that represents the products or services basic

characteristics, and is based on customer needs and expectations (Aubrey, 2005). The
89
quality of a product should meet the desired requirements at the lowest cost (Sharma,
2006). Dodo (2008) in Saad and Dahiru (2015) opined that the quest to achieving optimal
quality entails the effective functioning and integration of informed clients, skillful
designers, knowledgeable suppliers/component manufacturers, and competent contractors.
Ozaki (2003) stated that quality management has a three-fold meaning in construction,
namely getting the job done on time, ensuring that the basic characteristics of the final
project fall within the required specifications, and getting the job done within budget. Poor
quality impacts companies in two ways, namely higher cost and lower customer satisfaction
(Pyzdek, 2003). Quality can be defined as the totality of features and characteristics of an
entity that bear on its ability to satisfy stated or implied needs (ISO 9000:2000) where an
entity can be a product, a component, a service, or a process. Quality management has
developed from product-related quality control to company-related TQM, aiming for
continuous process improvement (Seaver, 2003). According to Mohammed (2011), TQM is
an umbrella term for continuous improvement and incorporates quality assurance (QA),
quality control (QC) and quality inspection (QI). Harris and McCaffer (2005) defined
TQM as a management approach that tries to achieve and sustain long-term organization
success by encouraging employee feedback and participation, satisfying customer needs
and expectations, respecting societal values and benefits, and obeying government’s
statutes and regulation. It is a senior management-led process to obtain the
involvement of all employees in the continuous improvement of the performance of all
activities, as part of normal business, to meet the needs and satisfaction of the customer,
whether internal or external.

Previous research indicates that TQM has been in use since the 1980s. In Nigeria, studies
have shown little usage of TQM despite its potential benefits to the industry. Although
TQM is not fully implemented, there is the Standard Organization of Nigeria (SON), the
activities of which are similar to that of the British Standard Institution (BSI) and ISO. SON
has officially adopted ISO 9000 series for quality management in Nigeria. The organization
was established by Enabling Act No. 56 of 1971 which has three amendments - Act No. 20
of 1976, Act No. 32 of 1984 and Act No. 18 of 1990. The activities of SON are similar to

90
those of BSI and ISO. In fact, SON has officially adopted ISO 9000 series for quality
management in Nigeria (Bamisile, 2004).

There are three (3) main standards covered by the ISO 9000 family of standards and they
collectively provide guidance for quality management and the general requirements for
quality assurance. These families are the ISO 9000, 9001, 9002 and 9003. The requirements
of this standard primarily aim at achieving client/customer satisfaction by preventing non-
conformity at all stages, from design through to servicing.

The ISO 9000 series is a framework for improving quality in the construction industry
(Kumaraswamy & Dissanayaka, 2000). The ISO 9000 standard was developed to move
away from the original ‘prescriptiveness’ approach of its predecessors to achieve a more
flexible framework which allows organizations to develop their own policies and
procedures (Harris & McCaffer, 2005). The ISO 9000 family of standards operates on the
assumption that the following factors can have an influence on the quality of a product or
service provided by an organization:
i. Design, purchasing and management
ii. Work patterns, job description, inspection and testing
iii. Reporting relationships, policies and procedures
iv. Record keeping systems, inventory control and training
v. Customers, technologies, resources and planning methods
vi. Production processes, transportation services and communication patterns
vii. Service delivery practices, employee knowledge and skills
According to Nee (1996), the ISO was founded in 1946. Its function then was to establish
worldwide common standards for manufacturing, communication and trade. The best
known series and recognised quality standards are the ISO 9000 series or families of
standards, (Pyzdek, 2003; Harris & McCaffer, 2005). The development of the ISO series
started with the formation of a Technical Committee 176 (TC176) in 1979. The ISO series
was first issued in 1987. The ISO standard embodies comprehensive quality management
concepts and provides guidance for implementing its principles. The concept which
underlies the ISO 9000 standards is that consistently high quality is best achieved by a

91
combination of technical product specification and management systems standards. The
standard allows more flexible framework to be achieved which allows organizations to
develop their own policies and procedures. It is necessary to mention some of the
requirements of ISO series. For instance, ISO 9001 (2008) requires that a quality
organization shall continually improve effectiveness of the quality of management system
through the use of the quality policy, quality objectives, and audits results, analysis of data,
corrective actions and management review. Similarly, ISO 9000 (2005) highlighted that the
aim of continual improvement of QMS is to increase the probability of enhancing the
satisfaction of customers and other interested parties.

Saad and Dahiru (2015) conducted research on multinational and indigenous construction
firms in Nigeria to determine how the concept of project quality management is employed
and implemented in their firms as well as its effects on construction processes and the
quality of the finished product. In their study, a total of one hundred (100) construction
firms were selected out of which 50 are multinationals whereas the other 50 are indigenous.
The results are as follows (See figures 6.1 to 6.3):

Figure 6.1: Comparison of efforts made in achieving quality between multinational &
indigenous construction firms

92
Figure 6.2: Comparison of the factors affecting quality achievement by multinational &
indigenous construction firms

Figure 6.3: Comparison of the level of adherence to project quality management plan by
both firms

93
From the study of Saad and Dahiru (2015) it can be concluded that:

i. There is disparity in terms of project quality delivery between multinational and

indigenous firms;

ii. An uncompromising attitude towards the stipulations of the contract documents in

terms of design and specifications by the multinationals puts them a step ahead, thus
a preference over their indigenous counterparts;

iii. Quality can be achieved when firms can invest in quality management which cost
far less than the cost of correcting defects and rework; and

iv. There are Quality Officers or Quality Managers in multinational construction firms
whereas they are absent in indigenous construction firms.

6.2.3 Lessons Learnt from Nigeria TQM

In construction industry, the aim of all professionals and consultants is to strive to provide
the owner or client with a product or service that will meet the standard that is intended for;
properly built with satisfactory performance and value for money. In furtherance to this
aim, the Standard Organization of Nigeria (SON) was put in place by the Federal
Government to ensure good quality of both materials and finished goods/products that are
produced in the country. In other words, SON is an organization charged with the
responsibility of formulating a national standard for a Quality Management System in
Nigeria. In this regard, many multinational construction firms in Nigeria have gone to a
great extent in introducing a quality management system such as TQM as an integral part of
construction management while most of these firms have chosen to structure their quality
systems in accordance with the standards contained in the ISO 9000 series.

It was realized that most of the construction firms in Nigeria are aware of the benefits of
TQM and the factors enhancing its implementation. However, the levels of adoption of the
TQM principles are very low, especially in indigenous construction firms. The principal
factors which prevent contracting organizations from adopting TQM in Nigeria is the
perception that it takes a long time to yield the desired benefits and that TQM involves

94
unnecessary extra cost. Other factors are lack of adequate training, lack of commitment to
change, and a poor cultural attitude. Furthermore, TQM is a new concept in Nigeria
construction industry and this may explain why though there is some level of familiarity
with it, it is yet to be widely adopted. It was therefore suggested that an extensive
enlightenment campaign on the importance of TQM in construction will be required in
Nigeria to further encourage its use. It was also recommended that government, being the
major client of the construction industry, should come up with an award for excellence for
quality construction to deserving organisations and a penalty for organizations which refuse
to adopt TQM. This will serve as an incentive for conformity to standards in various areas
of the industry as well as a deterrent to defaulters.

6.3 THE SOUTH AFRICAN CONSTRUCTION INDUSTRY

The construction industry was considered as one of the foundations of the country’s plan
for transformation (Bowen, Pearl & Akintoye, 2007). Compared to many other industries,
the construction industry plays a vital role in South Africa's economy and is a significant
contributor to economic growth (Construction Industry Development Board [CIDB], 2012;
Statistics South Africa [Stats SA], 2010). The South African construction industry has been
a key driving force behind the nation's economic growth ever since it was targeted for
reformation at the beginning of the post-apartheid era of the 1990s. In other words,
construction makes a significant contribution to the national economy; it creates
employment (especially for the least skilled members of society); it plays a role in the
development and transfer of technology, it creates many opportunities for enterprises, and it
contributes directly to improving the quality of life of the users of its products. According
to the Department of Public Works (1999), the industry contributes 35% to the total gross
domestic fixed investment and employs approximately 230,000 employees. The South
African Government is the single biggest construction client, making up between 40% and
50% of the entire domestic construction expenditure.

However, the sudden rise in interest rates in the late 1990s and the global economic
meltdown of 2008 had a significant impact on the construction sector of the South African

95
economy and many construction organizations did not survive the effect of these periods
(cidb, 2004; Joubert, Cruywagen & Basson, 2005; South African Reserve Bank, 2009).
Several challenges have also been identified as confronting and influencing the
performance, development and growth of the South African construction industry (Luus,
2003; Van Wyk, 2003, 2004; CIDB, 2004, 2007; Lewis, 2007; Boshoff, 2010; Mbande,
2010; Milford, 2010; Tomlinson, 2010). According to the Department of Public Works
(1999) reports, among other industry challenges are a sharp decline in employment, a steep
decline in gross domestic fixed investment (GDFI), slow delivery of public sector projects
due to poor capacity in public sector institutions and the contractors, low productivity and
poor quality workmanship, and low profit margins for contractors. Compounding these
challenges has been the rapid globalization of the South African economy. In this regard,
large South African contractors in particular are increasingly expanding into offshore
markets to grow revenues and to survive the economic recession affecting the South
African construction industry. This has meant that South African contractors need to be
more competitive to match the level of performance of their counterparts operating in
international markets.

The construction organizations that survived the recessive periods found themselves in a
highly competitive construction business environment with other foreign organizations
(Joubert et al., 2005). Competition was further exacerbated by the fragmented nature of the
construction industry, the modus operandi, as well as its structural features. In order to
confront the resultant challenges of uncertainties and fierce competition posed by the
business environment, it is essential that organizations recognize and establish a strategic
position that will integrate with their business undertakings and decisions (Dikmen &
Birgonul, 2003; Phua, 2006).

In order to improve the construction industry situation in South Africa, a government

department, The National Department of Public Works (NDPW), was tasked to develop a
remedial strategy. The launch of the framework document in 1999 (DPW, 1999) was
among the first decisive interventions aimed at addressing the situation, and set the tone for
government’s intention. Subsequently a Construction Industry Development Board (CIDB)
96
Act was passed in 2000, thus establishing a statutory body aimed at driving an integrated
Construction Industry Development strategy. The CIDB has established the construction
industry development strategy, performance targets and key performance indicators. The
promotion of best practice standards constitutes a critical component of the industry
development strategy. In strong support of the industry development strategy, the Council
for Scientific and Industrial Research’s (CSIR) Building and Construction Technology
Division has developed a model for assessing the overall performance of contractors with a
view to promoting best practice standards.

6.3.1 South African Policies on Construction

After the apartheid era, the South African Government put programmes in place to
empower previously marginalized citizens and improve the economy. This has given rise to
several pieces of legislation and policy documents that can be grouped under the heading of
Black Economic Empowerment. In response to a history dominated by colonialism, racism,
apartheid, sexism and repressive labour policies, the South African government came up
with a programme - the Reconstruction and Development Programme (RDP).

McCutcheon and Taylor Parkins (2003: 16) noted six aspects of employment that have been
disturbing in South Africa. These are summarized below:
i. The unemployment level is high and rising – in both formal and informal sectors;
unemployment grew from 7% in 1980 to between 27% and 37% in 2001.

ii. The formal employment sector was able to absorb 81% annual net additions to the labour
force in the 1960s but this shrank to under 10% in the 1990s.

iii. Unemployment varies geographically – some villages may have over 70% while others
may be about 30%.

iv. The level of unemployment amongst the uneducated between the ages of 16-19 is about
70%.

v. There are deleterious effects on people who are unemployed which lead to violence and
extreme stress.

97
vi. Because the South African economy has become more capital intensive, less than half
the additional employment was created between 1986 and 1990 compared to between 1971
and 1980.

The country’s major policy document, the Reconstruction and Development Programme
(RDP), was drawn up by the African National Congress (ANC) and the key objectives were
meeting basic needs, developing human resources, building the economy, democratising the
state and society, and implementing the RDP (ANC, 1994). The RDP is an integral,
coherent socio-economic policy framework that seeks to mobilize South Africans and the
nation’s resources towards the final eradication of apartheid and the building of a
democratic, non-racial and non-sexist future (ANC, 1994).

Developing human resources as defined by the ANC means involving South Africans in
decision-making processes, implementation, new job opportunities requiring new skills, and
managing and governing society (ANC, 1994). This gave birth to the Expanded Public
Works Programme (EPWP) which is aimed at providing poverty and income relief through
temporary work for the unemployed to carry out socially useful activities. The immediate
goal of the Expanded Public Works Programme Phase 1 was to help alleviate
unemployment by creating at least one (1) million work opportunities, of which at least
40% of beneficiaries would be women, 30% youth and two % people with disabilities.

The EPWP Phase 2 was launched in April 2009 at the University of the Western Cape. The
goal of EPWP phase 2 is to create two (2) million full-time equivalent (FTE) jobs for poor
and unemployed people in South Africa so as to contribute to halving unemployment by
2014 through the delivery of public and community services.

The EPWP has exceeded its aimed target that was set when the programme was launched in
May 1994. According to the South African Government Information (2009), at the end of
April 2008, the EPWP had already created 1 077 801 job opportunities ahead of its
scheduled 31 March 2009 time limit. The EPWP is managed by the Department of Public
Works which manages and provides the accommodation, housing, land and infrastructure
98
needs of national departments. The EPWP optimizes employment, growth and
transformation in the construction and property industries (South African Government
Information, 2009).

Two fundamental strategies underpin the government’s approach to reducing

unemployment: firstly, to increase economic growth so that the number of net new jobs
being created starts to exceed the number of new entrants into the labour market; and
secondly, to improve the education system such that the workforce is able to take up largely
skilled work opportunities which economic growth will generate (South African
Government Information, 2009). Short- to medium-term strategies have been put in place to
contribute towards these strategies, of which the EPWP is one. This has required the
various tiers and departments of Government to embark on labour-intensive projects and
insist that labour-intensive methods be considered on the major projects.

Not only does the Department of Public Works take steps to reduce unemployment in the
country, they also support and empower women-owned construction enterprises through
contractor development programmes such as the Emerging Contractor Development
Programme (ECDP). As part of the ECDP, the Contractor Incubator Programme (CIP) was
inaugurated with the intention of providing support to existing small- to medium-size
construction enterprises to enable them to become sustainable. By August 2008, a total of
136 contractors were registered on the CIP, of whom 62 were women-owned contracting
businesses (South African Government Information, 2009). The figures show that women
contractors owned about 46% of the total local companies in South Africa. This is a
noteworthy performance of the ECDP.

Another state entity that has contributed to the reduction in unemployment, poverty and
inequality through policies and programmes is the Department of Labour. The policies and
programmes of the Department of Labour are aimed at the following:
• Improved economic efficiency and productivity;

• Skills development and employment creation;

99
 • Sound labour relations;

• Eliminating inequality and discrimination in the workplace;

• Alleviating poverty in employment;

• Enhancing occupational health and safety awareness and compliance in the

workplace; and

• Nurturing the culture of acceptance that worker rights are human rights. (South
African Government Information, 2009).

The Employment and Skills Development Services (ESDS) is a unit under the Department
of Labour that helps to achieve one of the focuses of the South African Government – skills
development. There is also a Human Resource Development strategy, which has given rise
to a framework that stands on the following four pillars of strategic intervention:
• Capacity-development initiatives – developing human capital for high performance
and enhanced service delivery;

• Organizational support initiatives – enhancing organizational capacity and support

to maximize the productivity of human capital;

• Governance and institutional support initiatives – ensuring that the Human Resource
Development in the Public Service is effective;

• Economic growth and development initiatives – ensuring that the Human Resource
Development plans, strategies and activities seek to integrate, promote and respond
to the economic growth and development initiatives of government (South African
Government Information, 2009).

The Human Resource Development strategy helps in increasing the sustainable capacity
development for the low-income level citizens of South Africa.

The Sector Education and Training Authority (SETA) is another sub-department under the
Department of Labour that contributes to the raising the level of skills, to bring skills to the

100
employed, and assisting those wanting to be employed in their sector. They help to
implement the National Skills Development Strategy through ensuring that people learn the
skills that are needed by communities and employers. In most developing countries
construction skills are still mainly acquired through an informal apprenticeship system (ILO
2001: 38). The EPWP encourages formal training for all levels of the employees working
under the programme. This assists in ensuring that things are done right the first time to
reduce price and costs during execution (South African Government Information, 2009).

Aggressive affirmative action (AAA) policies were the first step taken to move black
employees quickly into the corporate ranks where they can be trained and developed. Two
basic arguments have been made in favour of such aggressive strategies (McFarlin, Coster
&Mogale-Pretorius, 1999: 65). The first is that the positive effects of economic growth will
take too long to filter down to affect corporate practices. If South Africa fails to redress
apartheid‟s inequalities quickly, it risks a level of racial polarization that could tear the
country apart. The second argument is that aggressive affirmative action is simply good
business since black consumers will dominate the South African economy in the future.
Moreover, affirmative action will help the business environment by lowering
unemployment and crime (McFarlin et al., 1999: 65).

However, despite the large government programmes since 1994 to eradicate the inequalities
of apartheid, almost 50% of the country’s population lives under the poverty line defined as
R1400 (or US$140). In South Africa, the poorest 10% of the population receives only 1.4%
of the total income, while the richest 10% receives 47.3% of the total income (Du Plessis,
Irurah &Scholes 2003: 242). The poverty level in South Africa has induced the government
to embark on development projects with all forms of programmes and policies to ensure
that the there is poverty reduction, growth and development in South Africa. These
programmes have been relatively effective in growing the economy of this developing
country.

101
6.3.2 Poor Contractor Performance and its Causes in South Africa
Many problems are faced by contractors when delivering construction projects. As a result
poor contractor performance, as characterized by poor work quality and low productivity, is
common in the industry. In a survey conducted among members of the South African
Property Owners Association (SAPOA) to investigate the clients’ perception relative to
contractors’ performance, Smallwood (2000) found the predominant problems to be
rework, poor productivity and poor quality. Smallwood concluded that the causes of poor
contractor performance, as perceived by clients, were a lack of concern for the
environment, late information, inadequate or poor planning, poor management of the design
activities, poor management and low skills levels among the workers. Other writers (Allens,
1994; Henry, 1994; Lobelo, 1996) strongly concur with Smallwood’s analysis.
Furthermore, they identified additional problems to be cost over-runs, rework, late
completion, an unacceptably high accident rate, insensitivity to environmental
considerations, poor work practices and adversarial relationships. A common thread
running through all these cases is the failure of many contractors to fully acknowledge the
significance of some key construction issues that seriously affect contractor performance,
such as integration of the design and construction process, as well as the quality
management process.

Ambrose and Claasen (2004) conducted a survey in South Africa and concluded the
following:

• Most of the projects were completed late for various reasons: some due to an increase in
work scope, others due to design deficiencies and others due to poor site management
on the part of the contractor executing the work.

• Many projects were characterized by adversarial sentiment, especially between the

consultant (project manager) and the contractor.

• Many projects experienced cost overruns due mainly to the time elapsed between business
plan formulations and funding approval, with inflation taking its toll on costs. In other
cases, unknown elements such as excess rock encountered caused the cost overrun,
while in a few cases, cost overruns was due to design deficiencies.

102
• The tender evaluations showed that many contractors are disqualified because they lack
the expertise and experience to carry out the work tendered for. In some cases, tender
committees appointed such contractors with alarming results.

6.3.3 Problems pertaining to Poor Quality in the South African Construction

Industry
Verster (1998) argued that, although the principles of producing quality are universal, the
construction industry differs in many aspects from the manufacturing industry. Much
research has been conducted on the management of quality and the implementation of TQM
in construction in South Africa. Some of the biggest problems identified are the following:

6.3.3.1 Commitment by Top Management

Ngowi (2001) state that top management of construction companies, both in South Africa
and Botswana, does not show commitment to quality or TQM and that the focus of
companies is more towards maximizing profit and saving time. Furthermore, management’s
commitment is either not clearly communicated to workers or workers do not have access
to the quality policy and quality goals of the company and therefore do not have the
motivation to deliver a quality product.

6.3.3.2 Training of Skilled and Unskilled Workers

One of the main reasons for poor quality in the South African construction industry is that
the workers who are responsible for the physical execution of projects, and therefore also
for the quality thereof, do not have sufficient training to deliver the expected results (Ngowi
2001). Companies either do not have the time to invest in workers’ training, or they are
unwilling to do so because the workers are contracted on a project-to-project basis and
therefore it is uncertain how long the workers are going to be in employment. According to
Ngowi (2001), the cost to train workers may be considered an unnecessary expense. Also in
South Africa there is very little, if any, correlation between the performance of workers and
their remuneration and this also contributes to the fact that workers do not strive to produce
a product of high quality.

103
 6.3.3.3 The Use of Subcontractors
Joubert et al. (2005) mentioned that main contractors found that the increasing use of
subcontractors leads to fragmentation of the process, with the result that such
subcontractors perform their tasks almost in isolation. A subcontractor’s aim is to complete
his part of the work as quickly as possible, to use as little expensive labour as possible and
to receive remuneration as soon as possible. This is not conducive to good quality.

6.3.3.4 Monitoring and Supervision

Ngowi (2001) found that a shortage of trained and experienced supervising personnel
results in a situation where monitoring and supervision do not take place on a regular basis
and therefore defects are not identified at an early stage. This leads to the continuous
rectification of defective work.

6.3.3.5 Tenders
In their study, Joubert et al. (2005) found that the practice where tenders are awarded solely
on price also has an influence on the quality of projects.

6.3.3.6 Culture
Ngowi (2001) conducted a study on the application of TQM in the construction industry in
Botswana. As a result of Botswana’s proximity to South Africa, many of the following
findings in this study are applicable to the South African construction industry.

Ngowi (2001) found that society in Botswana tends to be fatalistic rather than deterministic.
As a result of this, a culture exists which is largely contrary to a TQM culture. The
following are characteristics of the prevailing Botswanan culture:
• Workers do not take responsibility for their own actions as dictated by TQM
principles and tend to believe all responsibility rests with management.
• Where TQM places emphasis on prevention rather than detection, Botswana
workers leave initiatives towards preventive measures to higher authorities.

104
 • Societies in Botswana (traditionally administrated by chiefs) are more prescriptive
than achievement orientated and therefore workers do not aspire to achieve awards –
one of the aspects strongly emphasized by TQM.
• Also because of their fatalistic nature, workers feel that problems such as quality are
beyond their ability and should be solved by management.
• Most workers on construction projects are recruited on a project basis and therefore
do not spend much time with one company. This hampers another important
element of TQM, namely that of employee empowerment to make decisions at their
level of operation.

6.3.4 Implementation of TQM in South Africa Construction Industry

The international community has gone past quality management as a control system to
TQM as a management system and a culture. Theories and full methodologies for
implementing quality systems are available, both internationally and locally. Previous
research indicates that TQM has been in use since the 1980s. In the South African context,
Smallwood (2000) identified only three contractors that use strategies close to TQM but not
TQM itself. However, “…although the strategies do not constitute a formal TQM strategy
per se in terms of principles, supporting elements and steps, they do incorporate aspects of
the aforementioned” (Smallwood, 2000:5). Smallwood (2000) concluded that such lack of
quality management systems in South African construction companies is among the causes
of poor contractor performance. The findings by Smallwood (2000) are similar to the
findings by Ahmed et al. (2002) that the method and techniques to implement quality
management in the construction industry are still to be developed. The lack of quality
management systems in South African construction companies can be attributed to the
construction industry in South Africa being regulated by the Construction Industry
Development Board (CIDB). Within the CIDB structures, there are Construction Industry
Indicators (CIIs) that measure performance in terms of client satisfaction, health and safety,
quality of work delivered, and quality of tender documents and specifications used, among
others (CIDB, 2010).

105
The failure or successful implementation of TQM largely depends on the deep
understanding of the philosophy behind its origins and how it is integrated in the
organisation as a whole. TQM originated in Japan and it harbours the Japanese philosophy
of a holistic, integrated approach to quality management which in itself is a cultural
foundation. Implementing TQM in organisations which do not share the cultural values
upon which it was founded can result in failure (Ngowi, 2000).

A survey conducted in South Africa by Ambrose and Claasen (2004) indicated the
following:

• No formal TQM system is applied by any of the construction industry stakeholders in

South Africa. Some see it as not practical for this industry, but more suited to
manufacture.

• All the stakeholders identified a need for skills training of personnel involved in projects
to equip them to produce quality work. Furthermore, a need was also identified for
formal TQM education for construction personnel, and for TQM courses to be included
in tertiary qualifications offered.

• The implementation of TQM in the project process from initiation to handover must have
a principal/champion for it to work. This would in most cases be the project manager
(usually the consultant).

• The main factors that make production of quality construction work problematic are:

- Inexperience of the site management team – stakeholder representatives are often ill-
equipped for the on-site management task;

- There are too few barriers to the entry of unqualified participants;

- Personnel are not adequately trained and in many cases not suitably qualified; and

- There is no inherent pride in producing quality work.

• An autocratic approach is still prevalent – this needs to change to a participative and

preventative approach. Anticipation of possible problem situations and appropriate
action will enable the prevention of mistakes. Further measures to prevent problems that
106
 were identified are the employment of suitably qualified staff and the implementation of
a good quality control system.

• Confrontational and adversarial relations mostly develop during the construction stage.
This inevitably occurs when the quality standards have to be enforced, rather than
where the project team takes collective responsibility for quality control. Human ego is
a major stumbling block, and a culture needs to be developed where people welcome
having their work scrutinized.

• The focus of control must move from outside the individual to within so that project team
members can accept accountability for their actions.

• The customers’ needs must be accurately ascertained at project inception, and meeting
these needs must be a primary focus throughout the project process.

• All individuals interviewed agreed that implementing a TQM system would be beneficial
to any project, as well as to the individual organizations involved. These benefits
include enhanced reputation and minimized rework.

• The project process flow can only be effective if at each customer-supplier interface a
customer-supplier relationship exists, and meeting the customers’ needs is of primary
importance (the quality chain effect).

6.3.5 Lessons Learnt from the South African Construction Industry

The theories and full methodologies for implementing quality systems are available, both
internationally and locally. Although most of the construction companies in South Africa do
not have a formal quality management system in place, they believe that a formal quality
management system will ensure better quality on building sites. The failure or successful
implementation of TQM largely depends on the deep understanding of the philosophy
behind its origins and how it is integrated in the organization as a whole. It was realized that
the traditional tender system contributes to poor quality in South Africa construction
industry. Lack of training, lack of site supervision and the culture of labour also have a
negative impact on quality on local building sites in South Africa.

107
6.4 CONCLUSION
This chapter discussed TQM implementation in developing countries. The chapter looks
into the construction industry in Nigeria and South Africa and how the TQM concept has
been accepted in these countries. An evaluation of policies and government intervention
with regard to TQM implementation was also presented. The philosophical basis for TQM
implementation in Nigeria and South Africa were discussed. Also explored in this chapter
are the challenges facing TQM implementation in the Nigerian and South African
construction industries.

108
 CHAPTER SEVEN
7 TOTAL QUALITY MANAGEMENT IMPLEMENTATION IN THE
GHANAIAN CONSTRUCTION INDUSTRY

7.1 INTRODUCTION
This chapter discusses TQM implementation in the Ghanaian construction industry. The
chapter also examines how the TQM concept has been accepted by Ghana as a country. An
evaluation of policies and government intervention with regard to quality management
implementation is also presented. Furthermore, the philosophical basis for TQM
implementation in Ghana is discussed. Also explored in this chapter are the challenges
facing TQM implementation in the Ghanaian construction industry.

7.2 GENERAL OVERVIEW OF GHANA

The modern day Ghana was created from the British Gold Coast Colony, established in
1874, and the UK-administered Trusteeship Territory of Togoland, incorporated in 1956
following a referendum. Anxiety for independence grew strongly after the Second World
War. From the early 1950s, self-government was introduced with elections in 1951, 1954
and 1956. In 1957, Ghana became the first sub-Saharan country in colonial Africa to gain
its independence. Before independence on March 6, 1957 Ghana was called the Gold
Coast. The earliest Europeans to arrive here were the Portuguese in the 15th century. Upon
their arrival, they found so much gold between the River Ankobra and the Volta, they
subsequently named it ‘da Mina’, meaning ‘The Mine’ (Republic of Ghana, 2011).

The Republic of Ghana is a country located in West Africa Coast. It is a country (sovereign
state) located on the Gulf of Guinea and Atlantic Ocean in the sub-Saharan Africa frontier,
only a few degrees north of the Equator; therefore it has a warm climate. The country
covers an area of 238,500 square kilometers (92,085 square meters). Ghana consists of ten
territorial administrative regions bordered by the Ivory Coast to the west, Burkina Faso to
the north, Togo to the east and the Gulf of Guinea and Atlantic Ocean to the south. Map 1
shows the Republic of Ghana with the neighbouring countries and internal sub-division.
109
Map 7.1: Map of Ghana

Source: CIA Factbook, 2012

The regional capitals and land area are as shown in Table 7.1. The national census
conducted in Ghana recorded a population of 24,658,823 in 2010. The population was
6,726,815 in 1960 and increased to 18,912,079 in 2000 as shown in Table 7.2.

Table 7.1: Regions showing Land Area

Regions of Area (km²) Regions of Regions of Area (km²) Regions of

Ghana Ghana Ghana Ghana
Capital capital
Ashanti 24,389 Kumasi Northern 70,384 Tamale
Region Region
Brong- 39,557 Sunyani Upper East 8,842 Bolgatanga
Ahafo Region
Region
Central 9,826 Cape Coast Upper West 18,476 Wa
Region Region
Eastern 19,323 Koforidua Volta 20,570 Ho
Region Region

110
Greater 3,245 Accra Western 23,941 Sekondi-
Accra Region Takoradi
Region
Source: Ghana Statistical Service (2013)

Table 7.2: Population Size

Year Population
1960 6,726,815
1970 8,559,313
1984 12,296,081
2000 18,912,079
2010 24,658,823
Source: Ghana Statistical Service (2013)

Most of the country’s larger cities are in the south, which has a relatively high population
density and low poverty rates whilst the north is sparsely populated but with a very high
incidence of poverty. This population pattern also reflects the underlying economic activity
and hence the distribution of infrastructure in Ghana. Nevertheless, unlike many other
African countries, Ghana’s infrastructure backbones cover the entire national territory and
help to integrate the different regions. Two road corridors linking north and south, a
national power grid, and an ICT backbone interconnect all major population centres (Foster
& Pushak, 2011). According to Vision 2020, Ghana intends to maintain the ‘middle
income’ status by 2020 by UN/World Bank definition (Adusa-Poku, 2014).

Studies show that construction contributes between five % and 10% to the gross domestic
product (GDP) in all countries, employs up to 10% of the working population, and is
responsible for about half of the gross fixed capital formation (Hillebrandt, 2000). In
Ghana, the construction industry contributes 10.5% of the GDP and employs about 6% of
the population (GSS, 2013).

111
Developmental projects in the transport, water and sanitation, energy, health, education and
agricultural sectors in the economy are undertaken by the construction industry. For this
reason, developing a competitive construction industry should be an important objective of
government policy (World Bank, 1984). It is no wonder that the Government of Ghana had
the theme Infrastructural Development for Accelerated Growth and Job Creation in the
2012 National Budget. The aim of the government was to provide infrastructural activities
in various sectors of the economy to stimulate growth and support the private sector in job
generation, thus improving the wealth and health of the general population. Therefore,
construction sector plays an important role in the realization of these objectives.

7.2.1 Political Overview

In Ghana, the Ministry of Water Resources, Works and Housing and the Ministry of Roads
and Transport are responsible for the regulation and certification of onstruction contractors.
Although these two (2) Ministries issue certificates to contractors, this does not ensure the
sustainability and development of the construction industry.

The construction sector has two main classes of product. One is building which is
associated with housing, offices, hospitals, and factories, and the other is civil works
involving the infrastructure for water supply, transport, irrigation, and power generation.
The informal sector deals mainly with individuals in the construction of less complex
buildings with minimal skilled labour required. Such construction work does not undergo
strict adherence to standards. The formal sector is responsible for mostly infrastructural
works with the Government of Ghana, public institutions and multi-national companies as
the clients. The contractors in this formal sector are normally registered and have a large
capital base.

Housing and construction are generally accepted as potential game-changers to Ghana’s

developmental prospects but these sectors appear to be quietly slipping out of the hands of
local entrepreneurs. Generally, Ghana’s housing conditions are poor in nearly all areas
including poor access to improved water sources, poor or non-existent drainage and poor
sanitation reflected in inadequate, cooking, bathing and toilet facilities (GSS, 2013).
112
Although there is a deficit in housing stock, there is no clear-cut policy to support the local
contractors. However, the government initiates policies such as tax rebates and free land as
in the failed STX deal proposed by the government in 2011.

At a West Africa Building and Construction Exhibition and Seminar, a former Finance and
Economic Planning Minister, Dr Kwabena Duffuor, explained that the 120 local contractors
listed in the Ministry’s database do not have the financial muscle required to handle big
projects such as major road construction, bridges and housing. This inadequate financial
position therefore inhibits the growth of the construction industry (Adusa-Poku, 2014).

Ofori (2012) recommended the formation of a central agency to regulate and develop the
construction industry. He concluded that the absence of a regulatory body negatively affects
the standards and practices of this important industry. If the government fails in regulating
the industry and does not institute pragmatic measures to grow this industry, foreigners will
take advantage and win most government lucrative jobs. As in all industries, there are
standards that should govern the industry to ensure specifications and quality is achieved.
Thus, countries such as the UK and Hong Kong adopted ISO 9000 as an international
standard for certification of quality management systems (QMS).

In order to develop GCI strategically, politicians should initiate appropriate policies and
reforms to improve the standards and financial and regulatory status of the construction
business environment.

7.2.2 Economic Overview

Construction is the key sector for every economy. It is known that construction contributes
to the national socio-economic development by providing significant employment
opportunities at non-skilled and skilled levels.

Ghana’s economic history dates back to the days when the economy was dependent on
agriculture, trade and industry. Agriculture was booming during those pre-colonial days
113
when citizens were involved in some form of agriculture (planting or rearing of livestock).
Currently, Ghana is the second leading exporter of cocoa, a crop Tetteh Quarshie brought
from Fernado Po in 1878 (Adusa-Poku, 2014).

Before independence, the industries flourished in terms of mining, pottery, cloth weaving
and carving activities. The industries produced exports in which the Europeans traded.
There was active trading in gold, ornaments and cloth as the Europeans were involved in
the trade activities along the coast. After colonization the economy has grown steadfastly
under different governments depending on their ideological orientations. Various policies
initiated by the International Monitory Fund (IMF), and the World Bank have been
introduced to increase productivity and help boost the economy of the nation.

The major contributors to the GDP in Ghana in 2012 were services, followed by industry
and then agriculture as shown in Table 2.3. The GDP in Ghana was worth 40.71 billion US
dollars in 2012. The GDP value of Ghana represents 0.07% of the world economy.
Construction contributed the most to the GDP in the industry sector, as shown in table 7.4.

Table 7.3: Gross Domestic Product (GDP) at Current Market Prices by Economic
Activities (Gh¢ Million)

Economic 2008 2009 2010 2011 2012

Activity
Services 13,934.6 17,543.5 22,183.6 27,422.7 33,962.5
Industry 5,854.5 6,775.7 8,294.5 14,274.4 18,580.5
Agriculture 8,875.0 11,342.8 12,909.6 14,154.8 15,399.1
Source: Ghana Statistical Service (2013)

114
Table 7.4: Gross Domestic Product (GDP) at Current Market Prices by Economic
Activities

Economic Activity 2008 2009 2010 2011 2012

INDUSTRY 5,854.5 6,775.7 8,294.5 14,274.4 18,580.5
Mining and Quarrying 693.2 740.0 1012.7 4,689.9 5,956.1
Crude Oil 0.0 0.0 0 177.5 3,746.3 4,645.4
Manufacturing 2,276.7 2,478.4 2,941.5 3,842.5 4,680.1
Electricity 155.2 166.9 266.0 279.7 329.3
Water and Sewerage 228.9 246.4 368.3 467.4 505.3
Construction 2,500.5 3,144.0 3,706.0 4,994.9 7,109.6
Source: Ghana Statistical Service (2013)

According to the World Bank country classifications, the country moved from low-income
to lower middle-income status on July 1, 2011. To improve on this economic status, the
government intends to improve its infrastructure development especially in the rural and
northern sectors of the country where poverty rates are high when compared to the southern
sector. Foster and Pushak (2011) encouraged the country to maintain its middle income
status since raising the country’s infrastructure endowment to that of the region’s middle-
income countries could boost the annual growth by more than 2.7 percentage points.

Thus, the Minister of Finance stated that the establishment of the Ghana Infrastructure Fund
(GIF) as a quasi-fiscal body to deal with the huge infrastructure deficit and to focus on
strategic infrastructure will lead to job creation and growth of the economy (Terkper, 2013).
This confirms the assertion that construction is seen as a vehicle for dispersing economic
activity and raising income levels of the citizenry (Ofori, 2012).

One major challenge facing the GCI is access to finance to enable local contractors to
compete with foreign partners. In this regard, the Chairman for the Ghana branch of the
Chartered Institute of Building (CIOB), Rockson Dogbegah, is advocating setting up a bank
solely for the construction industry. He said this has become necessary owing to the high
financial cost involved in any construction project.
115
7.2.3 Sociological Overview
Education and employment are the factors that affect the social state of the country. People
with no formal education and those with primary and middle school (basic) education
dominate the working population of the country. Thus, nearly 70% of the workforce has no
formal or only basic education (GSS, 2013). The GSS report (2013) also stated that about
67% of the population were self-employed without employees, implying small-scale
enterprises dominate the economy with few avenues for employing other people.

The construction industry employs about six % of the population (GSS, 2013). This
industry has the capability of generating employment especially for the labour-intensive
activities. Apart from the labour directly linked to the project, other employment is
generated for food sellers, suppliers of goods such as electrical, plumbing, and the financial
sector (loans).

7.2.4 Technological Overview

Technology is the application of an existing body of knowledge to the production of goods
and services (Henderson & Ruiker, 2010). It embraces tools, equipment, techniques,
materials, systems, the products and their use (Adusa-Poku, 2014). There is a general
agreement that technology is critical to the development of a nation.

Technological advancement is introduced in the construction industry to increase

productivity and cost reduction leading to more profit. Studies have proved that technology
growth in the manufacturing industry has greatly improved profit margins. However, the CI
is slow in adapting new technologies to improve profit. After studying strategies in
improving technology implementation in the construction industry, Henderson and Ruiker
(2010) concluded that the degree of technology implementation success depends on the
degree to which changes are planned, managed and evaluated. Therefore, technology
implementation within construction organizations is not so much a technological problem
as it is a human behavioural one.

116
The development of the Ghanaian construction industry (GCI) is critical for improving the
economic state of the country since it is responsible for the provision of the infrastructural
development in the country. Moreover, the construction industry generates a great deal of
employment in both the private and public sectors. The main contributors to the Ghanaian
industry sector were the construction industry, which recorded a growth of 19.2%, mining
and quarrying contributed 5.2% and water and sewerage activities, which contributed 0.6%
(GSS, 2012). It is therefore imperative that the construction industry is developed and
improved to produce quality works in order to enhance the growth of the national economy.

Ofori (2012) stated that although the GCI has much strength, there are a myriad of
problems that have persisted for a long time. He suggested that an agency should be set up
to survey and develops comprehensive policy guidelines for the industry. In an article
titled‘ Construction and the Internet’ in The Economist (2000), it was noted that up to 30%
of construction costs are due to inefficiencies, mistakes, delays, and poor communication.
It is obvious that the GCI is affected by factors such as globalization, information
communication technology (ICT), increasing knowledge development and the ever-
increasing innovative products delivered daily to the market.

A typical construction project consists of the interaction between a client, consultant and a
contractor. In construction project management, where the objective is to meet clients’
needs and expectations, one must have knowledge of the management of time, quality,
human resources, and risk to ensure a successful completion. Lad and Beck (2009) noted
that construction quality management is the key to a successful project. They explained that
the selection of a Construction Quality Manager (CQM) early in a project gives the overall
project a better chance of delivery on time, to budget, with zero defects as well as high
availability and reliability. However, CQM effort can be a cultural change and a challenge
for contractors. It requires attention to detail, tracking of issues and accountability (Lad &
Beck, 2009). One of the means of achieving this feat is by applying the principles of
TQMTQM in construction.

117
7.3 IMPLEMENTATION OF TOTAL QUALITY MANAGEMENT IN THE
GHANAIAN CONSTRUCTION INDUSTRY
There are several project quality management techniques, standards, policies, systems,
procedures, audits, manuals, planning, records, and surveillance systems but they are
seldom applied during project execution in Ghana. This leads to low quality works, high
construction costs and dissatisfied clients (Ansah et al., 2015). With the country attaining a
middle-income status in addition to the oil production which started in the year 2010, there
is an increased demand for infrastructural work with its accompanying standard quality
requirements. Also, considering the influx of foreign construction companies in Ghana and
the increasing competitiveness in the construction sector, it is important that Ghanaian
contractors deploy appropriate quality management strategies such as TQM to reduce cost
and increase value and quality.

Previous researches indicate that TQM has been in use since the 1980s but in Ghana,
studies have shown little usage of TQM despite its potential benefits to the industry.
Although TQM is not fully implemented in Ghana, there are the National Building
Regulations (NBR) (L.I. 1630) which was enacted in 1996 to ensure building work satisfies
minimum constructional standards, energy conservation requirements and also the health
and safety of people occupying the building. The requirements of this standard primarily
aim at achieving client/customer satisfaction by preventing non-conformity at all stages,
from design through to servicing.

7.3.1 Overview of Building Regulation and Building Codes for Regulating

Construction Project Activities in Ghana
Building regulations had been in use by various municipal administrations from the 1940s
and in 1960. By the early 1970s the Code and Building Regulations were obviously out of
date and the Building and Road Research Institute (BRRI) of the Council for Scientific and
Industrial Research (CSIR) decided to produce a draft document for discussion,
modification and use as a basis for a final code to address the redundancy of the earlier
documents. The effort was undertaken by Dr. K. Amonoo Neizer and Parts 3 and 4 of the
Draft Code were published in 1977, followed by the Parts 1, 2 and 5 by 1988. With changes
118
in land use patterns, materials and construction methods as well as local government
structure, the Ministry of Works and Housing produced a document, the National Building
Regulations (NBR), 1996 which was gazetted as LI 1630. This document was to improve
upon the building regulations of the colonial times and to complement the BRRI Draft
Building Code.

The NBR (L.I. 1630) is a legislative instrument promulgated on 27 September 1996. This
regulation is set of rules and standards that must be followed to satisfy the minimum
acceptable levels of safety for buildings and non-building structures. The NBR is applicable
to the erection, alteration or extension of any building. L.I. 1630 consists of nineteen (19)
parts and one hundred and eighty-seven (187) regulations. The District Planning Authority
(DPA) is mandated by L.I. 1630 to implement the regulations on behalf of every local
authority. The DPA comprises heads of relevant departments of the local authority e.g.
District Town and Country Planning Manager, Head of District Works Department, District
Environmental Health Officer, District Fire Officer, Electricity Company of Ghana District
Manager, and the Ghana Water Company Ltd District Manager. The DPA appoints a
qualified building inspector who oversees and inspects daily work on buildings, erections
and installations to ensure compliance with the requirements of these regulations. The
nineteen parts of the regulations involve application of regulations and building plans; plot
development; site preparation and landscape; materials for building; structural stability;
structural fire precautions; access accommodation; air movement and ventilation; thermal
insulation; hearths, chimneys and heat-producing appliances; sound insulation; pest control
and protection against decay; drainage; sanitary conveniences; refuse disposal; water
supply; lighting and electrical installations; special requirements for rural building and
miscellaneous provisions (NBR, 1996).

Presently, there is an attempt to produce a Draft Building Code which will include all the
requirements of the existing building regulations and the BRRI Draft Building Code and be
in tune with requirements of a building code to answer current land use patterns, materials
for construction, construction methods, construction management practices, safety, and the
energy efficient use of resources. This document, the Draft Ghana Building Code has been
119
compiled under the auspices of the Ministry of Water Resources, Works and Housing
(MWRWH) with funding from the UNDP, through the National Disaster Management
Organization (NADMO). A technical committee made up of representatives of relevant
stake holding institutions in Ghana compiled the document, using the BRRI Draft Building
Code and the national Building Codes of India (2005) and Canada as templates (MWRWH,
2012).

This draft Ghana Building Code contains regulations which can be adopted immediately or
enacted for use by various departments, metropolitan, municipal and district assemblies,
private and public bodies. It lays down a set of minimum regulations designed to protect the
safety of the public with regard to structural efficiency, fire hazards and health aspects of
buildings. So long as these basic requirements are met, the choice of materials and methods
of design and construction is left to the ingenuity of the building professionals (MWRWH,
2012).

Part 11 of the Code covers the constructional planning, management and practices in
buildings; the storage, stacking and handling of materials, and the safety of personnel
during construction operations for all elements of a building as well as the demolition of
buildings. It also covers guidelines relating to maintenance management, repairs,
retrofitting and strengthening of buildings. The aspect of the code that deals with quality
indicates that the quality of a project should be planned for all activities from inception to
completion. It is desirable that the system planning gives adequate assurance and regulates?
how it shall meet project quality objectives. The system will cover a review of existing
requirements, subcontracting, materials, processes and controls during the process, auditing,
training of personnel, final inspection and acceptance. All activities shall be planned and
controlled. This development and overall pattern suggest that Ghana may benefit from a
systematic framework for regulating the quality of public services (Foster & Pushak, 2011).

120
7.4 LESSONS LEARNT FROM TOTAL QUALITY MANAGEMENT
IMPLEMENTATION IN THE GHANAIAN CONSTRUCTION INDUSTRY
In most developing countries such as Ghana, the principles of TQM are not fully employed
in construction, leading to low quality works, high construction costs and dissatisfied
clients. Foster and Pushak (2011) stated that despite Ghana’s success with increasing access
to infrastructural services, the quality of service remains low. Nevertheless, in the
construction industry the aim of all professionals/consultants is to strive to provide the
owner/client with a product or service that will meet the standard that it is intended for,
properly built with satisfactory performance and value for money. In the pursuit of
performance excellence in the construction industries and with an increasing awareness of
the construction quality, construction firms have no option other than to ensure that clients
get value for money. One of the means of achieving this is by fully applying the principles
of TQM in the Ghanaian construction industry.

In furtherance of this aim is the effort by the Government of Ghana to produce a Draft
Building Code which will include all the requirements of the existing building regulations
and the BRRI Draft and be in tune with requirements of a building code to ensure good
quality of both materials and finished goods/products that are produced in the country. In
other words, this Building Code will help to provide a national standard for a quality
management system in Ghana. Lad and Beck (2009) noted that construction quality
management is the key to a successful project. A study conducted by Rategan (1992)
indicated that a 90% improvement rate in employee relations, operating procedures,
customer satisfaction, and financial performance is achieved due to TQM implementation.
Studies on TQM further showed a reduction in cost and improved employee satisfaction
that ultimately increase client satisfaction (Low & Jasmine 2004). According to Harris et al.
(2006), a TQM approach is essential for long-term survival in all business, including the
construction industry. Although TQM is not fully implemented in Ghanaian construction
industry, there is the National Building Regulations (NBR) (L.I. 1630) which was enacted
in 1996 to ensure that building work satisfies minimum constructional standards. The
requirements of this standard primarily aim at achieving client/customer satisfaction by
preventing non-conformity at all stages, from design through to servicing. It is therefore

121
suggested that an extensive enlightenment campaign on the importance of TQM in
construction will be required in Ghana to encourage its use. It was also recommended that
government, being the major client of the construction industry, should come up with an
award for excellence for quality construction to deserving organisations and a penalty for
organizations which refuse to adopt TQM. This will serve as an incentive for conformity to
standards in various areas of the industry and as a deterrent to defaulters.

7.5 CONCLUSION
The chapter discussed TQM implementation in the Ghanaian construction industry. The
chapter looked into the construction industry in Ghana and how the TQM concept has been
accepted in this country. An evaluation of policies with regard to quality management
implementation was also presented. Also explored in the chapter are the challenges facing
TQM implementation in the Ghanaian construction industry. It was realized that in Ghana,
the principles of TQM are not fully employed in construction, leading to low quality works,
high construction costs and dissatisfied clients.

122
 CHAPTER EIGHT
8 RESEARCH METHODOLOGY
8.1 INTRODUCTION
This chapter discusses the various steps undertaken by the researcher to explore the
objectives of this research. As discussed in the introduction, the aim of this research is to
develop a holistic integrated TQM model for the Ghanaian construction industry. Company
vision and plan statement, product selection and design management, and company quality
system evaluation are new constructs that are peculiar to the present model to be validated
as they have not been previously considered in the existing models of TQM whilst other
variables have been measured in the majority of the previous studies. This chapter provides
details about the methodological research framework for the current study. The chapter
consists of the following sections: philosophical consideration in research methodology,
the quantitative study and the qualitative study, research design and methodology. The
research design and methodology section focuses on the research procedures, including the
choice of research methods and the selection of participants. The methodology adopted for
this study is based on the ontological and epistemological assumptions. These
philosophical underpinnings (epistemology and ontology) provide a guide to the
methodology followed in a research process. Hence, this investigation combined
quantitative and qualitative methods (mixed method): the Delphi study and a structured
questionnaire survey. The use of a mixed-method approach which is rooted in both
philosophical and practical reasons, are explained in section 8.3.3.

8.2 PHILOSOPHICAL CONSIDERATION IN RESEARCH METHODOLOGY

As stated above in Section 8.1, the choice of research methodology is usually influenced by
a set of assumptions underlying each research methodology (Crotty, 1998). The choice of a
method has to be supported by the statement of assumptions that have been brought into the
research process and are reflected in the methodology (Crotty (1998). These assumptions,
though varied, tend to fall into the philosophical areas of ontology and epistemology. A
brief discussion of these considerations follows in Sections 8.2.1 and 8.2.2.

123
8.2.1 Ontological Considerations
Ontology is concerned with assumptions about the variety of phenomena in the world. It is
said to be a theory of the nature of reality (Delanty & Strydom, 2003); it is a theory of being
and is concerned with issues of what exists and also refers to the claims that a
particular paradigm makes about reality or truth (Hitchcock & Hughes, 1989).

According to Hitchcock and Hughes (1989), “...ontology is about what exists, what it looks
like and how is the components interact with each other”. Likewise, as with epistemology,
these issues can sometimes have a major impact on methodology, and any contrasting
ontology of human beings can, in turn, sometimes demand different research methods
(Burrell & Morgan, 1979). Ontological assumptions revolve around the question of ‘what
is’ with the nature of reality (Crotty, 1998). In other words, it is an effort to elucidate what
reality is and why things happen the way they do. In an attempt to explain reality, Jean
(1992) advocates two opposite assumptions of reality, which are objectivity and
subjectivity. Jean (1992) viewed the objectivist stance as reality existing out there, intact
and tangible, but it is independent of individuals’ appreciation and cognition (Crotty, 1998;
Jean, 1992). Thus, regardless of whether or not individuals perceive and attach meaning to
this reality, it remains unchanged (Burrell & Morgan, 1994). Hence an individual is “...born
into and lives within the social world that has its own reality, which cannot be created by
that individual” (Burrell & Morgan, 1994:4). Thus, in order to create a better understanding
of reality, objectivist researchers propose the need to study the causal relationships among
the elements constituting reality (Jean, 1992; Burrell & Morgan1994) which is advanced in
the current research.

According to Crotty (1998), “… [the] objectivist view of reality is closely related to a

theoretical position called positivism”. Positivists postulate that the world exists as a
system of observable variables waiting to be discovered (Maguire, 1987). Similarly,
positivists believe that the use of scientific methods of inquiry can assist in discovering the
true meaning of reality (Maguire, 1987; Crotty, 1998). The results of such investigation
generate rules and theories that help to explain and sometimes provide a guide for
understanding social behaviour (Maguire, 1987). The current research aims to apply

124
scientific methods in the development of the model. In addition, it aims to devise the social
construct that will lead to TQM in the Ghanaian construction industry.

According to subjectivist interpretation, reality is not discovered but it is constructed by

human beings as they engage with the world in which they live (Crotty, 1998). In that way
understanding and interpretation of reality occur when human beings interact with their
environment and others and assign meaning to the world around them (Crotty, 1998). Thus,
in research, meaning of anything is “…an expression of the manner in which the researcher
as a human being has arbitrarily imposed a personal frame of reference on the world” (Jean,
1992: 89). Section 8.3.2 looked into ‘how’ reality or knowledge is created as the extension
of the discussion of philosophical suppositions that influence researchers’ choice of
methodology.

8.2.2 Epistemology
Epistemology is theory of knowledge embedded in the theoretical perspective that informs
the researcher (e.g. objectivism, subjectivism). Delanty and Strydom (2003) defined
epistemology as the study which investigates the possibility, limits, origin, structure,
methods and truthfulness of knowledge and how knowledge can be acquired, validated
and applied. Walker and Evers (1988) explained that epistemology is concerned with how
phenomena can be made known to the researcher. It is an inquiry of what differentiates
defensible belief from opinion. Epistemology can sometimes also have a major impact on
the data collection choices as well as on the methodology in a research process (Hitchcock
& Hughes, 1995: 19). Epistemology provides the grounds for the decision on the kind of
knowledge that is considered appropriate, adequate and legitimate for the research at hand
(Crotty, 1998). Furthermore, Hill (1995) was of the view that research methodology is
applied epistemology, and therefore methodology has to be supported by an epistemology.
Therefore, researchers are expected to point out, explain and justify the epistemology that
informs their choice of research methodology.

Consequently, the choice of epistemology is widely influenced by the ontological

considerations within a particular discipline (Quattrone, 2000). Both dimensions of
125
ontology (objective and subjective) play an important role in the epistemology and
ultimately, the methodology chosen for this research. Sections 8.2.1 and 8.2.2 clearly show
how ontological dimensions (objective and subjective) and epistemological considerations
affect the choice of research methodology.

8.3 QUANTITATIVE VERSUS QUALITATIVE RESEARCH METHODOLOGY

Qualitative research is collecting, analyzing, and interpreting data by observing what people
do and say, whereas quantitative research refers to counts and measures of things
(Anderson, 2006:3). Quantitative research options have been predetermined, a large number
of respondents are involved and measurement must be objective, quantitative and
statistically valid. Statisticians use formulas to calculate the sample size to determine how
large a sample size will be needed from a given population in order to achieve findings with
an acceptable degree of accuracy (Anderson, 2006).

Qualitative research is much more subjective than quantitative research and uses very
different methods of collecting information, mainly individual in-depth interviews and
focus groups. The nature of this type of research is exploratory and open ended. A small
number of people are interviewed in-depth, and a relatively small number of focus groups.
Participants are asked to respond to general questions. The interviewer or group moderator
probes and explores their responses to identify and define people’s perceptions, opinions
and feelings about the topic or idea being discussed and to determine the degree of
agreement that exists in the group. The quality of the findings from qualitative research is
directly dependent upon the skills, experience and sensitivity of the interviewer or group
moderator. This type of research is extremely effective in acquiring information about
people’s communications needs and their responses to and views about the specific
question (Anderson, 2006). This chapter explores the methodological options available to
undertake research. Therefore, an appropriate research methodology must be considered by
the researcher.

126
The methodology to be adopted is influenced by the research aim, as well as the type of
data to be collected. The choice of the different options of methodology is either
quantitative or qualitative or both. Any one of the methodologies or a combination of both
methodologies can be carried out by the researcher to provide appropriate answers to the
research questions. These methodologies are sometimes influenced by the research
paradigms (Jean, 1992) despite the influence of the research aim and the type of data to be
collected. A paradigm is a set of beliefs that researchers use to make sense of the world or a
segment of the world (Crotty, 1998:35). In other words, a paradigm provides insight into
the way in which researchers look at and perceive the world (Kuhn, 1996). Paradigms guide
the conceptual framework that researchers use in seeking to understand and make sense of
reality (Maguire, 1987). Paradigms set boundaries for researchers in terms of the manner in
which they can execute the research process with regard to research methods, strategies for
inquiry, as well as the purpose and use of knowledge (Crotty, 1998). Thus, paradigms
influence what researchers regard as accepted knowledge and ways of doing research
(Crotty, 1998) and shape researchers’ “…perceptions and practices within their research
disciplines” (Maguire, 1987:11).

Sometimes, the choice of method is typically influenced by major philosophical

considerations (ontological and epistemological) underlying the research process (discussed
in section 8.2). Both quantitative and qualitative research methodologies are based on the
epistemological assumptions regarding the nature of knowledge and the methods of
abstracting that knowledge, as well as ontological assumptions which relate to the nature of
reality or the phenomena being investigated (Jean, 1992). The philosophical considerations
which influenced the choice of the research approach for this thesis were discussed briefly
in section 8.2.

8.3.1 Quantitative Methodology

Quantitative research is about quantifying the relationships between variables and finally
measuring them. Statistical models are constructed to explain the observed variables.
Certain characteristics (variables) are searched for and endeavour to show something
interesting about how they are distributed within a certain population. The nature of the
127
research determines the variables in which the researcher is interested. A variable needs to
be measured for the purpose of quantitative analysis. Data is collected concerning many
variables through a questionnaire. The interested variables may
be dependent or independent. There are other features present in the problem that may be
constant or confounding.

The objectivist view of an integral and independent reality encourages researchers to adopt
the epistemology of positivism (Jean, 1992). The objectivist researcher strives to observe
measure, analyse and predict relationships between components that comprise reality (Kent,
1999: 11). He further stated that certain principles guide a positivist’s search for reality.
These principles include the following:
i. Only phenomena that can be observed can be used to validate knowledge;
ii. Scientific knowledge is arrived at through the accumulation of verified facts derived
from systematic observation or record-keeping;
iii. Scientific theories are used to describe patterns of relationships between these facts
to establish causal connections between them; and
iv. The process is neutral and judgment free. Observations are uncontaminated by the
scientist’ own prediction. Thus ethical issues can be included only if they are
included as part of the research.
The use of a scientifically guided research methodology where the aim explains and
predicts causal relations between elements that constitute reality is the positivist
epistemology (Jean, 1992; Quattrone, 2000). The current research adopted the positivist
epistemology. Data collected should be quantified and analysed using mathematical
formulas (Maguire, 1987) and this shows the success of positivist research. The scientific
presentation of results follows the views of positivist researchers (Kent, 1999). Statistical
rhetoric such as reliability, validity, correlation, causes and effect relationships are used in
the current research. However, the presentation of research findings under this methodology
follows an approach that emphasises explicit, exact, scientific and formal procedures just as
the use of quantitative methodology to explore and explain relationships between variables
is advocated by the positivist’s researchers (Sarantakos, 2005). This argument makes the
entire research process to be considered as highly neutral and judgment free with limited
128
room for personal bias. According to Sarantakos (2005:33), “…the task of the researcher is
to discover the scientific laws that explain human behaviour using quantitative methods,
similar to those of natural sciences”.
A researcher using quantitative methodology has to follow a number of steps in conducting
their research which usually include generating the research problem, coming up with
expectations based on reality, generating hypothesis, defining variables, sampling, data
collection, analysis of data, report of findings and relating findings to the theory (Kent,
1999:11). May (2001) indicates that the researcher and the research process move together.
Their perceptions, expectations, experiences and interpretations become part of the research
process. According to May (2001), the relationship between the researcher and the research
should be a continuous ebb and flow of information. Therefore the researcher’s subjectivity
is considered as an integral part of the research process. Advocates of subjectivity suggest
that it is a better option for undertaking research as opposed to objective quantitative
methods (Brieschke, 1992). Sarantakos (2005) asserted that scientific research with the
emphasis on explicit, exact, and formal procedures is appropriate for a quantitative
methodology. From the given scenarios, quantitative methodology should be supported by
qualitative methodology (Tashakkori & Teddlie, 1998).

Section 8.3.3 discusses the advantages of combining qualitative and quantitative

methodology in more detail. In the following section qualitative methodology is discussed.
The discussion of qualitative methodology in this section focuses on both the advantages
and limitations of qualitative methodology.

8.3.2 Qualitative Methodology

Qualitative research allows the subjects being studied to give much ‘richer’ answers to
questions put to them by the researcher, and may give valuable insights that might have
been missed by any other method. Not only does it provide valuable information to certain
research questions in its right but there is a strong case for using it to complement
quantitative research methods. If the area of interest has not been previously investigated
then qualitative research may be a vital forerunner to conducting any quantitative research;
for example, it is impossible to carry out a meaningful structured questionnaire survey on
129
TQM. At the other extreme, qualitative research may also help the researcher to understand
the findings of quantitative research; for example, it is very easy to discover that some
construction firms fail to implement TQM but uncovering the reasons for this can be more
difficult and conventional surveys may miss some of the important factors.

According to Jean (1992), “…the subjectivist’s view of reality advocates for appreciation of
human involvement in the creation and shaping of knowledge”. The subjectivist
epistemology thus suggests that meaning or reality is not discovered but rather is imposed
on the object by the subject, and in a research situation, imposed by the researcher (Crotty,
1998). In other words, with the subjectivist epistemology, the object being studied
contributes less to the meaning or reality. Thus, researchers’ input in the research process is
recognised under subjectivism. The research methodology recommended by subjectivists is
qualitative methodology. According to Jean (1992:92), qualitative research is “…a form of
social interaction in which the researcher converses with and learns about the phenomenon
being studied”. It is part of the research process and is actively involved in creating the
meaning of reality (Crotty, 1998). Qualitative research is suggested as more applicable to
the study of people and their environment (social sciences) than natural sciences (Bryman,
2001). Consequently, advocates of qualitative research advanced the use of qualitative
methodology when studying people as it enables the researcher to see through the eyes of
the researched (Bryman, 2001). Constructing meaning through engagement with people
involves interpretation. Interpretivism is the process by which information is extracted
through interpretation as pointed out by Sarantakos (2005). Under interpretivism,
researchers seek information relating to people’s views, opinions, perceptions and
interpretations of the social world (Crotty, 1998), which was also partly utilized in the
current research. Subjectivism, constructivism, and interpretivism form part of a broader list
of research methods commonly employed in qualitative research.

The qualitative methodology has been criticized for lacking in efficacy owing to its inability
to study the relationships between variables with a degree of accuracy (Sarantakos, 2005).
The researcher in qualitative research is the main player. He or she decides on what to
concentrate on during the data collection. Views may vary among different researchers and
it is difficult to replicate and generalize the findings with ease. A small number of cases is
130
studied as compared to the large sample sizes common in quantitative cases (Bryman,
2001). The population under the study may not be representative of the majority. However,
advocates of qualitative research argue that generalizations are made on the assumption that
the findings and inferences made during the research are supported by sound theoretical
reasoning (Mitchel, 1983). Representation in qualitative research is in accordance with the
subject of investigation, which is highly subjective and a narrow-minded view of events and
what is being observed (Ruyter & Scholl, 1998). The findings of qualitative research are
difficult to subject to rigorous quality verification requirements such as reliability and
validity (Creswell, 1994). It would be difficult to prove the validity of qualitative research
findings through measurement. Validity requires measurement of the object of enquiry and
that is not possible in qualitative research because its purpose is not to measure but to
generate ideas (Stenbacka, 2001).

On the other hand, reliability is concerned with producing the same result with consistency.
This is not possible under qualitative research because of the involvement, influence,
subjectiveness and the possibility of the bias of the researcher in qualitative research.
Qualitative researchers have, however, argued that quality verification using validity and
reliability checks is not necessarily applicable to qualitative research because it owes its
origin to scientific rhetoric and positivist paradigms common in quantitative research
(Creswell, 1994; Stenbacka, 2001). Stenbacka (2001:555), in further defending the
paradigm, suggests that “...new concepts relevant to qualitative research [have] been used
instead of quality concepts borrowed from quantitative research”.

Nevertheless, both qualitative and quantitative research methodologies can be used in

different situations, depending on the aims and objectives of the study. Most research is
centered on four primary objectives, namely “exploration, explanation, description and
prediction” (Ellram, 1996:98). Research where the objectives are either exploration or
explanation, or both, would normally require qualitative research methods. This is because
qualitative research has the ability to provide insight and explanation into a phenomenon
that was relatively unknown (Ruyter & Scholl, 1998). It provides answers to questions,
such as ‘how’ or ‘why’ which are common in exploration and explanation of phenomena
(Ellram, 1996). Research that is descriptive or predictive, or both, would require
131
quantitative research methods that utilize statistical techniques to predict and describe
relationships between variables (Ellram, 1996). The choice between the two methodologies
should be based on the aims and objectives of the study, as well as the nature of the study.
In some cases the two methods may be used jointly to cover for the weaknesses inherent in
each method (Tashakkori & Teddlie, 1998; Amaratunga, Baldry, Sarshar & Newton, 2002).
A similar process is known to Tashakkori and Teddlie (2003) as mixed method or
pragmatism. The combination of qualitative and quantitative methods is discussed in
section 8.3.3.

8.3.3 Combined Quantitative and Qualitative Methods

Mixed-methods research involves both collecting and analyzing quantitative and qualitative
data. Quantitative data includes closed-ended information such as that found on attitude,
behaviour or performance instruments. The collection of this kind of data might also
involve using a closed-ended checklist on which the researcher checks the behaviours seen.
The analysis consists of statistically analyzing scores collected on instruments, checklists,
or public documents to answer research questions or to test hypotheses. In contrast,
qualitative data consists of open-ended information that the researcher gathers through
interviews with participants. The general, open-ended questions asked during these
interviews allow the participants to supply answers in their words. The analysis of the
qualitative data (words or text or images) typically follows the path of aggregating the
words or images into categories of information and presenting the diversity of ideas
gathered during data collection. The open-ended versus closed-ended nature of the data
differentiates between the two types better than the sources of the data (Elliot, 2005).

Several researchers such as Uysal and Crompton (1985), Creswell (1994), Tashakkori and
Teddlie (1998), Bryman (2001), and Amaratunga et al (2002) have supported theoretically
the combination of qualitative and quantitative methods. Tashakkori and Teddlie (1998)
asserted that quantitative and qualitative research methods are not dichotomous but rather
complement one another to produce improved research findings.

132
However, those advocating for the use of combined methods rejected the choice between
positivism and constructivism as none of the methods work best in isolation (Tashakkori &
Teddlie, (1998). The use of combined methods, often called the mixed method, has been
found to alleviate the weaknesses linked with using either of the methods on their own
(Tashakkori & Teddlie, 1998; Bryman, 2001; Amaratunga et al., 2002; Mangan, Lalwani,
& Gardner, 2004). For example, Bryman (2001:450) suggests that “...in some instances
neither qualitative nor quantitative research methods may be adequate on their own, thus
researchers cannot rely on just one method and have to use both to support the research
process”. Quantitative and qualitative methods supplement each other by providing richness
and details that are otherwise unavailable if each method were pursued separately (Jack &
Raturi, 2006). Combining the methods provides a multidimensional insight into the research
problem, and thus assists in getting a broader understanding as well as a true analysis of the
situation at hand (Mangan et al., 2004), which is also one of the strong points of
consideration for the current research. The use of combined methods compensates for the
weakness embedded in each of the research method by “...counter-balancing the strengths
of another” (Amaratunga et al., 2002:23). Jack and Raturi (2006) informed that
triangulation provides confirmation of the research findings by improving the ability of
researchers to draw conclusions from their studies, thereby resulting in more robust and
generalizable research findings. The research design used in this research has been
discussed in the section. The study adopted a mixed-method methodology in order to
counter-balance the strengths and weakness embedded in each of the research methods
when used separately, as discussed earlier. Further details on the justification for and
explanation of how quantitative and qualitative methods were used to collect data in this
thesis are provided.

8.3.4 Mixed-Methods Approach

Mixed-methods research involves the use of more than one approach to or method of
design, data collection or data analysis within a single programme of study, with integration
of the different approaches or methods occurring during the programme of study, and not
just at its concluding point (Johnson et al., 2007:120). Mixed-methods research is a research
methodology with philosophical assumptions as well as methods of inquiry. As a
133
methodology, it involves philosophical assumptions that guide the direction of the
collection and analysis of data and the mixture of qualitative and quantitative approaches in
many phases in the research process. As a method, it focuses on collecting, analyzing, and
mixing both quantitative and qualitative data in a single study or series of studies. Its central
premise is that the use of quantitative and qualitative approaches, in combination, provides
a better understanding of research problems than either approach alone (Creswell & Clark,
2007). The mixed-method approach was adopted in this thesis based on the philosophical
and practical reasons discussed earlier. A quantitative survey provides a snapshot of
phenomena while the qualitative method or interview from the Delphi panelists (Delphi
study) provide contextual information and human subjective information to interpret and
inform the quantitative results. Creswell, Clark, Gutmann and Hanson (2003) identified six
commonly used designs in mixed-methods research. However, the present study uses two
of those, namely sequential explanatory and concurrent triangulation design. A visual
model of mixed methods, as discussed by Creswell et al (2003) and Teddlie and Tashakkori
(2009) is also used in the present study (Figure 8.1) to summarize and clarify the procedure.
The quantitative survey is the main driver of this study, complemented by the qualitative
study. The use of both methods provides a richer understanding of phenomena, an
explanatory account of triangulation and illuminates significant survey findings in what
Teddlie and Tashakkori (2009) referred to as crossover track analysis. Although the
quantitative and qualitative studies are independent, both sets of data and analyses are used
in analysis. The survey (quantitative) examination of the relations and associations between
the key variables and the Delphi study (qualitative) are presented in the preceding chapters.

The current study used mixed-methods research, which involved both quantitative and
qualitative approaches. According to Tashakkori and Teddlie (2003), mixed-methods
research is a research design with philosophical assumptions as well as methods of inquiry.
As a methodology, it involves philosophical assumptions that guide the direction of the
collection and analysis of data and the mixture of qualitative and quantitative approaches in
many phases of the research process. It focuses on collecting, analysing, and mixing both
quantitative and qualitative data in a single study or series of studies. The use of
quantitative and qualitative approaches in combination provides a better understanding of
134
research problems than either approach alone. The quantitative data in a typical mixed
method includes closed-ended information which was also adopted in this thesis. The
collection of this kind of data involves using a closed-ended checklist in which the
researcher checks the behaviour observed. Sometimes quantitative information is found in
documents such as census records or attendance records. The analysis of the qualitative data
consists of statistically analysing scores collected on instruments and checklists to answer
research questions, to test hypotheses or to answer the research questions (Creswell, 2003).
In contrast, qualitative data consists of open-ended information that the researcher gathers
through interviews with participants. The general open-ended or closed-ended structured
questions asked during interviews allow the participants to supply answers in their own
words.

QUAN Data
Collection
QUAL Data
Collection

QUAN Data
Collection QUAL Data
Analysis

Comparison and integration of

QUAN and QUAL results

Figure 8.1: Visual Model of Mixed Methods Design

Source: Adapted from Tashakkori and Teddlie (2003); Aigbavboa (2013: 248)

These can be thematically analysis and converted into qualitative data, which can also be
transcribed in quantitative data; for instance, when the Delphi technique is used,
frequencies of measures of central tendencies are used to draw consensus. Qualitative data
may also be collected by observing participants or sites of research, gathering documents
from a private or public source, or collecting audio-visual material such as video-tapes or
artefacts. The analysis of the qualitative data (words or text or images) typically follows the
path of aggregating the words or images into categories of information and presenting the
135
diversity of ideas gathered during data collection. The open- versus closed-ended nature of
the data differentiates between the two types better than the sources of the data. Mixed-
methods research is commonly used as a strategic research approach that is able “…(a) to
demonstrate a particular variable will have a predicted relationship with another variable
and (b) to answer exploratory questions about how that predicted (or some other related)
relationship actually happens” (Tashakkori & Teddlie 2003:15).

Both qualitative and quantitative methods were used in the current study to identify the
factors that are associated with TQM practices. It also indicates the statistical significance
of these factors in determining TQM in the Ghanaian construction industry. The current
study also looked into the relationship between the identified factors and TQM to be tested
(predicted). The impact of the identified independent variables on the TQM implementation
was also soughted in this study. The use of qualitative method was employed to explore and
gain a comprehensive understanding of the way the selected factors had an influence on
TQM. A quantitative method was further used to verify the results in the survey. The
factors that bring about TQM in the construction industry were obtained through the use of
the mixed-method approach.

The mixed-method approach confirms the findings of both the quantitative and qualitative
approaches (Flick, 2009). This implies that the findings of the study could be useful in the
establishment of the factors that should be considered for the development of TQM
principles. The current study adopted a qualitative method structured through the use of the
Delphi technique. The Delphi technique has been elaborated on in section 8.4.3. Findings
from the Delphi survey were used to refine the survey tool (structured questionnaire) for the
study and to validate the findings. The Delphi findings were used to resolve issues on TQM
in the construction industry through consensus. The quantitative method of data collection
for the study was the survey method with the use of a structured questionnaire. The analysis
was done using SEM with EQS Version 6.2, using SPSS in the development and validation
of the TQM model.

136
8.3.5 Justification of the Mixed-Method Approach
Both quantitative and qualitative methods have their strengths and weaknesses. Quantitative
methods have been criticized for being “…sanitized and lacking in contextual realism”
(Tashakkori & Teddlie 2003:516). Qualitative methods are suitable for addressing questions
of how and why things occur while quantitative methods are more appropriate for
answering ‘what’ and ‘how’ questions (Yin, 1994). The use of one method is not
appropriate in studying the variables that predict TQM in the construction industry and was
not enough to explore from the various stakeholders (identified in the qualitative Delphi
study as experts). The use of the mixed-method approach that integrated qualitative and
quantitative methods was required. According to Tashakkori and Teddle (2003:518),
“…one of the merits of a Mixed Method Approach in the current study is that the
techniques of the qualitative and quantitative domains, which are interwoven, helped to
maximize the knowledge yield of the research endeavour”. The mixed-method approach
was used by the researcher to discover and justify the model components within one study.

Qualitative research involves people in order to provide the realism and detail needed for
the generation of hypotheses and building of theory (Tashakkori & Teddle, 2003). Detail-
rich data was collected for this study by using qualitative techniques. The language and
context of the stakeholders and the people being studied were captured during the
questionnaire survey to collect the required data for the study. The qualitative data gathered
in both the first and second stages and the data analysis approach were aligned with the
positivist paradigm. The positivist paradigm sought to identify patterns and repetition
within each key research issue and also explored the level of impact, influence and
agreement through the use of scales. The mixed-method approach was adopted to answer
questions that would not have been possible to answer by either qualitative or quantitative
approaches alone. The research problem was addressed through the use of the mixed-
method approach which enabled the researcher to use both methods. The generalisability of
the research result was increased through the use of mixed-method approach which was a
major consideration in the present study. It also provided stronger evidence for a conclusion
through the convergence and verification of findings. It also added insights and
understanding that would have been missed if only a single method was used. In addition, it
137
provided complete knowledge necessary to inform theory, practice and to be able to answer
a broader and more complete range of research questions because the researcher was not
confined to a single method or approach. Individuals who intend to solve problems using
both numbers and words (combination of inductive and deductive thinking) naturally to
employ mixed methods research as the preferred mode of understanding the thesis
statement.

The Delphi technique was combined with the survey method in the current research to
provide the basis for the validation of the conceptual framework for the development of a
holistic TQM model for the Ghanaian construction industry and the major cities in Ghana
as a case study.

8.4 RESEARCH DESIGN

A research design gives a detailed outline of how an investigation will be carried out. It will
typically include how data will be collected, what instruments will be used, how the
instruments will be used and the intended means for analyzing data collected. The research
design has been defined as the framework for conducting research. It assists the researchers
to conduct the study successfully (Churchill, 2001). Usually, the research design is used to
justify decisions and choices relating to the research procedure (Sekaran, 2000). This
section of the thesis provided the outline of the research design as shown in figure 8.2. The
appropriate methodology used in this study is based on the ontological and epistemological
assumptions. The choice of research design is influenced largely by the methodology
(whether quantitative or qualitative) as well as the philosophical assumptions guiding the
research process (ontology and epistemology). Objectivist ontology influences a researcher
to adopt a more positivist epistemology. This gives emphasis to the use of quantitative
methods in the research process (Sarantakos, 2005) such as the constructivist ontology,
which culminates in qualitative methodology. The research design was fixed in line with
the requirements of objectivism, which favours a scientific way of abstracting data when
objectivist ontology is adopted. The instruments used in collecting data were determined by
the research design. In quantitative design, the survey method is used to collect the data.
The philosophical underpinnings (epistemology and ontology) provide a guide to the
138
methodology followed in a research process. A decision on the specific research design can
be hypothesised in the form of a connection beginning from the philosophical
underpinnings (epistemology and ontology).

ONTOLOGY

EPISTEMOLOGY

METHODOLOGY

DESIGNS

INSTRUMENTS

Figure 8.2: Steps in the research design process

Source: Adapted from Sarantakos (2005:29); Aigbavboa (2013: 253)

Following the decision on the methodology, the researcher has to decide on the research
design guided by the research questions and aim. The research design influences the the
choice of instruments to use in the execution of the research process (Sarantakos, 2005).
Figure 8.2 illustrates these connections. The exact justifications for a research design should
follow the five aspects - research purpose, theoretical framework, research questions,
research methods and sampling strategy – which are appropriately inter-connected,
according to Robson (2002). The current study follows these aspects of research design.
Hence, the choice of research methods for the current study was influenced by the research
aim, sub-questions and objectives.

Three considerations were made for the selection of the research methods to answer the
predetermined set of goals for the research. These are:
i. To be able to identify the variety of factors associated with TQM in the
construction industry;

139
 ii. To be able to predict the relationship between each of the identified factors and how
they can predict TQM in the construction industry; and
iii. To be able to provide in-depth information to be collected and analysed to show
how firms identified factors as important (influential) in determining TQM in
the construction industry.

The current study adopted the mixed method of research (quantitative and qualitative
combined) as stated, discussed and justified in the previous sections. The mixed-method
approach was adopted to answer the research questions and meet the research objectives,
thus, developing a TQM model that applies to the study area.

The following strategies were adopted to meet the research objectives stated in Chapter 1,
Section 1.6.2:
i. The first general objective was to establish the factors that determine TQM in the
construction industry. To this end, a literature review was conducted relating to
the factors that determine TQM implementation. Published articles,
development reports and status reports were reviewed. The expected outcome
from this objective was information and a global picture of the determinants of
TQM implementation. Both international and national literature was reviewed.
The review gave a general overview of how TQM in the construction industry is
formed and its relevance to construction firms.
ii. The second general objective was to establish the current theories and literature on
TQM implementation as well as to identify the gaps that needed consideration.
The constructs that were established were included as part of the theory for the
development of the holistic TQM model. Literature review was conducted from
a wide source of publications including journals, conference proceedings, books
and monographs. Specific theories of TQM studies were reviewed. The second
general objective was expected to provide information on the current theories on
TQM studies to determine the gaps which other scholars have not yet addressed,
to identify common themes and the type of methodologies that have been used

140
 in the research and how terms have been defined. This information was
necessary as it was the core literature to inform the current research project.
iii. The Delphi method was used to achieve the third and fourth objectives which were
to determine the main and sub-attribute(s) that bring about TQM and to examine
the similarities in attributes that determine compliance in other cultural contexts
as well as Ghana. In addition, the study aimed to evaluates the factors and
issues that affect the implementation of TQM in Ghana. Since experts’ views
were required on the factors that determine TQM, the Delphi Method was the
best to use in this instance. Apart from experimental procedures which were not
feasible for this study, the Delphi or focus groups, or both, could be used.

The focus groups could have been used except that there was the challenge of assembling
all workers and experts to deliberate from 8.00 am to 16.00 pm for a minimum of three
days. The focus group was also expensive and beyond the budget for the field work. This
might have defeated the purpose of conducting a rigorous process to achieve the objective.
This is not the case for a focus group. There was no bias from the experts because they
remain completely anonymous to each other and therefore there was no undue influence
from peers. A detailed explanation of the Delphi technique is explained in Section 8.4.3 in
order to give an idea to the reader of how the Delphi method was conducted and what
should be expected. The expected output is an estimation of the extent to which TQM was
influenced by the established factors. Also the consensus reached on the critical factors and
issues that affect the non-implementation of TQM in the Ghanaian construction industry.
The conceptual model was developed for the TQM from the factors and their
interrelationships
iv. The fifth general objective was achieved by drawing on the conclusions from the
extensive literature review and the results and findings from the qualitative
Delphi study.
v. An empirical questionnaire survey was conducted and analysed using SEM to
achieve the sixth objectives of the research.
The sixth objective was to test and validate the conceptual model developed from the RO5.
Data obtained from the questionnaire sought to establish interrelationships between the
141
factors that determine TQM to establish the relationship produced amongst them and which
constructs have a greater influence on the determination of TQM in the construction
industry. The aim was to establish the core determinants of TQM in the Ghanaian
construction industry. A detailed explanation of the survey concerning population, sampling
procedure and analysis of results is presented in Section 8.4.4. The expected output for the
sixth objective was information to validate the conceptualized holistic model and based on
this, to finalize the best fit model for TQM for the Ghanaian construction industry. Table
8.1 summarises the research methods employed to achieve the objectives of the research.
The methods used to achieve the objectives of the research have been explained in the
sections that follow.

Table 8.1: Research Procedure

Stage Research Data collection Data analysis Output

objective Method method
1.0 RO1: To Literature Information and a
Literature establish the review global picture of
review factors that the determinants
determines of TQM in the
TQM construction
implementation industry.
in the
construction
industry

RO2: To Literature Information on

investigate review the current
theories and theories on TQM
literature on in the construction
TQM industry and to
implementation determine the
in the gaps which other
construction scholars have not
industry addressed.
Outline of
constructs
(factors)
associated with
TQM
implementation in

142
 the construction
industry have not
been
considered in the
previously
developed models
2.0 RO3: To Delphi Descriptive Consensus on the
Delphi determine the Technique statistics influence and
Technique main and sub- impact level of
attributes of the
TQM in the various attributes
Ghanaian of TQM
construction implementation in
industry the construction
industry

RO4: To Delphi Descriptive Consensus on the

evaluate the Technique statistics critical factors and
critical factors issues that affect
that affect construction firms
construction for the
firms for not implementation of
implementating TQM in Ghana
TQM in Ghana .

RO5: To Desk study; Theory An integrated

develop an Questionnaire TQM model.
integrated TQM Survey;
model to guide Delphi
construction Technique
firms in
managing
construction
projects in
Ghana

3.0 RO6: To Questionnaire Structural Information to

Questionnaire determine the Survey Equation validate
Survey validity of the Modeling conceptual model;
integrated TQM (SEM); validated best-fit
model through a EQS model.
comparative
analysis of the
Delphi and
literature
outcome with
143
 the field
questionnaire
survey outcome

Stage 1 Stage 2 Stage 3

Questionnaire Evaluate factors that
Influence on ’ TQM
Develop conceptual implementation
model based on the Delphi
determinants of
Total Quality Management Analyse & model
results to validate
conceptual model

Literature review
Best fit model &
recommendations

Need for study

Figure 8.3: Research Design Outline

Source: Manu, Ankrah, Proverbs and Suresh (2010:29), Musonda (2012:91), Aigbavboa (2013:258)

8.4.1 Methods
An overview of the methods to be used to achieve the objectives of the research is given in
detail in this section. The methods are a literature review, the Delphi method and the
questionnaire survey. A detailed description of each of the methods listed has been given
under the analysis (how data were treated) section. An outline of how the study was
conducted is given in Table 8.1. The table details the methods (procedures) employed to
achieve the objective of the study. Figure 8.3 is also an outline of how the study was
conducted and gives a detailed description of the methods described in the next sections.

8.4.2 Stage One: Literature Review

Boote and Beile (2005) asserted that literature is the foundation of research. The most
important aspect of developing a study is to review literature on previous works carried out
by different researchers as well as the methodologies employed to investigate similar

144
problem in order to establish the trends on the use to solve several problems (Heppner &
Heppner, 2004:52).
For the current study it is necessary to know the following:
i. The theoretical and conceptual perspectives of TQM research;
ii. Gaps in TQM implementation in construction industry research;
iii. TQM issues in the construction industry;
iv. TQM implementation in develop countries; and
v. TQM implementation in developing countries, specifically those in Africa.
The literature review on the above aspects was important as it provides the broad context of
the study to the reader. It also highlights what has already been done before on the subject
under consideration. It further relates the present research to the on-going debate on the
subject and provides a framework for comparing the results of the present research with
other studies on the subject.

An effort was made to ensure that the literature analysis was thorough and comprehensive
to ensure the integrity of the study. Studies reviewed were well integrated and adopted
methods have been used in other studies. The reviewed and detailed analysis of the methods
used in other studies was not only on report made in the existing literature (Boote & Beile,
2005). Several materials were used for the literature review, such as books, reviews of
articles on the subject matter (both published and unpublished), theses and dissertations.
Names of leading authors and contributors on the subject matter were sourced from the
references of the consulted articles. This helped to obtain their publication history and the
search for information focused within research databases. The progression of research on
the topic under study was made possible from articles in the mentioned sources. The
methods of conducting a literature review as indicated by Boote and Beile (2005) below
were strictly followed in the study.

The process of conducting literature review specifically followed, among others the
following steps (Boote & Beile, 2005:13):
i. Sourcing a broad range of high-quality, specific articles, books, dissertations and
reviews directly related to the study;
145
 ii. Reading and re-reading to establish progressions and trends;
iii. Summarizing the studies read;
iv. Identifying methodologies adopted in the studies;
v. Relating the current study to those reviewed; and
vi. Writing the literature.
The output from the literature review was a clear perspective of the topic and an indication
of where the study fits in relation to other studies on the subject matter. This also provided a
framework for comparing the results of the study with other studies. According to Boote
and Beile (2005), “…good research is good because it advances our collective
understanding”. A great deal of energy has been expended in the review of literature on the
current study. Findings from the literature review have indicated various factors which
determine TQM practices in different contexts. Other factors such as key constructs which
should bring about TQM should be considered. These factors were not considered in the
previous models developed since there are missing universal factors which give assurance
on TQM in the Ghanaian construction industry. Theories were developed about the
influence of the missing factors and their interrelationships with other factors to determine
TQM implementation in the construction industry. A test needed to be carried out to
ascertain their influence on TQM in the Ghanaian construction industry. This was achieved
by using theDelphi method described below.

8.4.3 Stage Two: Delphi Method

The Delphi technique was first or originally developed by two scientists, Olaf Helmer and
Norman Dalkey of the Rand Corporationin in US in the 1950s, and named after the greatest
of all Greek oracles (Buckley, 1995). In other words, Delphi technique was named after the
ancient Greek temple where the oracle could be found. The technique was developed as a
tool for forecasting and problem solving of complex topics at the Rand Corporation. Dalkey
and Helmer (1963) identified that the Delphi technique is a procedure to solicit opinions,
judgement, and consensus from a group of experts in a given subject field. They added that
the process is very interactive and solicits and collates opinions on a particular topic
through a set of designed sequential questionnaires interspersed with a summarized

146
feedback of opinions derived from earlier responses (Dalkey & Helmer, 1963). On the other
hand, Miller (1993) explained that the Delphi technique is a qualitative methodology
seeking to produce consensus of a group of experts on an issue of concern through a survey
consisting of rounds. Miller (1993) added that the technique is based on structural surveys
and makes use of intuitively available information of the participants (experts) in their
various fields. The method provides both qualitative and quantitative results, and has
beneath its explorative, predictive and even normative elements (Cuhls, 2003). Hence, there
is agreement that Delphi is an expert survey in two or more ‘rounds’ in which the second
and later rounds of the survey (the results) of the previous round are given as feedback.
Thus, the experts answer from the second round based on the influence of the other experts’
opinions but the panel members have no idea who is involved. The idea is that the joint
judgment of experts is a relevant measure of the outcome of the research. The Delphi
method was used during the second stage of the current study to identify the main attributes
that bring about an understanding of TQM in the construction industry and to examine
whether the attributes that determine TQM in other cultural contexts as identified from the
literature are the same within the construction industry in Ghana.

The Delphi technique requires knowledgeable and expert contributors individually

responding to questions and submitting the results to a central coordinator (researcher). The
coordinator (researcher) processes the responses, looking for central and extreme tendencies
and their validations (Grisham, 2008). The results are then fed back to the input provided by
the coordinator (researcher). The experts are then asked to resubmit their opinions, aided by
the input provided by the coordinator (researcher). This process continues until the
coordinator sees that consensus has been reached on the questions asked. The method was
intended to remove the bias that is possible when diverse groups of experts meet, which is
common with other methods of decision making. With the Delphi method, the experts do
not know who the other experts are.

Hence, the standard Delphi method is a survey which is directed by a coordinator as already
stated and comprises several rounds with a group of experts, who are kept anonymous and
for whose subjective-intuitive prognoses a consensus is aimed at (Cuhls, 2003). After each
147
survey round, standard feedback about the statistical group judgment calculated from the
median, the percentages and the interquartile range of single projections is given and if
possible, the arguments and counter-arguments of the extreme answers are fed back. In the
Delphi process, nobody ‘loses face’ because the study is done anonymously using a
questionnaire. Rowe, Wright and Bolger (1991) and Häder and Häder (1995) stated that it is
commonly assumed that the method makes better use of group interaction whereby the
questionnaire is the medium of interaction. The Delphi method is especially useful for long
range forecasting, as expert opinions are the only source of information available.

The Delphi technique is part of a group of decision-making (policymaking) techniques that

includes the nominal group technique (NGT) and the interacting group method (IGM). The
Delphi technique differs in various ways from the NGT and IGM respectively, but
primarily due to the fact that Delphi is individual based, anonymous and independent. The
element of group interaction is eliminated from the process and feedback to questionnaires
is in written format (Loo, 2002). Over time, the Delphi method has gained popularity as a
method of inquiry across many scientific disciplines. Czinkota and Ronkainen (1992) cited
in Aigbavboa (2013) indicated that the Delphi method has gained considerable approval
across disciplines. They further claimed that it has been used as a study instrument in the
fields of library and information science (Buckley, 1995), in the medical disciplines
(Linstone & Turoff, 1975), in multi-country studies of communications in Europe, and by
actuaries to predict economic conditions (SOA, 1999). Czinkota and Ronkainen (1992)
cited in Aigbavboa (2013) further report that those experienced with the Delphi technique
report that the method produces valuable results which are accepted and supported by the
majority of the expert community. Similarly, in the business field, the technique has been
highly rated by some as a systematic thinking tool, but has been challenged in its ability to
serve as an identifier of strategic issues (Schoemaker, 1993).

The Delphi technique is well suited as a research approach and method for the current
study. The technique has not been used in a similar study in Ghana or in any other
developing country. The study was aimed to attract a wide spectrum of inputs from various

148
geographically dispersed experts in Ghana. The Delphi method was preferred to common
survey methods as the current study was addressing the ‘what can - if’ kind of questions, as
opposed to the ‘what is’ kind of questions. Delphi is more suited for these kinds of
questions to explore concepts that are difficult to measure except through experimental
methods. Unfortunately, an experimental survey was not feasible or appropriate for the
current study. The Delphi’s strength lies in the rounds used, unlike ordinary survey
research, which provided an opportunity for initial feedback, collation of feedback, and
distribution of collated feedback to participants for further review (Stitt-Gohdes & Crews,
2004:62). Therefore, the Delphi method was also considered to be a robust method of
rigorous query of experts. This unique process requiring group communication is central to
the strength of the Delphi (Stitt-Gohde & Crews, 2004:62). Also, Loo (2002) opined that
the Delphi process should be used when investigating policy-making or policy-evaluation
strategies that will set the future direction for the public or private sector respectively. The
Delphi method was considered as a useful tool for the current study. The thesis is aimed at
future direction for TQM in the Ghanaian construction industry.

However, the Delphi method was alleged to have failed to follow accepted scientific
procedures, in particular the lack of psychometric validity (Sackman, 1974). Coates (1975)
was of the view that the Delphi technique is of value not in the search for public knowledge
but rather in the search for public wisdom; not in the search for individual data but in the
search for deliberative judgment. Sackman’s (1974) view is not applicable to the current
study. However, it should be noted that the approach deals with areas that do not lend
themselves to traditional scientific approaches; henceHelmer (1977) argues that the
forecasting tendency, one of the major applications of the Delphi, is inevitably conducted in
a domain of what might be called ‘soft data’ and ‘soft law’. Helmer (1977) further
determines that standard operations and research techniques should be augmented by
judgmental information and that the Delphi method cannot be legitimately criticized for
using mere opinion and for violating the rules of random sampling in the ‘polling of
experts’. Such criticism, Helmer (1977) argued, rests on a gross misunderstanding of what
the Delphi method is; it should be pointed out that a Delphi inquiry is not an ‘opinion poll’.
As all the above definitions illustrated, in no instance is reaching a majority opinion the
149
ultimate goal in a typical Delphi study; it is rather the reaching of agreement (consensus).
According to Buckley (1995), “Delphi is a tool for discovering agreement and identifying
differences rather than forcing consensus”. Buckley (1995) further informs that in principle,
agreement alone is not a sufficient condition for arguing the acceptance of the Delphi
method.

Linestone and Turoff (2002) asserted that majorities of research methods, the method of use
and application has an enormous influence on the eventual success of the inquiry. Hence,
where no agreement is achieved, the Delphi still helps to clarify the issue being
investigated. Linestone and Turoff (2002) asserted that one of the common reasons for
failure in a Delphi study is ignoring and not exploring disagreement. The current research is
not only about reaching or forcing a consensus but recognizing disagreement and exploring
the reason for such.

In addition to the above criticism of the Delphi technique, different authors have pointed
out their various views of the weaknesses of the Delphi technique as follows:
i. It has not been shown consistently that the results from the Delphi method are any
better than those achieved through other structured judgmental techniques
(Rowe, Wright & Bolger, 1991);
ii. There is possibility of biases of the coordinating or monitor team (researcher), who
chooses the experts, interprets the returned information and structures the
questions for the Delphi study. There is an enormous debate whether the experts
should be chosen from within or outside of the organisation initiating the study
and whether they should be experienced in the subject area of the study in
question (Masini, 1993);
iii. Linstone (1978) disagreed with the process and how the questionnaire was
structured, which Linstone (1978) believed can lead to bias (like IQ tests), which
assumes a certain cultural background. Hence, the experts may give responses
they think the monitoring group wants to hear, or they may not respond at all.
Consequently, the cultural background of respondents will have an impact upon
the results;
150
 iv. Simmonds (1977) debated that one of the key flaws in the Delphi technique is that
certain questions are not asked as they do not seem important when the study
begins. Nonetheless, once the study begins, new questions cannot be added,
which in turn can weaken the study considerably;
v. Lang (1995) states that the process of choosing the panelists is often not considered
seriously enough. Yet, it is the caliber of the panelists that determines the quality
of the outcomes of the study (Lang, 1995);
vi. In the process of achieving consensus, extreme points of view run the risk of being
suppressed, when in fact they may provide important new information or
insights (Lang, 1995);
vii. The flexibility of the technique means it can be adapted to a whole range of
situations, which in turn can makes it vulnerable to misrepresentation and sloppy
execution (Amara, 1975); and
viii. Garrod (2008) found that the Delphi technique can be extremely sensitive to the
level of panelists’ expertise, the composition of the panel, the clarity of the
questions, the way the research or coordinator reports reasons for outliers, and
the administration of the questionnaire.

Despite the limitations noted above from different scholars, Brill, Bishop and Walker
(2006) describe the Delphi as a particularly good research method for developing consensus
amongst a group of entities having expertise on a particular topic where information
required is subjective and where participants are separated by physical distance (Linstone &
Turoff, 1975). Brill et al. (2006) further state that the Delphi method has been validated in
the literature as a reliable empirical method for reaching consensus in a number of areas.
Amongst these areas are distance education (Thach & Murphy, 1995), journalism (Smith,
1997), visual literacy (Brill et al., 2006), electronic commerce (Addison, 2003), health care
(Whitman, 1990) and others. Beside these areas, the method has also been used in many
other disciplines, such as in information technology (IT) research to identify and rank key
issues for management attention (Delbecq, Van de Ven &Gustafson, 1975); scientific study
of GIS (Hatzichristos & Giaoutzi, 2005), quality management (Saizarbitoria, 2006),
terrorism (Parente, Hiob, Silver, Jenkins, Poe & Mullins, 2005), banking (Beales, 2005),
151
social sciences (Landeta, 2006), the privatization of utilities (Critcher & Gladstone, 1998),
and education (Yousuf, 2007). Based on the extensive usage of the method over time, the
Delphi method in a research is an accepted practice. However, it may not be appropriate for
all research activities.

8.4.3.1 Epistemological Approach towards the Delphi Design

The variance amongst the various group techniques, the definition of the Delphi method as
compiled by various scholars and cognisance of the various criticisms form the
epistemological foundation for defining the approach towards a typical Delphi study design.
Amongst these include reducing the effects of personal bias. This is done by assuring that
all expert feedback is anonymous. Through this, the technique captures the opinions,
experience, and knowledge of each panel member. Personal knowledge is harvested and
interpersonal interaction biases are stripped away. According to Scheele (2002), “…the
concreteness of the framework of the Delphi design is vital in researching the overall
objective of the study”. The basic premises of the Delphi research design towards a typical
TQM (TQM) study is entrenched in some form of general agreement and consensus
regarding the core ingredients and components of the subsequent framework.

Given the current status of TQM issues in Ghana and the absence of generally agreed TQM
implementation principles, the search for consensus and a point of departure in issues on
TQM that will better serve the construction industry is therefore justified through the use of
the technique. Hence the objective of the Delphi design for this study is to obtain the most
reliable consensus of opinion of a group of experts in the specific field being studied,
namely TQM in the construction industry. According to Lang (1995), “…the Delphi
technique is mostly used to solicit the opinions of experts to determine the timing and
possible occurrence of future events”. It is a method that is best used where there is little
past data available applicable to extrapolate from, and where social, economic, ethical and
moral considerations are pre-eminent. Considering the outcome of the literature review of
the current research, there is no structured research so far carried out which has adopted the
technique with regard to TQM implementation in the Ghanaian construction indusrty. It is

152
therefore justified that the Delphi technique is the best method to explore the subject of the
research and to achieve the aim and objectives.

8.4.3.2 When to use Delphi Technique

The Delphi method is mainly used when long-term issues have to be assessed such as the
subject of the current research. This is because it is a procedure used to identify statements
(topics) that are relevant for the future; it reduces the tacit and complex knowledge to a
single statement and makes it possible to judge upon (Cuhls, 2003). Hence the use in
combination with other methodologies like survey design can be interesting. It is also
suitable if there is the (political) attempt to involve many people in processes (Eto, 2003).
Hence, Linstone and Turoff (2002) argued that one or more of the following properties
could lead to the need for the use of the Delphi technique:
i. When the problem of inquiry does not lend itself to precise analytical
techniques but can benefit from subjective judgments on a collective basis
(Buckley, 1994);
ii. The research needs to contribute to the examination of a broad or complex
problem with no history of adequate communication and may represent
diverse backgrounds with respect to experience or expertise, which is a
major premise of the research;
iii. More individuals are needed than can effectively interact in a face-to-face
exchange;
iv. Time and cost to make frequent group meetings is limited;
v. The efficiency of face-to-face meetings can be increased by a supplemental
group communication process;
vi. Disagreements among individuals are so severe or politically unpalatable
that the communication process must be refereed and anonymity assured; or
vii. The heterogeneity of the participants must be preserved to assure the validity
of the results, such as the avoidance of domination by quantity or by strength
of personality, called the “bandwagon effect”.

153
According to Cuhls (2003), “…the Delphi method as a foresight tool seems to possess
certain degrees of invariance to survive in the changing challenges of the past 50 years”.
Hence the process could serve different understandings of predicting or premonition and is
probably understood by the users as being relevant for covering technical perspectives
organisational perspectives, but also personal perspectives. Cuhls (2003) further emphasises
that what the users of the Delphi technique especially like are the sets of data about the
future that are collected. Writing down future topics seems to have an immense
psychological effect because it transfers implicit tacit knowledge to the more visible and
explicit, and therefore transferable knowledge.

8.4.3.3 Components of the Delphi Technique

The main components of the Delphi technique, according to Loo (2002), consist of the
following five major characteristics, which were adopted in the study.
i) A study should consist of a panel of carefully selected experts representing a broad
spectrum of opinions on the topic or issue being examined;
ii) The participants are usually anonymous;
iii) The coordinator (researcher) constructs a structured questionnaires and feedback reports
for the panel over the course of the Delphi;
iv) It is an iterative process often involving three to four iterations called ‘rounds’ of
questionnaires and feedback reports; and
v) There is an output, typically in form of a research report containing the Delphi results,
the forecasts, policy and programme options (with their strengths and weaknesses),
recommendations to senior management and possibly an action plan for developing and
implementing the policies programmes.

Likewise, Hasson, Keeney and McKenna (2000) recommended that the following research
guidelines for using the Delphi technique be addressed in designing a Delphi approach,
namely research problem identification, understanding the process;, selection of experts,
informing or invitation to experts, data analysis, and presentation and interpretation.

154
i. Research problem identification: Turoff (1970) outlined four objectives that call for the
use of the Delphi technique. One of those objectives was to relate informed judgments on a
topic that spans a wide range of disciplines. Reid (1988) contended that the decision to use
the Delphi technique must centre upon the appropriateness of the available alternatives.
Reid (1988) claims that the use of experts in a field of study is a perfectly suited technique
if the technique has not been utilized in the past, based upon the research performed, such
as the current study that has not employed the Delphi Technique as a tool of investigation in
TQM in Ghana.

ii. Understanding the process: The Delphi technique is a multistage process designed to
combine opinions into group consensus (McKenna, 1994). The process consists of the
following:
• Pilot testing of a small group;
• Initial questionnaire - qualitative comments solicited (not in all cases);
• Initial feedback - quantitative after statistical analysis of the initial opinions;
• Subsequent questionnaire-qualitative comments solicited again; and
• Subsequent feedback - quantitative after statistical analysis. This provides an
opportunity for participants to change their opinions.
iii. Selection of experts: It is important to select panel members who are impartial, and are
interested in the topic. Some studies have over 60 experts, some as few as fifteen. Selection
of people knowledgeable in the field, and their commitment to multiple rounds of questions
on the same topic are essential. In the section that provides more details regarding the
practical design and execution of the Delphi study for this thesis, further details on how the
experts were chosen for the study arepresented.
iv.. Informing or invitation to experts: It is imperative to explain what is required of them,
how much time it will require, what they will be required to provide, what the objective of
the study is, and what will be done with the information.
v. Data analysis: This is the process where opinions of the experts are solicited. According
to Green, Jones, Hughes and Williams, 1999) two or three rounds are preferred. Green et al.
(1999) suggest that an 80 per cent consensus should be the goal.

155
Likewise, Crisp, Pelletier, Duffield, Adams and Nagy (1997) suggest that percentages
should not be used, but rather the process should stop when stability of the data occurs.
However, percentage estimation was found suitable for this study as one of the means to
achieve consensus, hence, a 60 per cent consensus goal was set for the three round Delphi
studies. Also, analytical software can be utilised to analyse the responses and provide
feedback to the experts on the central tendencies (median and interquartile range) and on
the levels of dispersion (standard deviation). Hence, Lincola and Guba (1985) state that the
criteria for qualitative studies such as the Delphi technique should be credibility
(truthfulness), fittingness (applicability), audit ability (consistency), and confirmatory
ability.
vi. Presentation and interpretation: There are a number of methods for presenting the data
from a typical Delphi study, with two methods being graphical and statistical. These two
methods have been used in the current research.
Therefore, given the nature of the current research, it is further believed that the Delphi
technique is well-suited to obtain credible inputs from experts in industry, academia and
government to serve as key input in the development of a TQM model for Ghanaian
construction industry. The next section provides an overview of how the Delphi technique
is used in this thesis.

8.4.3.4 Designing, Constructing and Executing the Delphi Technique Study

Given the rationale behind the Delphi technique and the main features previously explained,
the design, construction and execution of the Delphi study for the current research follows a
sequential process as suggested by Loo (2002). Four vital planning and execution activities
are followed as pointed out by (Loo, 2002):
i. Problem definition;
ii. Panel selection;
iii. Determining the panel size; and
iv. Conduction the Delphi iterations.
Supporting Loo (2002) approach, Delbecq, Van de Ven and Gustafson (1975) suggest a
basic Delphi methodology that includes distinct stages such as Delphi question
development (objective), expert panel selection, sample size, first questionnaire, first

156
questionnaire analysis and follow-up questionnaires. This methodology forms the basis of
the current Delphi research study and is explained in the subsequent sections. Table 8.2
gives a summary of the Delphi design, construction and execution.

Phase 1: Delphi Questionnaire Development

The formulation of the Delphi question is vital to the whole process. It is paramount that the
panel of experts should understand the broad context within which the questionnaire is
designed. The current research looked at the concept of what determines TQM
implementation in the construction industry. Hence the concept had to be broadly clarified.
Key questions were asked to achieve the objectives of the Delphi study. The basis of
constructing the questions for this current study was the guidelines given in Table 8.2, with
corresponding wording and phrasing given for this study.

Table 8.2: Key Delphi Questions and Phrasing for the Study

Key Delphi questions? Phrasing for this study

Why are you interested This study was initiated because of the belief that not all
in this study? construction firms in Ghana implement TQM.
This assumption is concrete because there is lack of
understanding of the diverse features that determine TQM
implementation in construction industry.
What do you need to Despite the knowledge about the features that brings about TQM
know that you do not implantation in construction industry, Carry on here
know now? these have not been put together as a model for the construction
industry in Ghana. The attribute that determines TQM in the
Ghanaian construction industry emerges the end of this study.
How will the results The results of the Delphi study enable the development of a
from the Delphi study conceptual framework for the TQM model to be developed. The
influence TQM in attributes which collectively predict and assure TQM
construction industry? implementation in the Ghanaian construction industry are
established.

157
Phase 2 – Delphi Expert Panel Selection

A critical phase in conducting a Delphi interview technique is the selection of the right
experts (panellists, participants or respondents). The experts’ role is also crucial to the
success of the research (Hasson, Keeney & McKenna, 2000). To ensure a high commitment
response rate in the subject under examination depends on the selection of the experts, their
interests and involvement. According to Hasson et al. (2000), “…controversial debate
occurs when a professional becomes an expert”. The claim that a group represents valid
expert opinion has been criticized as scientifically untenable and overstated. McKenna
(1994) defined ‘an expert’ as a panel of informed individuals (otherwise called experts
hereafter). The definition of ‘an expert’ by McKenna (1994) was supported by Goodman
(1987:730) who stated that the Delphi technique “…tends not to advocate a random sample
of panellists… instead the use of experts or at least of informed advocates is
recommended”.

Delphi inquiry is not an opinion poll, relying on drawing a random sample from the
population of experts is not the best approach, rather, once a set of experts has been
selected (regardless of how-but, following a predetermined qualifying criteria), it
provides a communicative device for them that uses the conductor of the exercise as a
filter in order to preserve anonymity of responses’, which is the core of the Delphi
Technique (Helmer, 1977:18-19).

Therefore, Linstone and Turoff (2002) asserted that the most significant danger in selecting
the panel of experts lies in the path of ‘least resistance’ through the selection of a group of
cosy friends and or like-minded individuals, which thus negates the strength of the process.
Since panellists form the cornerstone of the Delphi technique, clear inclusion criteria should
be applied and outlined as a means of evaluating the results and establishing the study’s
potential relevance to other settings and populations (Igbal & Pipon-Young, 2009).

The selection of panellists for the study was based on criterion sampling. Panellists were
selected for a purpose to apply their knowledge to a concept raised in the study based on the

158
criteria that were developed from the research questions under investigation. A Delphi
study does not depend on a statistical sample that attempts to be representative of any
population. It is a group decision mechanism requiring qualified experts who have deep
understanding of the issues (Okoli & Pawlowski, 2004). Therefore, one of the most critical
requirements is the selection of qualified experts as it is the most important step in the
entire Delphi process because it directly relates to the quality of the results generated (Hsu
& Sandford, 2007). Stitt-Gohdes and Crews (2004:61) argued that careful selection of the
panel of experts is the keystone to a successful Delphi study. Dalkey and Helmer (1963)
opined that there are detailed criteria for the selection of panel experts, recommending that
in a typical Delphi study experts should meet two recommendations. These
recommendations were also postulated by Rodgers and Lopez (2002). The two
recommendations are as follows:
i. The experts should exhibit a high degree of knowledge of the subject matter. and
ii. They should be representatives of the profession so that their suggestions may be
adaptable or transferable to the population.

Adler and Ziglio (1996) stated that the Delphi participants in any study should meet four
‘expertise’ requirements, namely knowledge and experience with the issues under
investigation, capacity and willingness to participate, sufficient time to participate in the
Delphi studies, and effective communication skills.
In choosing panellists for this study, each expert was required to meet at least five (5) of the
following minimum criteria:
i. Residency: Have lived in any of the Metropolitan/Municipal/Districts in Ghana for at
least more than one (1) year;
ii. Knowledge: Has knowledge of TQM in the construction industry;
iii. Academic Qualification: Has been presented an earned degree (National
Diploma/Bachelor’s degree/Master’s degree/Doctor of Philosophy) related to field of
study, certification of employment/experience focusing on construction
development/project management issues;
iv. Experience: Has a history of or is currently performing consultation services for the
government of Ghana, individuals, businesses, agencies, companies, and/or
159
 organizations relating to construction project management. The experts must exhibit a
high degree of knowledge of experience in the subject matter as well as extensive
theoretical knowledge.
v. Employment: Currently serves (or has previously served) in a professional or voluntary
capacity (e.g. at place of employment - institution, business, agency, department,
company) as supervisor or manager of establishment that is involved with construction
project management construction in Ghana.
vi. Influence and Recognition: Has served/is currently serving as a peer reviewer for one or
more manuscripts received from a journal editor prior to its publication in the primary
literature, with focus of the manuscript(s) on construction project management.
vii. Authorship: Is an author or co-author of peer-reviewed publications in the field of
construction with emphasis on Ghana, has prepared and presented papers at
conferences, workshops or professional meetings focusing on construction project
management and TQM.
viii. Research: Has submitted one or more proposals to or has received research funds
(grant/contract) from national, local government, regional, and/or private sources that
support construction project management and studies related to TQM.
ix. Teaching: Has organised, prepared, and successfully presented one or more TQM
training workshops focusing on the group for which expertise is sought. The workshop
or course must have been on TQM practices. Or, has served as an individual or as a
collaborative instructor in the teaching of one or more polytechnic or university courses
focusing on TQM or a related field.
x. Membership: Member of a professional body (as listed on the expert questionnaire).
Should be the representative of a professional body so that their opinions may be
adaptable or transferable to the population.
xi. Willingness: Must be willing to participate fully in the entire Delphi process.

160
 Step 1:

Literature Review of Total Quality Management

Step 2:

Total Quality Management expert Selection

Step 3:

Development of 1st Round Questionnaire

Step 4:

Questionnaire Quality Check

Step 5:

Completion of 1st Round

Questionnaire
Step 6:

Analysis of 1st Round Questionnaire

Step 7:

Reiterate Steps 7-9 until consensus is reached

Development of 2nd Round Questionnaire
Step 8:

Completion of 2nd Round

Questionnaire
Step 9:

Analysis of 2nd Round Questionnaire

Step 10:
No

Has Consensus been reached?

Step 11:

Produce Report

Figure 8.4: The Delphi process

161
The selected participants should represent a wide variety of backgrounds to guarantee a
wide base of knowledge (Rowe, Wright & Bolger, 1991). The recommendations of Rowe et
al. (1991) were adopted for the current study. The number of respondents should be large
enough to ensure that all perspectives are represented but not so large as to make the
analysis of the results unmanageable by the researcher (Linstone & Turoff, 1975). The
adoption of five of these criteria was considered more stringent than the recommended
number of at least two criteria by Rogers and Lopez (2002) and Dalkey and Helmer (1963).
The five minimum criteria were framed after the four recommendations made by Adler and
Ziglio (1996), with the inclusion of experts’ residency status, which was considered to be
compulsory for all selected experts. This was considered significant because experts were
required to have a wide-ranging understanding of TQM practices within their locality.

Phase 3 - Determining the Panel Size

The Delphi technique requires a qualitative approach rather than a quantitative approach.
The number of participants or experts in a Delphi technique is expected to be fewer than
normal quantitative surveys. Various scholars have recommended different sample sizes for
determining the minimum number of experts to participate in a typical Delphi survey.
Dalkey and Helmer used a panel of seven experts in their original Delphi experiment in
1953 (Helmer, 1983). According to Linstone (1978:296), “…a suitable minimum panel size
is seven”. Linstone (1978) justified this by indicating that the research runs the risk of
accuracy and deteriorates quickly when the numbers increase. The justification given by
Linstone (1978) was supported by Cavalli-Sforza and Ortolano (1984:325) who postulated
that a “…typical Delphi panel has about eight to twelve members”, while Phillips
(2000:193) stated that the optimum number of participation should be between seven and
twelve. A similar number was cited by Linstone (1978). Miller (1993) was of the view that
any additional responses beyond the first thirty responses would not generate new
information. Similarly, Dunn (1994) suggested ten to thirty participants, indicating that as
the complexity of the policy issue increases, the sample size needs to be larger to include
the entire range of participants both for and against the policy issue area. Dunn (1994)
further emphasised the inclusion of both formal and informal stakeholders who have an

162
interest in the policy issue in the study. They should also have varying degrees of influence,
hold a variety of positions, and be affiliated to different groups.

These requirements are the basis upon which the present Delphi study was based.
According to Andranovich (1995), “…if the group of experts is fairly homogeneous
(sharing similar opinions) then ten to fifteen panellists will be enough and if there are
diverse interests present amongst the experts, then the size of the group will need to be
increased to ensure balance (Zami & Lee, 2009). Most community-oriented Delphi studies
have a panel of thirty experts. Since the Delphi technique is a labour-intensive procedure,
the greater the number of panellists, the greater the information load (Zami & Lee, 2009).

The following are the factors highlighted by Skulmoski, Hartman and Krahn (2007) to be
considered in order to determine the sample size for a Delphi technique:
i. Heterogeneous or homogeneous sample: Where the group is homogeneous, then a smaller
sample of between ten to fifteen people should yield sufficient results.
Nevertheless, if an unrelated group is involved, for instance in an international study, then a
larger sample will most likely be required and several hundred people might participate
(Delbeq et al., 1975). However, the researcher needs to exercise caution because
heterogeneous groups can greatly increase the complexity and difficulty of collecting data,
reaching consensus, conducting analysis, and verifying results.
ii. Decision quality/Delphi manageability trade off: There is a reduction in group error (or
an increase in decision quality) as sample size increases (Linstone & Turoff, 2002).
However, above a certain threshold, managing the Delphi process and analysing the data
becomes cumbersome in return for marginal benefits.
iii. Internal or external verification: The larger the group, the more credibly the results can
be said to be verified. However, a smaller sample might be used with result verification
conducted through follow-up research. The current research adopted a smaller sample
premise and verified through a follow-up questionnaire survey.

However, the selection of an initial respondent panel for the Delphi study varies. It was
concluded from the literature review that a typical sample size varies between seven to fifty
163
panellists. There is no agreement on the desired ‘typical’ number of panellist to be adopted
in a Delphi study. Rather, the method can be modified to suit the circumstances and the
research question. Owing to time constraints and conflicting schedules of the experts, the
current study did not involve a large number of experts. A sample size of a panel of 10
experts was adopted based on the following premise. This was also in conjunction with the
qualifying criteria as established in phase two of the Delphi study:
i. Experts should be fairly and practically split between academics and practitioners. The
two categories may provide input for various perspectives and balance the theoretical and
practical considerations.
ii. Panellists in both categories should have extensive experience relating to TQM and
project management in general.
The number of panel of experts also depends on the topic area, as well as the time and
resources at the researcher’s disposal. The adopted number of panel of experts of nine (9)
seems appropriate, considering the data required and subsequent analyses each panel of
expert will generate.

Phase 4 – Conducting the Delphi Iterations

Data collection through Delphi

Sequences of questionnaire rounds are used to obtain iterative responses to issues in a

Delphi study (Masser & Foley, 1987). According to Woudenberg (1991), “…two to ten
rounds is appropriate because accuracy is expected to increase over rounds and the need for
repetition of judgment and group pressure for conformity”, while Critcher and Gladstone
(1998) suggested between two to five rounds. Two rounds of iterative process were used in
the current study for the Delphi method. This was done to achieve consensus between the
panel members on the influence and impact of TQM in the construction industry. The
attached Delphi questionnaire in Appendix D was sent out electronically to all panel experts
to respond to the questions according to their ability and expertise. The Delphi
questionnaire was developed based on the findings from the literature review and was
specifically designed to address and achieve the Delphi-specific objectives defined for the

164
study. The experts took a minimum of two months to complete the questionnaires. A
questionnaire was designed for each round based on the responses to the previous one.

However, the Round One questionnaire was designed based on a summary of the
comprehensive review of literature highlighting sets of attributes and sub-attributes that are
potentially relevant to TQM in the construction industry in Ghana (see Appendix D).. These
attributes were structurally and constructively put together to frame the first round of the
Delphi survey. Therefore, Round One of the Delphi study was intended to be a
brainstorming exercise used to produce a list of empirical attributes that determine TQM
implementation in construction industry. Other issues relating to TQM implementation in
Ghana and other subsequent issues relating to the objective of the Delphi study were also
considered in Round One. Closed-ended questions were used in this round (round one). The
responses were analysed and formed the basis of Round Two (see Appendix E) of the
study. Frequencies were obtained to measure the degree of consensus reached amongst
participants regarding the attributes that determine TQM in the Ghanaian construction
industry and for other related questions. The purpose of the second round of the study was
to allow experts to review and comment on the attributes that determine TQM and other
related issues relating to TQM implementation in Ghana, which were proposed by the panel
of experts in Round One. Closed-ended questions were used in this round to investigate
participant comments expressing agreement, disagreement or clarification concerning
proposed attributes that determine TQM implementation in Ghana. The specific nature of
the closed-ended questions stimulated participants’ reactions. Frequencies were likewise
obtained to measure the degree of consensus reached amongst participants regarding the
attributes that determine TQM implementation and for other related questions.

The final Round Two (Appendix E) was specifically aimed at:

i. Informing the experts of the findings of the analysis of responses to the questionnaire of
Round One; and
ii. Requesting their final affirmation or comments on attributes and issues that did not
receive any consensus in Round One.

165
The Round Two questionnaire was designed based on the measures of frequency responses
to the questionnaire of Round One. Frequencies were obtained to indicate consensus
reached amongst experts regarding attributes that determine TQM and TQM issues as
presented in the study. Consensus was reached regarding most of the attributes that
determine TQM in the Ghanaian construction industry over the two rounds of the Delphi
survey.

Based on the findings of the analyses of responses to the Delphi rounds, a list of attributes
that determine TQM was prepared, which informs the conceptual framework for the
broader study, while issues surrounding TQM implementation were highlighted. The
Delphi survey was conducted via electronic mail, and follow-up emails were used to
encourage prompt responses to the questionnaires. Using email provides a free and faster
means of communication.

With regard to the Delphi questionnaire, a panel of experts were requested to rate the
likelihood of an attribute influencing TQM, and the impact of sub-factors in predicting
TQM in the Ghanaian construction industry, if they were present. The probability scale
ranged from one to ten representing zero to 100%. Interval ranges were set at ten (Table
8.7). The panel of experts were asked to rate the negative impact that would result if a
particular TQM attribute were also absent. This was based on a 10-point ordinal scale
ranging from ‘negligible’ to ‘very high impact’. This aspect indicated the importance of the
TQM as shown on Table 8.8.

Table 8.3: Influence or Likelihood Scale

0-10 % 11-20 % 21-30 % 31-40 % 41-50 % 51-60 61-70 % 71-80 % 81-90 % 91-100
% %
1 2 3 4 5 6 7 8 9 10
x x

Table 8.4: Impact Scale

No impact Low impact Medium impact High impact Very high impact
1 2 3 4 5 6 7 8 9 10
x
166
The panel of experts were also required to state their level of agreement using a 10-point
Likert scale with certain statements and to support their choices where necessary with
regard to TQM policy issues and TQM practices to arrive at a consensus. Group medians
were calculated as a measure of the central tendency for each response on each element.
Thus, an indicator of whether consensus had been reached on the questions for each
element was determined. The median was deemed to be more suitable for the type of
information that was being collected. This is because the median eliminates bias, takes into
consideration outlier responses, and makes a consensus notion more reasonable. The mean,
on the other hand, may not reflect a reasonable central tendency as it considers only the
outlier responses.

Group medians and the absolute deviations from the Delphi First Round were computed for
each element. These were then sent back to the expert panel members so that responses in
the second round could be made, taking into account the group median. The panel of
experts in the second round were asked to either maintain their original responses made in
Round One, or to change their initial responses to be more in line with the group median.
Panel members, who had responses to units above or below the group median in the second
round, were requested to state their reasons for sticking to a response that does not agree
with the group median. Group medians and the absolute deviations were again computed
for the second round. From these calculations and after two rounds of the Delphi process, it
was determined that consensus had been reached. Reasons for other experts who stuck to
their ratings were, however, taken into consideration.

Calculations for the second round of the Delphi process indicated that there was no need to
proceed to the third round as there was no further value that could be added to the degree of
consensus attained at that level. Throughout the entire Delphi process, anonymity of panel
members was maintained to avoid undue influence on other members. The aspect of
anonymity is crucial to the credibility of the Delphi technique. There was a rigorous
establishment of the complex ‘what would happen if’ kind of questions that ideally should
be established from an experimental study. This was in fact extremely difficult and not
feasible to do. Figure 8.5 shows an outline of how the Delphi study data was collected.
167
 Researcher actions Expert panel actions

Round 1 Circulates questionnaire to Rates probable housing satisfaction

Q1 expert panellists due to factors a influence / impact and
other issues

Reviews individual rating in view of

Round 2 Compute group medians and the group’s median. Gives reason if
Q2 re-circulate questionnaires required

Round 3 Re-compute group medians, Experts given opportunity to

Q3 standard deviations & reconsider initial rating
compile comments

Determine consensus and

terminate process

Figure 8.5: Outline of Delphi Process

Source: Adapted from Thangaratinam and Redman (2005:124); Musonda (2012: 102);
Aigbavboa (2013: 284)

8.4.3.5 Specific Objectives of the Delphi Study

There are various characteristics and factors which determine TQM in the construction
industry from literature, as measured from different TQM theories and categories such as
TQM practices. It is not clear from the literature what specific level or extent the identified
factors contribute to the TQM implementation in the construction industry. Attempts have
been made through various studies to determine the influence of these factors in TQM
studies but none has been specifically related to TQM implementation. Various factors
determine TQM implementation albeit varied in different cultural and TQM backgrounds
and typological settings. Also, previous models have not been adequately organized into a
model to form a holistic attribute which determines TQM implementation in Ghana. Based

168
on the foregoing, a more reliable measure of the determinants of TQM was therefore
necessary in order to establish not only whether these factors have an influence on TQM
implementation but also the extent or level of their influence. In addition, it was necessary
to identify which factors have the greatest influence in the Ghanaian context together with
the identified gap of factors from the literature. Based on the context of the thesis, this kind
of investigation would ordinarily call for an experimental kind of research. However, the
experimental method of research was not feasible and practical considering the time frame,
ethical issues and the willingness of would-be participants. Hence, the Delphi method was
considered the most suitable method to achieve the general objective of determining the
influence and impact of the identified factors on TQM in the construction industry.

The broader research aim was to develop an integrated TQM framework for the Ghanaian
construction industry. The Delphi method was therefore chosen at the first stage to
formulate the conceptual model. This had to be validated later from responses obtained
from the questionnaire survey analysed using the SEM Software EQS Version 6.2. At the
Delphi stage, factors that were identified from literature that defined and determined TQM
were formulated into questions. Experts were asked to give their rating as being influential
or having an impact on TQM. The output from the Delphi process was a set of attributes
which determine TQM that would be implemented.

The following are the specific objectives for achieving the Delphi survey based on research
objectives 3 and 4:

SOD1. To identify the main and sub-attributes that determine TQM implementation in the
Ghanaian construction industry;

SOD2. To determine the factors that enables construction companies to implement TQM;

SOD3. To identify the factors that affect construction companies for not implementing
TQM in Ghana;

SOD4. To evaluate organizational factors that affect the implementation of TQM; and

SOD5. To identify the effects of implementation of TQM in the construction industry in

Ghana
169
The philosophy behind the above objectives is to do away with the tendency of a non-
coherent dialogue on TQM in Ghana. The following are the outcomes in achieving the
above objectives:
i. Determining the key factors and constructs that are of critical significance (influence) to
determine TQM in the construction industry; and
ii. Developing a holistic conceptual model on TQM for the Ghanaian construction industry.

8.4.3.6 Computation of Data from Delphi Study

Computation of data from the Delphi study was conducted using Microsoft Office Excel, a
spreadsheet software program. The first stage involved analysis to determine consensus on
responses to the predetermined criteria. This involved determining the group median
responses for each question. After the second round of the Delphi, absolute deviations (Di)
of the group medians (m(X)) of each rating for the relevant questions as pre-determined
were calculated using Equation 1.0 below.

Di=[x i-m(X)]
Equation 1.0

Where Di =Absolute deviation

x i =Panel rating

m (X) = Measure of central tendency

A computation of each question element was completed for the likelihood and impact of the
attributes in predicting TQM implementation in Ghana. The influence or impact of the
absence or presence of a particular TQM practices element on the overall TQM of the other
elements is presented. For every round of responses from the experts, besides the group
median value computation, their respective interquartile deviation (IQD) were also
computed as a measure of the central tendencies to determine consensus. The median value
was adopted as a measure of central tendency because of its effect to minimize the effects
of potentially biased individuals while the IQD scores were used to summarize the

170
variability in the data. The IQD helped to identify which measures were most appropriate to
influence TQM. A clearer picture of the overall data set was provided as the IQD removes
or ignores outlying values through the use of the IQD. The inter-quartile range is a measure
that indicates the extent to which the central 50% of values within the data set are dispersed.
This is based upon, and related to, the median. The median is adopted as a measure of
central tendency for studies of this nature, as opposed to the mean and IQD, although it is
sometimes used. To compute the variation of the median from the responses for each
question in each round, the absolute deviation given in Equation 1.0 was done. This is the
absolute difference between a response within a data set and a given point. The point from
which the deviation is measured is a measure of central tendency, which is the median. The
results from the Delphi analysis will be presented as numbers and percentages in tables in
the results (Chapter 9, Section 8.3). The outcomes show the predictions of the influence of
TQM factors and other issues in TQM practices in Ghana.

8.4.3.7 Determination of Consensus from the Delphi Process

It is required that consensus should be reached on all questions. Consensus was determined
by measuring the central tendency of the various responses on all questions. The group
median and the IQD were computed for all responses. In order to achieve consensus, the
deviation of all responses about the group median was determined not to be more than one
(1) unit, likewise for the IQD. This is considered to be suitable as the scale that was used
for both probability (influence) and impact was 1 to 10. The deviation of all responses was
calculated using the absolute median (Equation 1.0), while the IQD was calculated based on
the recommended statistical process of the absolute value of the difference between the 75th
and 25th percentiles. A percentile (or centile) is the value of a variable below which a
certain per cent of observations fall. The percentile is often used in the reporting of scores
from norm-referenced tests, as in the present situation. The 25th percentile is also known as
the first quartile (Q1), the 50th percentile as the median or second quartile (Q2), and the 75th
percentile as the third quartile (Q3).

Hence, the deviation between the 75th and 25th percentiles give an absolute value referred to
as the interquartile range or deviation. The interquartile range deviation is a robust statistic,

171
having a breakdown point of 25%, and is thus often preferred to the total range, with
smaller values indicating higher degrees of agreement (consensus). Smaller values in the
inter-quartile range would then indicate higher degrees of consensus. However, consensus
is difficult to measure in Delphi studies. The foregoing has been established from literature,
namely that there is no consensus on how to determine consensus regarding a set of
opinions. Holey, Feeley, Dixon and Whittaker (2007:2) suggested that consensus is the
same as agreement and that agreement can be determined by the following:
i. The aggregate of judgments;
ii. A move to a subjective level of central tendency; or
iii. By confirming stability in responses with the consistency of answers between successive
rounds of the study.
Other researchers have used frequency distribution to assess agreement and the criterion of
at least 51% responding to any given response category being used to determine consensus
(McKenna, 1994). Other studies, such as one conducted by Rayens and Hahn (2000), have
used means and standard deviations with a decrease in standard deviations between rounds
indicating an increase in agreement. Inter-quartile deviation (IQD) has also been used to
determine consensus (Rayens & Hahn, 2000), which has also been adopted for the current
study. Studies conducted by Rayens and Hahn (2000) have included another criterion to
determine consensus in addition to the IQD to achieve stability. The criterion to achieve
consensus was that the IQD should equal one (1) unit for which more than 60% of
respondents should have answered either generally positive or generally negative. Items
which had an IQD ≠1 for which the percentage of generally positive or generally negative
responses was between 40% and 60% were determined to indicate a lack of consensus or
agreement. An IQD of 1.00 or less was identified by Raskin (1994) as an indicator of
consensus. Spinelli (1983) also considered a change of more than 1.00 IQD point in each
successive stage as the criterion for convergence of opinion. How to use or interpret IQD as
a method of data analysis for the Delphi process to achieve consensus has not been
identified in literature review. The potential range of IQD values depends on the number of
response choices, with larger IQDs expected as the number of response choices increases.
The use of a particular IQD as a cut-off for consensus requires consideration of the number

172
of response choices. The following criteria for determining consensus have been used by
Holey et al., (2007):
i. Percentage response;
ii. Percentages for each level of agreement for each question to compensate for varying
response rates;
iii. Computation of median, standard deviation and their associated group rankings;
iv. Computation of the means, standard deviation and their associated group rankings using
the importance ratings; and
v. Computation of the weighted kappa (k) values to compare the chance eliminated
agreement between rounds.

Holey, Feeley, Di and Whittaker (2007) opined that consensus is reached when the
following are present:
i. An increase in percentage agreements;
ii. Convergence of importance rakings;
iii. Increase in kappa values;
iv. A decrease in comments as rounds progressed;
v. A smaller range of responses; and
vi. Smaller values of standard deviations.
These studies suggest that there is little agreement on how to measure consensus in a Delphi
study. It is however agreed that for consensus to have been achieved, there has to be a
convergence of ideas and reasoning towards a subjective central tendency measure. Hence,
in the current study, consensus was determined to have been reached if the following were
achieved:
i. More than 60% of responses are generally positive or negative with certain questions;
ii. The average of the absolute deviation was not more than one unit. The absolute deviation
is calculated from Equation 1.0., and
iii. The IQD was less than 1.00, meaning that items with IQD = 0.00 were considered to
have reflected high consensus.
Therefore the scales of consensus adapted for this research are as following:

173
i. Strong consensus - median 9-10, mean 8-10, interquartile deviation (IQD) ≤1 and ≥80%
(8-10);
ii. Good consensus - median 7-8.99, mean 6-7.99, IQD≥1.1≤2 and ≥60%≤79% (6-7.99);
and
iii. Weak consensus - median ≤ 6.99, mean ≤5.99 and IQD≥2.1≤3 and ≤ 59% (5.99).

8.4.3.8 Reliability and Validity of the Delphi Process

Reliability is the extent to which a procedure produces similar results under constant
conditions at all times (Els & Delarey, 2006:52). This kind of statistical reliability is not
possible in a Delphi study because another panel may reach a different conclusion
depending on their knowledge of the subject area and interest. Care was taken that
credibility showed in truthfulness, fittingness exhibited in applicability, audit ability shown
in response consistency and conformability was exhibited in the responses from all
participants to reach reliability. Credibility was also assured during the selection of the
panel. All panels of experts had distinguished themselves based on the set criteria for the
selection of panel of experts and the depth of their knowledge and experience as presented
in Section 8.4.3.4 (Phase 2: Delphi expert panel selection). Validity was boosted by the
removal of preconception or influence from other members by keeping all members
completely anonymous from each other and hence eliminating the ‘bandwagon’ effect,
which is one of the strengths of the Delphi method. Furthermore, the number of iterations
that were implemented in the Delphi study also enhanced the internal validity. Thus, expert
panellists were given a chance to change their opinion or maintain it with a written
explanation or argument for dissenting views.

Feedback to the researcher and constant communication between the researcher and the
panellists individually was another way of ensuring internal validity of the study. The
external validity of a study deals with the extent the results from the study can be
generalised to a larger population. This is usually determined by how participants are
selected to be part of the study. However, this process was not necessary as the validation
process of the conceptual model had been done using the questionnaire survey. The
selection of participants for the Delphi study guaranteed external validity as scientific

174
criteria as predetermined (Section 8.2.3) based on previous scholarly works were adopted.
The panel comprised members from varied sectors, all with in-depth knowledge on project
management. All the panel of experts reside in one of the major cities in Ghana. They were
highly experienced with sound working experience. The study therefore fulfilled the
requirements for external validity in line with standard research ethics.

8.4.4 Stage Three (Questionnaire Survey)

Phase Three of the research involved collecting data from the field through the use of
questionnaires in order to meet the general objective RO6 of the overall research
respectively. Phase Three formed the pinnacle of the research. The Delphi study findings
were that TQM is a multidimensional construct, which consists of the Leadership and Top
Management, Company Supplier Quality Management, Company Quality System
Evaluation, Client Focus and Involvement, Company Vision and Plan Statement, Product
Selection and Design Management, Construction Process Management and Improvement,
and Construction Employees’ Involvement and Motivation. These factors have been
collectively considered for the development of a holistic integrated TQM model in this
study. Five of the factors have been previously considered in the development of a TQM
(TQM) model in other cultural contexts. None of the existing models to date have included
Company Vision and Plan Statement, Product Selection and Design Management, and
Company Quality System Evaluation as factors to develop a model that will help
construction companies in Ghana produce quality project. The following specific objectives
of the questionnaire survey were used to validate the findings from Delphi study:
QS1. Identify the factors that are more influential regarding TQM implementation in the
Ghanaian construction industry;

QS2. Establish the effects of implementation of TQM in the construction industry;

QS3. Identify the factors that are more influential and affecting construction companies’
for not implementing TQM in the Ghanaian construction industry;

QS4. Determine the factors that are more influential to enable construction companies to
implement TQM;

175
QS5. Determine the goodness-of-fit of the hypothesized integrated TQM model to the data
sampled.

The previous models of TQM established in the developed countries cannot be relied on in
developing countries. The findings of what determines TQM in the construction in
developing countries are rarely known from the previously conducted research. There is a
lack of research into the overall impact on and influence of the direct and holistic active
involvement of TQM constructs. Owing to the absence of an integrated TQM model, the
achievement of TQM implementation in the Ghanaian construction industry is unlikely.
The integrated conceptual model shown in Figure 10.1 (Model 1.0; page 257) was made up
of the following interrelationships:
1. Leadership and Top Management features have an impact on TQM in the
construction industry and greatly influence its determination;
2. Company Supplier Quality Management features have an impact on TQM in the
construction industry and greatly influence its determination;
3. Company Quality System Evaluation features have an impact on TQM in the
construction industry and greatly influence its determination;
4. Client Focus and Involvement have an impact on TQM in the construction industry
and greatly influence its determination;
5. Company Vision and Plan Statement have an impact on TQM in the construction
industry and greatly influence its determination;
6. Product Selection and Design Management features have an impact of TQM in the
construction industry and greatly influence its determination;
7. Construction Process Management and Improvement features have an impact of
TQM in the construction industry and greatly influence its determination;
8. Construction Employees’ Involvement and Motivation features have an impact of
TQM in the construction industry and greatly influence its determination; and
9. The integrated holistic TQM model describes the determinant (constructs) which
determine TQM implementation in the Ghanaian construction industry.

176
As a result of the objectives of the study, it was obvious that collecting facts would involve
professionals and top management employees in the building construction industry. A field
survey was considered the most suitable method of collecting the required data. A complete
biographical section that related to the building construction industry was required. Data
relating to Leadership and Top Management, Company Supplier Quality Management,
Company Quality System Evaluation, Client Focus and Involvement, Company Vision and
Plan Statement, Product Selection and Design Management, Construction Process
Management and Improvement, and Construction Employees Involvement and Motivation
were required. Other data required were factors that have an influence on TQM and factors
that have an influence on TQM implementation in the Ghanaian construction industry.
These types of information could not be obtained through other means of data collection
except a field survey. This is because other means of data collection would not give an
adequate representation of the above stated relationships. Also, the decision to choose a
survey method was based on a number of factors. The factors included the sampling
technique to be adopted, the type of population, the question form, the question content, the
response rate, and the duration of data collection. The most appropriate survey method for
this research was a personally administered questionnaire.
This method was chosen for the following reasons:
i. The questions could be answered by crossing the proper response. Respondents could
seek clarity on any question in order to meet consistent question objectives (Sekaran, 2000;
Aaker, Kumar & George, 2009);
ii. A higher response rate of almost 100% can be assured since the questionnaires were not
left with the respondents but collected once they had been completed (Malhotra, 1999;
Sekaran, 2000);
iii. Higher levels of anonymity of respondents were assured because respondents were not
required to disclose their identities (Sekaran, 2000; Burns & Bush 2002).

Apart from the above reasons for adopting a personally administered questionnaire survey
method, the following reasons justify the use of the questionnaire survey method:

177
 1. The philosophy underpinning the research is based on the positivist theory as
discussed above, which uses quantitative methods and collection of data by the use
of questionnaires;
2. Validation of the conceptual model developed at Phase One (literature review) and
Two (Delphi) entailed using an alternative method to the previous two to validate
the findings. This therefore eliminated the use of methods similar to the Delphi and
its derivatives. Hence, this called for collection of data on the TQM in the
construction industry through a questionnaire survey;
2. The field survey was considered to be more descriptive in that most of construction
firms are located in the regional capitals of Ghana;
3. Likewise, the interpretation and presentation of the data can easily be done and
understood by various readers when adopting a positivist philosophy of research as
it follows a logical explanation of the method;
4. A large number of research questions can be asked in a questionnaire to target many
respondents within a stipulated time frame;
5. A questionnaire requires minimal investment and is relatively easy to obtain
generalizations from (Bell, 1996);
6. Specific information about views, attitudes and perceptions of a group of
respondents, which are difficult to measure using observational technique, can be
easily elicited via a questionnaire (McIntyre, 1999; Yuen, 2007);
7. Another reason for choosing a questionnaire was because data collected through a
questionnaire can be analysed easily; and
8. Data entry and analysis for the questionnaire can be easily done using computer
software packages, such as the SPSS and EQS (Bell, 1996; Hishamuddin, 2007;
Yuen, 2007).

8.4.4.1 Questionnaire Survey Instrument

Pinsonneault and Kraemer (1993) defined a questionnaire as a method used to gather
information related to the opinions of a large group of people. A standardised questionnaire
exposes each respondent to same set of questions (Brace, 2008). This study applies a formal
standardised questionnaire in order to achieve the research objectives. A structured

178
questionnaire was used to collect data during the field survey. The questionnaire was based
on the literature review conducted in the first stage of the research, as well as the findings
from the Delphi study in stage two. The questionnaire consisted of six sections. The first
section was designed to collect information about the profiles of the respondents and firms.
This information included biographical, socio-economic and other information, as deemed
necessary to meet the research objectives.

Section Two of the questionnaire included questions on the TQM in the construction
industry as related to Leadership and Top Management, Company Supplier Quality
Management, Company Quality System Evaluation, Client Focus and Involvement,
Company Vision and Plan Statement, Product Selection and Design Management,
Construction Process Management and Improvement, and Construction Employees’
Involvement and Motivation. This section was meant to collect information on the extent of
the TQM for each sub-attribute variable as provided. The questionnaire was designed to
assess the influence of the identified constructs on TQM. The first section had fourteen
questions relating to the profiles of the respondents and firms. Section Two had nine sub-
headings with different numbers of questions in each section. The respondents were asked
to rate each of the items on a five-point Likert scale regarding the extent to which they
agreed or disagreed with factors that determine TQM in the construction industry. The
length of the questionnaire was ten pages, including the cover letter (see Appendix F). This
was in line with the recommended length of between five (5) to 12 pages.

To avoid bias resulting from the questionnaire design, the questions were constructed in
such a way that they were direct, short, comprehensible, avoided ambiguity, not vague, no
generalizations, not leading, not double-barrelled, presumptuous questions, but simple and
familiar to the respondents. Instructions for the questionnaire were kept simple with no
technical or specialized words being used. However, it was recognised that this type of
questionnaire has a few weaknesses as follows:
i. There is an absence of probing beyond the answer given;
ii. There is a lack of control over who answers the questionnaire; and
iii. They can be characterized by a low response rate because of cost.
179
The above weaknesses were addressed by refining the questions and keeping them simple
but care was taken not to deviate from the objectives of the instrument, keeping the overall
questionnaire within the recommended limits and ensuring that only the right person(s)
completed the questionnaire by the lead researcher constantly communicating with the
fieldworkers. The absence of further probing is characteristic of this type of questionnaire.
This aspect was not a major concern as the data to be collected was meant to validate the
integrated conceptual model initially developed in the previous phase of the study.

One type of response format was used, namely close-ended. To obtain the respondents’
extent of agreement and disagreement towards the model’s identified constructs; a labelled
scale response format was used. Apart from the simplicity of administering and coding in
further statistical analysis (Burns & Bush, 2000), a labelled scale response format is
appropriate for TQM, as it allows the respondent to respond to questions to varying degrees
that describe the dimensions being studied (Aaker, Kumar & George, 2009). Labelled
Likert scales were appropriate to measure the responses for this study. This scale was
adopted based on the following reasons:
i. It yields higher reliability coefficients with fewer items than the scales developed using
other methods (Hayes, 1998);
ii. This scale is widely used in TQM research and has been extensively tested in both
marketing and social science studies (Garland, 1991);
iii. It offers a high likelihood of responses that accurately reflect respondent opinion under
study (Zikmund, 2000; Burns & Bush, 2002); and
iv. It helps to increase the spread of variance of responses, which in turn provide stronger
measures of association (Aaker et al., 2009; Wong, 1999).

8.4.4.2 Variables
The research instrument was designed to measure the exogenous variables, namely
Leadership and Top Management (LTM), Company Supplier Quality Management
(CSQM), Company Quality System Evaluation (CQSE), Client Focus and Involvement
(CFI), Company Vision and Plan Statement (CVPS), Product Selection and Design
Management (PSDM), Construction Process Management and Improvement (CPMI),

180
Construction Employees’ Involvement and Motivation (CEIM), and Endogenous Variables
(TQM Implementation). These variables were hypothesised to be characterised by indicator
variables, which collectively constituted the questionnaire items. Factors influencing
construction firms for the implementation and non-implementation of TQM in the Ghanaian
construction industry were also measured by the questionnaire. Table 8.5 gives a
comprehensive summation of the latent and indicator variables.
Table 8.5: Conceptual Model Indicator Variables

Latent Variable Construct Measurement Variables

Leadership and Top Management Leadership style

(TML) Leadership ability in solving quality problem
Leadership skills
Leadership initiatives
Leadership and top management knowledge in
TQM
Top management commitment in TQM
Top management interaction with workers
Top management participation
Top management learning
Top management empowerment
Top management encouragement
Top management’s role model
Company Supplier Quality Partnership with suppliers
Supplier selection criteria
Management (CSQM)
Participation in suppliers

Supplier performance evaluation

Supplier quality audit

Supplier communication

181
 Supplier knowledge of TQM
Location of supplier
Suppliers’ past records
Suppliers’ commitment
Cooperation from suppliers
Suppliers’ orientation
Client Focus and Involvement
Client brief/ input
(CFI)
Client complaint information/feedback

Market investigation

Client satisfaction survey

Quality warranty

Client information system

Client services

Client cooperation

Client orientation

Company Quality System Evaluation of strategy

Evaluation (CQSE)
Evaluation of overall company performance

Evaluation of departments’ performance

Evaluation of employee performance

Evaluation of quality costs

Evaluation of quality manual

Evaluation of quality control system

Evaluation of quality system procedures

182
 Evaluation of end results

Quality audit

Quality benchmarking

Quality information system

Company Vision and Plan Vision statement
Statement (CVPS)
Quality policy

Overall business performance plan

Product quality goal

Quality improvement plan

Formulation of vision and plan

Concreteness of future plan

Employee contribution to the vision

Involvement of employees in the development of

vision statement

Product Selection and Design Client brief/input

Management (PSDM)
Cost implication

Environmental issues

Socio cultural issues

Appearance/finishes required

Strength of required

Intended purpose of the material

Where the material to be used

Design methods/techniques

Design detailing
183
 Concurrent engineering

Design experiments/testing

Quality function deployment

Value engineering

Computer-aided design

Construction Process Project monitoring and control improvement

Management and Improvement
Equipment maintenance and innovation
(CPMI)
Inventory management

Use of quality materials

Use of quality tools

Use of quality manual

Understanding of work instructions/quality

manual

Quality system procedures improvement

Obtaining ISO 9000 certification for operation

Understanding of quality control system

Continuity of quality control circle activities

Appropriate use of system structure and standards

Targets and priority measures

Safe working environment

Utilization of analysis results

184
Construction Employees Education and training
Involvement and Motivation
Salary promotion
(CEIM)
Position promotion

Employee rewards

Bonus scheme

Conducive working environment

Involvement of employees within-functional team

Employee participation

Employee suggestion

Improving employee commitment

Job rotation

Involvement in quality control circle

Employee involvement in cross-functional team

Employee recognition

185
 TQM Implementation (TQMI) Greater reduction of construction costs

Higher product quality

Improved schedule performance

Elimination of reworks

Defect-free product at first attempt

Higher safety standards

Higher productivity

Good methods of working

Higher profitability

Higher reduction of construction waste

8.4.4.3 Population
Cooper and Schindler (2006) and Neuman (2006) defined a study population as the entire
group of items in which the researcher has an interest. The population for this study is made
up of the small to large (D4K4 to D1K1) construction companies in Ghana. The majority of
these construction companies are located in the ten regional capitals of Ghana (Accra,
Kumasi, Takoradi, Cape Coast, Koforidua, Ho, Sunyani, Tamale, Bolgataga and Wa).

8.4.4.4 Sample Frame

The sample frame was established by obtaining a list of registered construction companies
in good standing from the Association of Building and Civil Engineering Contractors of
Ghana (ABCECG). There were 1 282 construction companies in good standing which were
registered with the ABCECG as at March 2017. The selection of the number of construction
companies in the list was made possible based on a probability sampling technique
186
(discussed in the next section). The construction companies selected from the list were
instructed as in the cover letter of the questionnaire (see Appendix F) that their responses
were of utmost importance to the researcher.

8.4.4.5 Sampling Method

Sampling is the process of selecting a sample consisting of units such as people, or
organisations from a population of interest. By studying the sample we may answer the
questions posed regarding some aspects of the population from which they were chosen
(Trochim & Donnelly, 2007). The two general sampling methods are probability and non-
probability sampling, which are usually differentiated by their randomness. A non-
probability sample is also known as a non-random sample whereby samples are selected in
some way not suggested by probability theory but using something other than a
mathematically random process (Neuman,). On the other hand, probability sampling allows
each segment of the population to be represented in the sample. Probability sampling is also
known as random sampling, hence Cooper and Schindler (2006) state that a probability
sample is one based on the concept of random selection – a controlled procedure that
assures each population element is given a known non-zero or allows each member of the
population to have an equal chance of being selected.

In this case the samples were chosen from a larger population by a process known as
random selection. The various sampling techniques employed in the selection of a
probability sample are simple random, stratified random, systematic, and cluster sampling.
Simple random sampling allows the sample to be chosen by simple random selection where
every member of the population has an equal chance of being selected while stratified
random sampling occurs in populations which consist of different strata or groups. In order
to have equal representation in a stratified sample, the researcher selects samples equally
from each one of the strata or group whereas cluster sampling sub-divides an expansive
area into smaller units, for instance, a country could be sub-divided into regions, and further
into towns. The clusters must be as similar to one another as possible, with each cluster
containing an equally heterogeneous mix of individuals and a subset of the identified
clusters is randomly selected.

187
The current study used the probability sampling method, which allows all segments of the
construction companies as defined above to be represented in the sample, making sure that
a representative sample of companies was selected for this study. Therefore, a simple
random sampling technique was used which allows each member of the population to have
an equal chance of being selected (Kerlinger & Lee, 2000). The rationale for selecting this
method of sampling was based on the nature and composition of the companies in Ghana.
The selection of a representative sample for this study was also based on the justification by
Smith (2004) who informed that random sampling must be used for a study of this nature,
hence it was adopted.

8.4.4.6 Sample Size

A sample is a subset of items a researcher selects from a specific population (Neuman,
2006). Sample size is therefore an important feature of any empirical study in which the
goal is to make inferences about a population from a sample, as in the case of the current
study. Sample size, in other words, is the number of observations or replicates to be
included in a statistical sample. The sample size used in the study was determined based on
the expense of data collection and the need to have sufficient statistical power to validate
the conceptual model. The sample size for the current study was not based on the entire
population of the study; therefore the sample size was not equal to the population size. The
sample size was based on the number of respondents from 250 to 500 or more using the
required number under SEM. The question of how large a sample should be depends on the
following: the kind of data analysis the researcher plans to use; how accurate the sample has
to be for the researcher’s purposes and the population characteristics (Neuman, 2006).

Malhotra (1999) indicated that the determination of sample size depends on the proposed
data analysis techniques, financial and access to sampling frame. The proposed data
analysis technique for this research is SEM utilizing EQS software, which is very sensitive
to sample size and less stable when estimated from small samples (Tabachnick & Fidell,
2001). As a general rule of thumb, at least 300 cases is deemed comfortable, 500 as very
good and 1000 as excellent (Comrey & Lee, 1992; Tabachnick & Fidell, 2001), thus for
this study it was decided to target a sample size of 50% from the entire population of the
188
study. This decision was based on the assertion by Leedy and Ormrod (2005) that if the
population size is around 500, 50% of the population should be sampled and if the
population size is around 1500, 20% should be sampled. Therefore because the entire
population of the study is 1282, a sample size 50% (641) chosen from the population was
deemed adequate for the study. Neuman (2006) informs that a large sample size alone does
not guarantee a representative sample. This is because a large sample size without random
sampling or with a poor sampling frame is less representative than a smaller one with
random sampling and an excellent sampling frame. However, the larger the size of the
sample, the more likely its mean and standard deviation will be representative of the
population’s mean and standard deviation. A larger sample also makes it less likely that the
researcher will obtain negative results or fail to determine the truth.

Hence, Leedy and Ormrod (2005) opined that researchers should endeavour to maximize
the sample size and provide the following guidelines for selecting a sample size:
i. For small populations with fewer than 100 people or other units, there is little point in
sampling; therefore the entire population should be sampled;
ii. If the population size is around 500, 50% of the population should be sampled;
iii. If the population size is around 1500, 20% should be sampled, and
iv. Beyond a certain point (at about 5000 units or more), the population size is almost
irrelevant and a sample size of 400 should be adequate.
Neuman (2006) recommended that for small populations, a researcher needs a large
sampling size and for moderately large populations, a smaller sample size of about 10 per
cent is needed to be equally accurate. However, Cooper and Emory (1995) and Cooper and
Schindler (2006) disagree on the 10 per cent sample size recommendation for smaller
populations. Cooper and Schindler (2006) inform that a sample size does not have to be
large for it to be representative of the population. They state that the absolute size of a
sample is much more important than its size compared to the population, and how large a
sample should be is a function of the variation in the population parameters under study and
the estimating precision needed by the researcher. They suggested that a sample of 400 may
be appropriate sometimes, while more than 2 000 are required in other circumstances; and
in another case, perhaps only 40 are called for.
189
Furthermore, Smith (2004) simplified the process of sampling size by recommending that
one may use 20 cases or 5 per cent, whichever is greater for the population. Schiller further
informed that the sample size should vary with the type of study, and that a routine review
study would require 5 per cent or 30 observations; a query review study would require 10
per cent or 40 cases, whichever is greater; an intensive review study would require a sample
size of 15 per cent or 60 cases and a sentinel event would require a 100 per cent of the
observations. Hence, given the kind of data analysis method (SEM) used in this study and
the avoidance of negative results which would jeopardize the model goodness-of-fit, thus
failing to establish the truth with regard to the constructs which predict TQM in the
construction industry, a large sample size of 641 was considered. This is because the role of
sample size is crucial in SEM analysis (Lucko & Rojas, 2010). Therefore the sample size
requirement in this thesis was a function of the model framework development
consideration. Harris and Schaubroeck (1990) proposed a sample size of 200 at least to
guarantee robust SEM. Kline (2010) suggested that a very complicated path model needs a
sample size of 200 or more, while Bagozzi and Yi (2012) proposed that the sample size
should be above 200. Also, based on Smith’s (2004) research classification, the study is
both a query and an intensive review; hence the selection of the sample size demanded a 10
per cent or 40 cases and 15 per cent or 60 cases, whichever is greater. The study sample
size also agrees with Neuman’s (2006) recommendation of a 10 per cent sampling size.
Therefore, the total sample size of the respondents from the entire population of the study
was 641, which aligned perfectly with the analysis of covariance structure estimation
requirement.

8.4.4.7 Sample Selection

This step required a detailed specification of all the steps discussed above (Malhotra, 1999).
A total of 641 respondents were chosen for the research, which was equivalent to the
sample size. The names of construction companies (entire population) were assigned with
codes and entered into a computer and the 641 sample for study was generated randomly
from the computer. This process was essential in obtaining true representativeness of the
entire sample.
190
 8.4.4.8 Fieldworkers
Fieldworkers were recruited prior to the actual survey to assist with the administering of the
questionnaires. A fieldworker is defined as an objective collector of data. He or she may or
may not have formal qualifications but is perceived to have access to a particular
community. Maart and Soal (1996:1) determined that a fieldworker primarily mediates or
facilitates the learning of individuals and groups to create an environment in which people
can participate. Tamblyn and Shelton (1996) defined in a comprehensive market research
manual the data collection skills that fieldworkers should possess, and thus recommended
that fieldworkers must be selected with great care and be trained for a stipulated minimum
of four hours before undertaking quantitative data collection. They concluded that it is
essential for fieldworkers to have a good understanding of the area and the respondents and
to be trained in the skills necessary for relating to people, analysing situations and designing
strategies.

For this study, fieldworkers were recruited from the Bachelor of Technology students of the
Department of Building Technology at the Cape Coast Technical University. They were
selected based on the researcher’s working knowledge of their ability and competence in
construction management issues. Their resident status in the survey areas was also taken
into consideration. The fieldworkers were trained by the researcher on the use
(administering) of the questionnaire. Five hours of an intensive training workshop was
organised for the fieldworkers. The selection and training of the fieldworkers took place
after determining the sample size for the study. During the collection of data, a total of six
(6) fieldworkers were used. This was based on the sample size (641) and number of days
(60) available for the data collection. The researcher helped the fieldworkers to identify and
reach the selected construction firms for the study through contact telephone numbers and
email addresses of the firms. The researcher conducted the first day survey with the
fieldworkers to ensure that the fieldworkers followed instructions as stipulated on the
survey instrument. He controlled the data quality by checking for errors during the survey
and after each survey, checked whether the questionnaires had been completed fully and
correctly. Following this procedure ensured that all the respondents answered the questions
fully and correctly and identified problem areas, which were adequately resolved.

191
 8.4.4.9 Data Collection
The questionnaire was personally administered to the companies by the researcher and six
selected trained fieldworkers. A personally administered or face-to-face structured
questionnaire for data collection was the preferable option (Fowler, 1993) used for the
current study. It took approximately 25 minutes to complete each questionnaire, although
respondents were informed in the cover letter that it would take 20 minutes to complete.
The process of data collection took two months, from the third week of the month of
January 2018 to third week of March 2018. Most of the questionnaires were completed
immediately by the respondents and where necessary, they were given clarity. Other
respondents who were busy at the time when the questionnaire was administered and could
not complete the questionnaire at the time requested to complete the questionnaire in their
leisure time. The data collection process took a long time because of this process. Out of
641questionnaires administered, 536 were fully completed and retrieved. All the 536
questionnaires collected were sent for data capturing by the Statistics Department of the
University of Johannesburg.

8.4.4.10 Data Analysis from the Questionnaire Survey

Coding the responses, cleaning, screening the data and selecting the appropriate data
analysis strategy are steps involved in data analysis. Coding of the questionnaire involved
identifying, classifying and assigning a numeric or character symbol to data, which may be
done in two ways: pre-coded and post-coded (Wong, 1999). The aspect of data analysis
from the questionnaire survey has been elaborated in Chapter 11. Taken from the list of
responses, a number corresponding to a particular selection was given. This process was
applied to every question that needed this treatment. Upon completion, the data was then
entered onto a Statistical Analysis Software Package (SPSS) for the next analysis steps. In
choosing the appropriate statistical analysis technique, the following were research
elements considered, namely the research problem, objectives, characteristics of data and
the underlying properties of the statistical techniques (Malhotra, 1999). To meet the
purposes of this study, descriptive and inferential analyses and the measures of goodness-
of-fit of model were applied where necessary. The data analysis involved the use of
multiple analytical techniques to facilitate ease of communicating the results, while at the

192
same time improving its validity. Hence, the researcher chose to use SEM utilizing EQS
software. Raw data from the questionnaire was entered into the SPSS software and was
later exported to the SEM software EQS Version 6.2 for analysis. The motivation for the
choice of the SEM and particularly the use of the software EQS is explained in the next
session.

Inferential analysis refers to the cause-effect relationships between variables which the
current study hopes to establish between the identified model constructs. Inferential
statistics use the results obtained from samples to generalize about a population (Forzano,
2008). Inferential statistics used for this research were correlations, exploratory factor
analysis (EFA) confirmatory factor analysis (CFA) and SEM. SEM was used for the
development and validation of the TQM model. The statistical significance of the constructs
was evaluated. The result’s statistical significance was expressed by p-value (Forzano,
2008). When the p-value is high, there is less possibility of an association between two
variables (McClave, Benson & Sincich, 2008), while a smaller p-value gives a better
likelihood of association. The p-value chosen in the present study is 0.05, which implies a
95 per cent chance that the population mean is within a listed range of values (McClave et
al., 2008). SEM is currently the most inclusive statistical procedure in social and scientific
research catering for all operations of the general linear modelling (GLM) group of
statistics such as analysis of variance (ANOVA), multivariate analysis of variance
(MANOVA) and multiple regression (Kline, 2005:14). Though there are many ways to
describe SEM, it is most commonly thought of as a hybrid between some form of ANOVA
or regression and some form of factor analysis. In general, it can be said that SEM allows
one to perform some type of multilevel regression or ANOVA on factors. SEM is
conceptually used to answer any research question involving the indirect or direct
observation of one or more independent variables or one or more dependent variables.

However, the primary goal of SEM is to determine and validate a proposed causal process
or model, or both. In the current study, the conceptualized holistically integrated TQM
model for Ghanaian construction industry is being validated. SEM takes a confirmatory
approach to the analysis of a structural theory bearing on some phenomenon (Byrne, 2010).
193
However, Dion (2008:365) claims that SEM simultaneously estimates all coefficients in the
model and therefore it is able to assess the significance and strength of a relationship in the
context of the entire postulated model. Hence, considering the conceptualized model in this
thesis of unobserved (exogenous) variables which had to be estimated from the observable
variables, methods of analysis such as ANOVA could not be used as they lack a direct way
of distinguishing between observed measures and the underlying constructs (Kline,
2005:14). A clear distinction is made in SEM between true variance and error variance,
which implies that model parameters are estimated by taking measurement error into
consideration. CFA was carried out on each exogenous variable to determine the best fit for
the model before the SEM was performed.

The choice of the software EQS for analysis was enhanced by the benefit of utilizing the
Satorra-Bentler scaled statistics (S – Bχ2), which provide an adjusted, more robust measure
of fit for non-normal data. This approach is more accurate than the normal chi-square test
statistics (χ2)(Byrne, 2006:22). Likewise Kline (2005:83), Musonda (2012) and Aigbavboa
(2013) inform that EQS offers several different estimation methods for non-normal data as
well, including the robust maximum likelihood (RML). EQS Version 6.2, a software
package, was used for SEM as it is user-friendly software that provides a graphical user
interface, which is easy to understand. EQS also enables data to be imported directly from
SPSS. Other reasons why the researcher adopted EQS 6.2 and SPSS 20.0 software include
first, the software is available at the University’s Postgraduate and Statistics Centre. It was
easier for a postgraduate research student to access the software; second, as revealed from
the literature, EQS had seldom been used by previous researchers to enhance conceptual
understanding of TQM research as compared to other techniques, such as AMOS and
LISREL (Tong, 2007). Being a user-friendly graphically modelling interface, EQS offers a
wider variety of goodness-of-fit measures (Tong, 2007).

8.4.4.11 Data Screening and Preliminary Analysis

The use of the SPSS program version 20.0 was to ensure consistency in data and provide
meaningful interpretation of results. Screening of data was carried out during the

194
preliminary analyses. This includes data cleaning, the handling of missing data, the
normality test and the outliers. The screened data were further analysed using more
complex analyses including EFA, CFA and the SEM.

8.4.4.12 Data Cleaning

To ensure the accuracy of the data being coded and entered into the data file, a verification
procedure was carried out. In this process, data were examined using descriptive statistics
and graphic representations of the variables (Tabachnick & Fidell, 2007). Data cleaning can
be achieved by frequency tables, histograms, bar stem-and-leaf displays, and box plots
(Pallant, 2007). Summaries of the values for the respondents’ profiles and firms were
obtained in frequency tables. Percentages and graphic displays were used for the descriptive
methods to simplify and characterize the data.

8.4.4.13 Normality

According to Hair et al. (1998), sample size affects a study’s findings where the outcome of
smaller samples has too little statistical power for the test to realistically identify significant
results. Sample size can also be easily ‘over-fitting’ to the data in that they fit the sample
very well, but yet have no generalizability. Large sample sizes of more than 200 to 400
respondents have disadvantages due to making the statistical tests overly sensitive as a
result of the increased statistical power from the sample size (Hair et al., 1998) which such
data can incur non-normality. Therefore, the data obtained were analyzed for normality to
ensure their suitability using standard multivariate analysis. Normality of data can be
examined through statistical approaches such as skewness and kurtosis, the Kolmogorov-
Smirnov test and graphical approaches, for example, histograms and box plots (Pallant,
2007). The variable’s frequency value distribution should approximate the bell-shaped
curve or a straight diagonal line to attain normality of the data (Pallant, 2007; Hair et al.,
1998). The skewness and kurtosis were used for this study and established that the data was
slightly non-normal. To overcome the non-normality of the data, maximum likelihood
estimation with robust standard errors and chi-square (MLR) EQS program version 6.2 was

195
used in this study when the CFA and SEM were analyzed. This estimator method rectifies
non-normal data.

8.4.4.14 Treatment of Outliers

Outliers are cases with extreme values on a single variable (univariate) or on a combination
of variables (multivariate) (Pallant, 2007). Some causes of outliers are data entry errors,
unusual events, unexplainable observations, and unusual or unique combined patterns (Hair
et al., 1998). Univariate outliers were identified by examining the box plot of each variable
(Tabachnick & Fidell, 2001). Few outliers were identified in the current study, therefore the
items were removed.

8.4.4.15 Criteria for Determining Reliability and Validity

The content validity is the extent to which a constituent variable belongs to its
corresponding construct (Pallant, 2007; Leedy & Ormrod, 2010). Since content validity
cannot be tested using statistical tools, an in-depth literature survey was necessary to
specify the variables and define the latent variables in order to keep the researcher’s
judgment on the right track. An exploratory study with TQM experts using the Delphi
method was conducted. The finalized variables are indicated in Appendix D deriving the
achievement of content validity.

8.4.4.16 Content Validity

The content validity is the extent to which a constituent variable belongs to its
corresponding construct (Pallant, 2007; Leedy & Ormrod, 2010). Since content validity
cannot be tested using statistical tools, an in-depth literature survey was necessary to
specify the variables and define the latent variables in order to keep the researche’s
judgment on the right track. An exploratory study with TQM experts using the Delphi
method was conducted. The finalized variables are indicated in Appendix D deriving the
achievement of content validity.

196
 8.4.4.17 Reliability
Reliability tests were conducted after establishing the content validity and preliminary data
analyses. Scale reliability is the correlation between two scores ranging from 0 to 1.00
where the Cronbach's alpha is the most common form of internal consistency reliability
coefficient. A lenient cut-off of 0.60 is common in exploratory research. The generally
agreed upon lower limit for alpha is 0.70 (Hair et al., 1998) and a cut-off of 0.80 for a good
scale (Lingard et al., 2011). The adopted cut-off alpha for this study is 0.70 and measures
below 0.70 were eliminated. Composite or construct reliability for CFA was calculated after
the re-specification of the measurement model. Composite reliability represents a better
choice; it draws on standardized loadings and measurement error for each item; however, it
was reported that only 20% of SEM studies reported composite reliability to exceed 0.60
(Shook et al., 2004). This was calculated as follows: {(Sum of standardized loadings)²} /
{(sum of standardized loadings)² + (sum of indicator measurement error)}.

8.4.4.18 Convergent Validity

The analysis of convergent validity was carried out after the establishment of the content
validity and preliminary data analyses. Convergent validity was tested by determining
whether the scores of items in one scale correlate with the scores on the other scales and
converge or load together on a single construct in the measurement model (Hair et al.,
2006). Furthermore, Hair et al. (2006) opined that factor loadings of 0.30 and 0.40 are
considered significant for sample sizes of 350 and 200 respectively. No factor loadings
were less than 0.30 in the current study. This would have been considered not to be
significant (Field, 2005; Hair, et al., 2006). The current study used a cut-off value of 0.30 to
retain the variables as suggested by Hair et al. (2006).

8.4.4.19 Discriminant Validity

The analysis of discriminant validity was carried out after the establishment of the content
validity and preliminary data analyses. Discriminant validity is the extent to which items
representing a latent variable are unique and capture some phenomena that other measures
do not (Hair et al., 2006). Items on one construct should not load or converge too closely
197
with items from other scales. Different latent variables that correlate too highly may indeed
measure the same construct instead of different constructs. Hair et al. (1998) indicated that
items in a construct attaining inter-correlations below 0.90 suggest that there is no
multicollinearity, but that the construct has discriminant validity. Another method of
attaining discriminant validity as indicated by Fornell and Larcker (1981) is the average
variance estimate (AVE). The AVE was calculated as follows: {(Sum of squared
standardized loadings)} / {(sum of squared standardized loadings) + (sum of indicator
measurement error)}. Despite the two methods of analyzing discriminant validity, the inter-
correlation of the items in a construct-element was used for this study.

8.4.5 Methods of Data Analysis of the Quantitative Survey

Descriptive and multivariate correlation data analyses were conducted after screening the
data. Three major steps were carried out in the multivariate correlation analysis: EFA, SEM
which includes CFA and structural model testing. Analyses were conducted on selected
respondents’ profiles and firm variables with the elements of the TQM model.

8.4.5.1 Exploratory Factor Analysis

EFA was performed to gather information about the unidimensionality of the factors to
yield their factor analysability. The EFA was conducted using SPSS version 20. Seo,
Torabi, Blair and Ellis (2004) asserted that EFA is a precursor to the SEM. EFA was used
in the current study to confirm the validity and reliability of the eight element-constructs of
the proposed construction TQM model in the construction industry. The maximum
likelihood, with a minimum eigenvalue of one, together with principal axis factoring with
Oblimin Kaiser normalization was specified as the analysis method for this study. A
bivariate correlation was performed to assess the strength among the research elements of
the TQM model.

8.4.5.2 Structural Equation Modelling

SEM was performed immediately after EFA. The development of the methods of analysis
involving SEM with latent variables has provided researchers with considerable means to
construct test and modify theories (Anderson & Gerbing, 1982). SEM represents a
198
component of the methodological instrument of the social sciences. It is a comprehensive
statistical approach for testing hypotheses about relationships among observed and latent
variables (Kline, 2005; Molenaar, Park & Washington, 2009). It assesses whether the
sample covariance matrix is consistent with hypothesised model (Sweeny, 2009). In
comparison to other statistical analytical techniques such as factor and regression analysis,
SEM is a relatively young field. The cradle of SEM was pioneered in the 1920s, when
Sewell Wright attempted to solve simultaneous equations to disentangle genetic influences
across generations (Murayama, 1998). The increasing complexity of research questions in
the social and behavioural sciences and appearance of flexible user-friendly computer
software programs, especially with the advent of windows applications (Marcoulides &
Hershberger, 1999), has increased the interest in SEM (Kelloway, 1998). It has been argued
to be the most important multivariate correlational analysis technique (Byrne, 2010). SEM
can also be used to analyse and test theoretical models (Schumacker & Lomax, 2004).
Recent research has shown that SEM has become increasingly popular in the construction
industry literature. For example, Molenaar et al. (2009), Mohamed (2002), Chinda and
Mohamed (2008), Musonda (2012), and Agumba (2013) used SEM to develop models in
the construction industry. Therefore SEM was considered the most suitable method of
analysis for the third phase of the current study in testing the conceptualised model.

8.4.6 Structural Equation Modelling Analysis

A two-step process was followed after the EFA to test the structural model. Anderson and
Gerbing‘s (1988) recommended two approaches which included CFA and structural model
testing. Firstly, it is vital to confirm the measurement model before the structural model can
be finalized. Tabachnick and Fidell (2007) indicated that measurement models test
relationships which are paths between the measures, also known as manifests or observed
variables, and constructs, also termed latent variables, whilst the structural model clarifies
the casual relationships, as well as the degrees of influence.

8.4.6.1 Structural Equation Modelling (SEM) Process

Structural model testing can be carried out through five basic steps involved in SEM
analysis, namely model specification, model identification, model estimation, model testing
199
and model modification (Schumacker & Lomax, 2004). The study adopted the Schumacker
and Lomax (2004) process. However, Hair et al. (1998) proposed a seven-step process for
SEM analysis which included developing a theoretical model, constructing a path diagram
of causal relationships, converting the path diagram into a set of structural equations and
measurement models, choosing the input matrix type (correlation matrix or covariance
matrix) and estimating the proposed model, assessing the identification equations and
measurement models, choosing the input matrix type (correlation matrix or covariance
matrix) and estimating the proposed model, assessing the identification of model equations,
evaluating the results of goodness-of-fit and making the indicated modifications to the
model if theoretically justified.

8.4.6.1.1 Step 1: Model Specification

Model specification is the first vital step in SEM. It involves developing a theoretical model
(Schumacker & Lomax, 2004). This process must be guided by a combination of theory and
empirical results from previous research (Hair et al., 2006), although the role of informed
judgement, hunches and dogmatic statements of belief should not be discounted (Kelloway,
1998). This was the first step performed in this study. In particular, attention must be paid
to include all relevant variables and only those variables that are relevant. If the theoretical
model is not consistent with the true model, the theoretical model is said to be mis-specified
and lacks validity. This may occur if the researcher fails to include an important variable or
an important parameter or alternatively, if an unimportant parameter or variable was
included in error (Schumacker & Lomax, 2004).

Having developed the theoretical framework of the model, the next step is to illustrate this
in a path diagram, which is a pictorial representation of all relationships in the model (Hair
et al., 1998). This is a graphical representation of how the various constructs of the model
relate to one another (Byrne, 2010). This is essentially the first step in the SEM process
(Byrne, 2010; Kline, 2005). While it is not a formal requirement of SEM, construction of a
path diagram offers important benefits. Specifically, the hypotheses contained in the model
are much more easily comprehended in visual form than in either verbal or mathematical
terms. It may also help improve the conceptualization of the model by drawing attention to

200
omitted links or excluded variables (Diamantopoulos, 1994), therefore decreasing the
possibility of specification error (Kline, 2005). Path diagrams not only enhance the
understanding of structural models, but substantially contribute to the creation of the correct
input files (Raykov & Marcoulides, 2000).

Path models adhere to common drawing specifications that are utilized in SEM models. The
observed variables are enclosed by boxes or rectangles. The relationships between the latent
variables and their corresponding indicators are represented by arrows which originate at
the latent variable and end at the indicators. Each indicator is also associated with an error
term representing the errors in measurement. The error terms associated with the
endogenous variables represent error in equations and indicate that the dependent variables
in the model are not perfectly explained by the independent variables. A curved double
headed line between two independent variables indicates covariance (Byrne, 2010; Kline,
2005).

8.4.6.1.2 Step 2: Model Identification

The second step of SEM is model identification. This is crucial because identification
problems should be resolved prior to the estimation of parameters (Schumacker & Lomax,
2004; Kline, 2005). Identification revolves around the question of whether one has
sufficient information to obtain a unique solution for the parameters to be estimated by the
model (Byrne, 2010). Identification determines whether it is possible to find unique values
for the parameters of the specified model (Kline, 2005). It concerns the correspondence
between the information to be estimated (the free parameters) and the information from
which it is to be estimated (the observed variances and covariances) (Hoyle, 1995).

Models can be under-identified, just-identified or over-identified. A model is considered

just-identified if it has only one estimate for each parameter and generates zero degrees of
freedom and therefore cannot be rejected (Byrne, 2010). A just-identified model will
always provide one unique solution that will be able to perfectly reproduce the correlation
matrix (Kline, 2005). However, the solution is uninteresting because it has no
generalizability (Hair et al., 2006). An under-identified model is obtained when one or more

201
parameters are not uniquely determined; that is the number of unknowns exceeds the
number of equations, therefore there is no much empirical information to allow its unique
estimation (Schumacker & Lomax, 2004; Kline 2005) and therefore its estimation should
not be relied upon (Kline, 2005). The most accepted situation is one in which there are
more indicators than unknown variables and the model is over-identified and has positive
degrees of freedom (Byrne, 2010). Only models that are identified can be estimated (Kline,
2005). In an over-identified model, there are a number of possible solutions, and the task is
to select the one that comes closest to explaining the observed data within some margin of
error (Kelloway, 1998).

Hoyle (1995) cautions that determining the identification status of a model can be difficult.
Providing some guidance, Raykov and Marcoulides (2000) urge researchers to simply count
the number of parameters in the model and subtract this from the number of non-redundant
elements in the sample correlation matrix, determined as follows:
N (N + 1) / 2, [where N = the number of observed variables in the model]. The resulting
difference is referred to as the degrees of freedom. If positive, the model is considered to be
identified (Hair et al., 2006). The current structural and measurement models were over-
identified and were therefore appropriate for testing.

8.4.6.1.3 Step 3: Model Estimation

The third step is model estimation. The purpose of estimation is to generate numerical
values for free parameters within the model that produces the implied matrix Σ so that the
parameter values yield a matrix as close as possible to the sample covariance matrix Ѕ. The
estimation process involves the selection of a particular fitting function to minimise the
difference between Σ and S (Schumacker & Lomax, 2004; Kline, 2005) and this can be
achieved by use of EQations (EQS). EQS can estimate both structural equation models and
path models for a single or multiple groups. In addition, EQS can estimate models with
regressions among combinations of continuous latent variables and observed variables. It
can also estimate factor indicators and other observed dependent variables when they are all
continuous. Several fitting functions or estimation procedures are available to be used. EQS
can use different estimator choices: maximum likelihood (ML), maximum likelihood with

202
robust standard errors and chi-square (MLR), generalized least squares (GLS), and
weighted least squares (WLS) (Muthèn & Muthèn, 2007). This study adopted MLR in EQS
program because of slight non-normal data identified in the process of data screening and
its robustness to rectify non-normality.

8.4.6.1.4 Step 4: Model Testing

Once the parameter estimates are obtained for a SEM model, the fourth step is to determine
how well the data fit the model (Schumacker & Lomax, 2004). Assessing whether a
specified model fits the data is an important step of SEM (Yuan, 2005) as it determines
whether the model being tested should be accepted or rejected. Model fit refers to the extent
to which a hypothesized model is consistent with the data (Schumacker & Lomax, 2004).
Goodness-of-fit under SEM is defined by Hair et al. (2006) as the degree to which the
actual/observed input matrix is predicted by the estimation model. A model is said to fit the
observed data to the extent that the covariance matrix it implies is equivalent to the
observed covariance matrix (Hoyle, 1995). The process of estimation results in an implied
covariance matrix Σ which is as close as possible to sample covariance matrix S; the closer
Σ is to S, the better the fit of the model (Schumacker & Lomax, 2004). Model fit represents
one of the most controversial areas of SEM as indicated by Barrett (2007).

If the model does not fit the data in such a way that the observed covariance matrix is
statistically different from the covariance structure of the model, either the model or the
data should be rejected (Fornell, 1983). The issue of model fit assessment has been a real
problem (Barrett, 2007). It represents a major challenge facing theory developers and
researchers as different indexes and values are reported (Kline, 2005; Barrett, 2007), unlike
many statistical procedures that have a single, powerful fit index, for example, F-test in
analysis of variance (ANOVA). In SEM there is an increasingly large number of model fit
indices (Schumacker & Lomax, 2004). Determining the tests that best suit the model is a
matter of discretion. As such, there is a possibility that only those fit measures that fall
within the acceptable range and support the proposed model are reported (Kline, 2005).
Model fit indices provide no guarantee that a model is useful. Fit indices provide
information on a model’s lack of fit and do not reflect the extent to which the model is

203
plausible (Kline, 2005). Even if a model fits well, data can never confirm a model. They
can only fail to disconfirm. Other equal or better fitting models may exist (Maruyama,
1998).

Discussion of the Selected Fit Indices

Assessment of model fit under SEM is an important process (Yuan, 2005; Kline 2005) as fit
measures are continually evolving (Kline, 2005). It determines whether the proposed model
should be accepted or rejected (Hu & Bentler, 1999). It is unnecessary and unrealistic to
include every index included in the program’s output (Hooper, Coughlan & Mullen 2008).
Similarly the cut-off criteria for fit-indexes vary in different publications. According to Hu
and Bentler (1999), it is difficult to designate a specific cut-off value for each fit index
because it does not work equally well with various conditions. Similarly, Schumacker and
Lomax (2004) highlighted the controversy and discussion on subjective interpretation and
appropriateness under specific modelling conditions used. It is now common practice to use
multiple tests when evaluating and reporting overall model fit (Hu & Bentler, 1999).
Although there is not an agreed and strict list of fit indexes to be examined and criteria to be
met while evaluating the model fit, Hooper et al. (2008) indicated the following three types
of goodness of fit measures used:
i. Absolute fit measures or accurate fit measures, which measure the overall model fit
of ?both measurement and structural models, with no adjustment for the degree
of over-fitting that might occur using root mean square error of approximation
(RMSEA) and standardised root mean squared residual (SRMR);
ii. Incremental fit measures, that compare the proposed model to a baseline model
specified by the researcher using the Tucker-Lewis fit index (TLI) and
comparative fit index (CFI); and
iii. Parsimonious fit measures, which adjust the measures of fit to provide comparison
between models with differing numbers of estimated coefficient using chi square
divided by degree of freedom x²/df.

Based on the aforementioned discussions the following model fit indices were selected to
be used for this study. The most common model fit index is the chi-square (x²) goodness-
204
of-fit test (Kline, 2005). It has been reported in previous SEM studies (Agumba, 2013;
Molenaar et al., 2009; Chinda & Mohamed, 2008). A significant chi-square indicates the
rejection of the null hypothesis, suggesting that the model is not plausible in the population.
Goodness-of-fit indices usually represent indexes ranging from zero to one, with zero
indicating a complete lack of fit and one indicating perfect fit. However, with regard to the
chi-square test, zero indicates perfect fit and large numbers indicate extreme lack of fit; as
such the chi-square test has been referred to as a lack of fit test (Mulaik, James, Van
Alstine, Bennet, Lind & Stilwell, 1989).

However, goodness-of-fit tests tend to be quite sensitive to sample size. This is especially
true for the chi-square test (Hair et al., 2006). When the sample size becomes very large, the
chi-square test increases in sensitivity and becomes impractical (Kelloway, 1998).
Similarly, a poor fit based on a small sample may result in a non-significant chi square,
implying one should accept the model (Kline, 2005). Based on the sensitivity of the chi-
square index to sample size (Schumacker & Lomax, 2004; Kelloway, 1998), a number of fit
indices were chosen for this study to supplement the chi-square index. The normed chi-
square was used as a supplement index to the chi-square. The normed chi-square reduces
the sensitivity of the sample size. Its value is achieved when the value of chi-square (x²), is
divided by the degrees of freedom (df), (x²/df), which results in a lower value called the
normed-chi-square. The minimal acceptable cut-off value has not been clearly established
(Kline, 2005). In a recent study by Hsu et al. (2012), they proposed values of less than 2 or
3 to be standard and applicable. However, some studies in construction industry have used
cut-off values of less than 5 as acceptable to data fit (Chinda & Mohamed, 2008; Agumba,
2013).

A cut-off value less than 5 was adopted in this study. The root mean square error of
approximation (RMSEA) focuses on the discrepancy between the model and population
covariance matrices per degree of freedom. It is generally regarded as one of the most
informative fit indices (Kline, 2005). RMSEA has been reported in previous structural
equation modelling studies (Agumba, 2013; Molenaar et al., 2009; Chinda & Mohamed,
2008; Fernandez-Muniz et al., 2007). An RMSEA value of 0.05 or less is indicative of good
205
fit. Values between 0.05 and 0.08 are considered reasonable. Values between 0.08 and 0.10
are considered moderate fit while values greater than 0.10 indicate poor fit (Hsu et al.,
2012). Furthermore, Steiger (2007) argued that a cut-off value of 0.07 is considered
acceptable. A cut-off value of less than 0.05 was adopted for this study. Standardised root
mean squared residual (SRMR) has a lower bound of zero and an upper bound of one.
Generally values below 0.05 are interpreted as indicating good model fit (Hair et al., 2006;
Hsu et al., 2012) whereas values ranging from 0.05 to 0.09 are considered acceptable fit
(Hu & Bentler, 1998). Kline (2005) postulates that values less than 0.10 are favourable.
This index was therefore adopted for this study.

However, the CFI is the revised form of the normed fit index (NFI) which takes into
account the sample size (Byrne, 2010). CFI performs well even with small samples
(Tabachnick & Fidell 2007). In1990 Bentler introduced the CFI and subsequently included
it as part of the EQS program (Kline, 2005). The CFI assumes that all latent variables are
uncorrelated (null/independence model) and compares the sample covariance matrix with
the null model. The statistical values for CFI range between 0.00 and 1.00 with values
closer to 1.00 indicating good fit. A cut-off value for CFI greater than or equal to 0.90 (Hsu
et al., 2012) was initially accepted as good fit. However, recent studies have shown that a
value greater than 0.90 is needed in order to ensure that misspecified models are not
accepted (Hu & Bentler, 1999). The CFI greater than or equal to 0.95 is presently
recognized as indicative of good fit (Hu & Bentler, 1999).

The CFI index has been reported in previous studies by Fernandez-Muniz et al (2007),
Chinda and Mohammed (2008), Larsson et al. (2008), Martinez-Corcoles et al (2012), and
Agumba (2013) and was therefore adopted for this study. A cut-off value of greater than
0.80 was adopted for this study. The Tucker-Lewis index (TLI), also known as non-normed
fit index (Schreiber, Stage, King, Nora & Barlow, 2006), penalises model complexity. The
TLI is one of the indexes that are less affected by sample size. TLI greater than or equal to
0.90 (Hsu et al., 2012) indicates acceptable model fit. Some authors have used the cut-off
values of 0.80 since TLI tends to run lower than the goodness-of fit-index (GFI) (Sweeny,
2009). Hu and Bentler (1999) support the use of TLI because it is relatively insensitive to
206
sample size, sensitive to model misspecifications, relatively insensitive to violations of
assumptions of multivariate normality, and is relative insensitive to estimation method.
Previous studies in construction industry have reported on this index (Chinda & Mohamed,
2008; Molenaar et al., 2009; Agumba, 20132013). Therefore this index was adopted for this
current study because of its benefits. A cut-off value of 0.80 was used for this study.

8.4.6.1.5 Step 5: Model Modification

In the final step of SEM, if the fit of the implied theoretical model is not as strong as
desired (which is often the case with initial models) then the next step is to modify the
model and subsequently evaluate this modified model (Schumacker & Lomax, 2004).
Modifications can be made by linking the indicators to the latent variable from fixed to free
or vice versa, allowing or constraining correlations among measurement errors or allowing
or constraining correlations among latent variables (Hair et al., 2006). This process is
known as model modification (Kline, 2005). This is done to improve the model. It implies a
better fitting model and/or more parsimonious model that are substantively more
interpretable. To assist in the process most SEM software provides modification indices for
each fixed parameter. This value indicates the minimum improvement that could be
obtained in the chi-square value if that parameter were fixed for estimation (Kline, 2005;
Schumacker & Lomax, 2004).

Providing a slightly different approach, the Wald index indicates how much a proposed
model’s chi-square would increase if a particular parameter were fixed to zero (Raykov &
Marcoulides, 2000). Another method is to examine the residual matrix, the differences
between the observed covariance matrix S and the model implied covariance matrix Σ.
They should be small in magnitude and should not be larger for one variable than another.
Large values overall signify general model misspecification whereas large values for single
variables indicate misspecification for that variable only (Schumacker & Lomax, 2004).
Testing and revising of models should not become a procedure completely determined by
statistical results and devoid of theoretical underpinnings. Theoretical considerations must
guide model modifications (Kline, 2005) because adjusting a model after initial testing

207
increases the chance of making a Type I error. Blind use of modification indices can lead
researchers astray from their original goal (Raykov & Marcoulides, 2000). In this current
study, the structural model was not modified, but some of the measurement models were.

8.4.7 Ethical Considerations

It is relevant to consider the proper conduct of this research. This research has taken the
responsibility to protect the interests of the survey respondents. With regard to the survey
respondents, no one was forced to respond to this survey. Respondents were asked to
participate of their own free will. They were told of their rights not to participate or to end
their participation if they so wished. In addition, they were briefed about the purpose of the
study and how or why they were chosen. As such they were free from deception or stress
that might arise from their participation in this research. The respondents were also
guaranteed protection through anonymity and all information that may reveal their
identities, which are held in strict confidence.

8.5 CONCLUSION
This chapter presented the methodology adopted for the conduct of this research. It also
provided the justifications for the philosophical position and methods of data collection.
The research design described in this chapter has linked three important elements of the
research methodology, namely the underlying philosophical assumptions, the research
methods/approach, and the process followed in the questionnaire administration, as well as
an introduction to the data analysis. In testing the theoretical model proposed for this study,
the survey method was discussed. The descriptive and multivariate correlation data analyses
were also discussed. Furthermore, the methodology for analysing the proposed conceptual
model was discussed, including the process to be followed when using SEM. The three
main steps were discussed for conducting the multivariate correlation analysis, namely
EFA, CFA and SEM.

Moreover, ethical considerations pertaining to the collection of data were discussed.

Finally, issues relating to validity and reliability were also discussed. The next chapter
presents the results of the data analysis from the Delphi study.
208
 CHAPTER NINE
9 RESULTS FROM THE DELPHI STUDY
9.1 INTRODUCTION
The influence (probability) and factors affecting TQM in the construction industry were
determined by soliciting experts’ views through a Delphi study. Thus, the study identified
the main and sub-attributes that bring about the understanding of TQM in the Ghanaian
construction industry. Two rounds of the Delphi process were conducted to seek experts’
views on TQM in the Ghanaian construction industry. Consensus was reached by the
experts on various questions asked after the second round. The summary of the results from
the consensus reached by the experts after the second round is presented in this section. The
analysis of the results from the questions on the influence (probability) and impact of the
attributes that predict TQM in the Ghanaian construction industry were computed and
presented in the preceding sections. The preceding sections also provided a description and
general background to the Delphi study. This section of the study describes the composition
of the selected panel of experts’ and the findings of the study. The discussion of the
summary of the findings based on the objectives of the Delphi study has also been provided
in this section.

9.2 OBJECTIVES OF DELPHI STUDY

Delphi study was conducted to meet the following objectives:

DSO1. To identify the main and sub-attributes that determine TQM implementation in
Ghanaian construction industry;

DSO2. To determine the factors that enable construction companies to implement TQM;

DSO3. To identify the factors that affect construction companies for not implementing
TQM in Ghana;

DSO4. To evaluate organizational factors that affect the implementation of TQM; and

DSO5. To identify the effects of implementation of TQM on the construction industry

209
The above Delphi study objectives were set to:

i. determine key factors and constructs that are of critical significance (influence) to
determine TQM in construction industry and

ii. develop a holistically integrated conceptual model on TQM for the Ghanaian
construction industry.

9.3 EXPERT PANEL SELECTED FOR THE DELPHI STUDY

The involvement of experts in the Delphi process was a key consideration for this study.
Panel members (experts) for the study were identified from three sources. They comprised
professionals in construction firms, academicians and consultants with a construction
project management background. Some of the selected experts were authors identified from
conference proceedings of the West African Built Environment Research (WABER) and
Association of Researchers in Construction Management (ARCOM) held in Ghana. The
selected participants represented a wide variety of backgrounds to guarantee a wide base of
knowledge as recommended by Rowe, Wright and Bolger (1991). Panellists were recruited
via e-mail, with a brief overview of the study objective included therein. Thereafter, those
that consented to the preliminary invitation were sent a letter inviting them to participate
with a detailed description of the Delphi study (see Appendix C). Experts were asked to
send their curriculum vitae in order to confirm their areas of expertise and to ascertain
whether they meet the qualifying criteria. Seventeen (17) invitations were sent out based on
the set criteria. The five criteria set for the study were met by all experts selected for the
current study. A panel of fourteen (14) experts agreed to participate but only ten (10)
completed the first round and remained as panellists throughout the study. In other words,
the eventual withdrawal of four (4) experts during the first round of the study led to the
completion of the study with ten (10) experts.

This number of panellists was considered adequate based on literature recommendations

from scholars which have employed the technique previously. Ten (10) to fifteen (15)
panellists as suggested by Delbecq et al. (1975) could be sufficient if the background of the
panellists is homogenous, which was achieved in the current study. Rowe and Wright
210
(1999) indicated that the size of a Delphi panel ranges from three (3) to eighty (80) in peer
reviewed studies. Okoli and Pawlowski (2004) and Skulmoski, Krahn and Hartman (2007)
also indicated a panel size of about ten (10) to eighteen (18) members whilst Hallowell and
Gambatese (2010) suggested that since most studies incorporate between eight (8) and
sixteen (16) panellists, a minimum of eight (8) is suggested. Hallowell and Gambatese
(2010) further argued that the size of a panel should be dictated by the study characteristics,
the number of available experts, the desired geographical representation and the capacity of
the facilitator. Therefore a panel of ten (10) experts for this study was considered adequate
based on the fact that the Delphi method does not depend on statistical power but rather on
group dynamics for arriving at consensus amongst experts. Most of the potential experts
who agreed to participate during the initial stage withdrew from the study, probably owing
to the rigorous nature and time required of the Delphi method.

The selected experts’ were required to have a thorough understanding of TQM in their
respective locations. The residence of the experts’ was considered as compulsory for the
selection of participants owing to the importance of the study. Hence, the setting for the
study was based on five metropolitans in Ghana (Cape Coast, Accra, Takoradi, Koforidua
and Kumasi). It is therefore means that all the panel members were Ghanaians. Three of the
experts are currently working in Cape Coast and one of the experts each are in Koforidua,
Takoradi and Kumasi respectively. Four of the remaining experts are in Accra (Table 9.1).

Table 9.1: Residential location of experts

Regional Capitals /Cities Number of Experts

Cape Coast 3
Koforidua 1
Kumasi 1
Takoradi 1
Accra 4
Total 10

211
The highest qualifications held by the experts are tabulated in Table 9.2 below. Three of the
experts had a Doctor of Philosophy (PhD) degree and the remaining seven experts had a
Master of Science (MSc). All experts were involved in construction project quality
management.

Table 9.2: Qualification of Panel of Experts

Highest qualification Number of experts

Doctor of Philosophy (PhD) 3
Master of Science degree (MSc) 7
Total 10

Table 9.3 shows that one panel of experts had 1-5 years of experience, four had 6-10 years
of experience, four had 11-16 years of experience, and one had 16-20 years of experience.
All experts were professionally registered at the highest level with various professional
regulating bodies such as the Ghana Institute of Construction (GIOC), the Ghana Institute
of Engineers, and the Ghana Institute of Surveyors. Others are the Institute of Engineers
and Technology, the Royal Institute of Chartered Surveyors, the Chartered Institute of
Marketing, and the International Association of Valuators, Consultants and Analysts.

Table 9.3: Panel of Experts’ Years of Experience

Years of experience Number of

experts
1-5 1
6-10 4
11-15 4
16-20 1

Total 10

212
Consideration was also given to the selection of the panel of the experts to achieve a
balance between professionals and academicians in the fields of TQM in the construction
industry. In terms of their current occupations, seven (7) of the selected experts were
employed by construction firms in Ghana as project managers and directors and the
remaining three (3) experts were employed by technical universities (Coast Technical
University -2 and Koforidua Technical University – 1). All three (3) panel experts from the
technical universities were Senior Lecturers and held senior positions in the technical
universities. They also handle project management-related courses in their institutions. The
seven (7) experts employed by construction firms held the highest positions at their various
firms (Project Managers – 5, Quality Control Manager – 1, Consultant/Director - 1) and
were involved in TQM. Their curriculum vitae analysis shows that all the experts were
involved in both construction works and related fields (Table 9.4).

Table 9.4: Panel of Experts’ Fields of Specialization

Field of specialization Number of

experts
Construction Project Management 8

Construction Health and Safety Management 1

Civil Engineering (Structural Engineering). 1

Total 10

213
9.4 FINDINGS FROM THE DELPHI STUDY

Frequencies were obtained to measure the degree of consensus reached amongst

participants regarding the attributes that determine TQM implementation in the Ghanaian
construction industry. The results of each Delphi round were reviewed and compiled by the
researcher. The median, mean, standard deviation, percentages and IQD scores of each
question were calculated. All statements were examined individually for consensus. In other
words, the quantitative results were statistically analysed after each round of questionnaires
to determine whether consensus had been reached for each question or statement using the
provided scale for each question or statement. Consensus is achieved with 100% of the
participants in agreement but two-thirds in the agreement is considered a common consent
(Stitt-Gohdes & Crews, 2004). The goal for this study was that each question or statement
should have consensus but common consent would be acceptable. Common consent was
obtained if 60% of the experts agreed on each statement that was achieved in the study. If
consensus was reached prior to the final round that question or statement was no longer
required (asked/required) in the next rounds. After the second round Delphi survey,
consensus was reached regarding most of the attributes that determine TQM
implementation in the construction industry in Ghana. If consensus had not been reached at
the second round, the data from this second round would have been analysed again by
calculations for median, mean, standard deviation and percentages scores and sent to the
experts for consideration in responding to a third round. The researcher predicted the use of
three rounds to achieve consensus but this prediction was incorrect. A third round was not
needed, and the participants were informed of this when the second round questionnaires
were sent out.

Based on the findings of the analyses of responses from the Delphi rounds, a list of
attributes that determine TQM implementation in the construction industry was prepared
which inform the conceptual framework for the broader study. Issues relating to TQM
implementation in the construction industry in Ghana were highlighted, which responded
accordingly to the set objectives of the Delphi study. The results of the Delphi study about
the specific objectives of the Delphi are presented below.

214
DSO 1: To identify the main and sub-attributes that determine TQM implementation in
Ghanaian construction industry
A set of main attributes and sub-attributes that are relevant to TQM were emphasised
through a comprehensive review of literature. Although the reviewed literature was based
on studies from the developed countries, they were collectively used to examine the
attributes that determine TQM in the Ghanaian construction industry. The influence of the
attributes on TQM was obtained as a product of the impact on the construction industry in
Ghana. The main attributes were based on the level of influence as categorized on the
questionnaire. This was established by assessing the extent to which the listed attributes
will determine the TQM implementation in Ghanaian construction industry. The impact of
the sub-attributes in determining TQM was assessed if they were present or lacking. The
rating was based on an ordinal scale of one to ten with one being ‘low influence’ or ‘no
impact’ and ten being ‘high influence’ or ‘very high impact’. The levels of influence and
impact were obtained as a product of the consensus achieved as detailed in Chapter 8,
Section 8.4.3.7.

By applying the interquartile deviation (IQD) to determine whether an attribute reached

consensus or not, four (4) attributes, namely Leadership and Top Management, Company
Quality System Evaluation, Product Selection and Design Management, Construction
Process Management and Improvement reached consensus with the IQD cut-off (IQD ≤1)
score set to reach consensus (Table 9.5).. Six (6) of the attributes had strong consensus,
with the exception of Client Focus and Involvement, and Product Selection and Design
Management, which had good consensus (Table 9.5). Seven (7) attributes from the eight (8)
identified main attributes that determine TQM in the construction industry were considered
by the experts to have a very high influence, with the exception of one (1) attribute (Product
Selection and Design Management) which had a high influence.

215
Table 9.5: Total Quality Management (TQM) in the Construction Industry Main
Attributes

Total Quality Management in the

M SD IQD
Construction Industry Main Attributes
Leadership and Top Management (LTM) 9.10 10.00 1.85 1.00
Company Supplier Quality Management
8.40 9.00 1.43 1.25
(CSQM)
Client Focus and Involvement (CFI) 8.30 8.00 1.06 1.50
Company Quality System Evaluation
9.10 9.00 0.57 0.25
(CQSE)
Company Vision and Plan Statement
8.40 9.00 1.43 1.25
(CVPS)
Product Selection and Design
7.80 8.00 0.79 0.25
Management (PSDM)
Construction Process Management and
8.60 9.00 0.84 0.50
Improvement (CPMI)
Construction Employees’ Involvement
9.00 10.00 2.21 1.25
and Motivation (CEIM)
M=Median; =Mean; =Standard Deviation (SD); IQD=Interquartile Deviation

Twelve (12) attributes were identified from the first sub-attribute (Leadership and Top
Management). By applying the interquartile deviation (IQD) to determine whether an
attribute reached consensus or not, the experts reached consensus on six attributes with the
IQD cut-off (IQD ≤1) score set to reach consensus (Table 9.6). Eight (8) attributes had
strong consensus, with the remaining four (4) having good consensus. The outcome of these
results has been depicted in Table 9.2. The experts considered eight (8) attributes out of
twelve (12) attributes to have very high influence on the attributes that determine TQM in
the construction industry. The experts considered the remaining four attributes to have high
influence. Eight (8) out of the twelve (12) attributes had a very high impact (VHI -9-10) on
TQM in the construction industry, while the remaining four (4) attributes had a high impact
(HI-7-8.99) on TQM in the construction industry (Table 9.6).

216
Table 9.6: Leadership and Top Management

Leadership and Top Management M SD IQD

Leadership style 8.78 9.00 1.09 0.00
Leadership ability in solving quality
8.70 9.00 0.95 0.00
problem
Leadership skills 8.20 8.50 0.92 2.00
Leadership initiatives 8.89 9.00 0.33 0.00
Leadership and top management
9.30 10.00 1.34 1.25
knowledge of TQM
Top management commitment in TQM 9.00 9.00 0.67 0.50
Top management interaction with
8.50 9.00 1.18 1.25
workers
Top management participation 8.56 9.00 1.24 2.50
Top management learning 7.70 8.00 0.95 1.25
Top management empowerment 7.90 8.00 1.10 1.25
Top management encouragement 8.80 9.00 0.63 1.00
Top management’s role model 7.70 8.00 1.25 0.75
M=Median; =Mean; =Standard Deviation (SD; IQD=Interquartile Deviation

From the twelve (12) identified attributes under the second sub-attribute (Company
Supplier Quality Management), the experts reached consensus on seven attributes with IQD
cut-off (IQD ≤1) score set to achieve consensus (Table 9.7). Five (5) attributes had strong
consensus, while six (6) had good consensus and the remaining one (1) (Location of
supplier) had weak consensus in determining TQM in the construction industry. The
representation of all the attributes is shown in Table 9.7.

Table 9.7: Company Supplier Quality Management

Company Supplier Quality

M SD MAD IQD
Management
Partnership with suppliers 7.80 8.00 1.03 0.40 0.00
Supplier selection criteria 8.10 8.00 0.74 0.30 0.00
Participation in suppliers 7.90 8.00 0.32 0.10 0.00
Supplier performance evaluation 8.80 9.00 0.42 0.20 0.25
Supplier quality audit 8.60 9.00 0.70 0.40 1.00
Supplier communication 8.30 9.00 1.34 0.70 1.25
Supplier knowledge in TQM 8.40 9.00 1.43 0.80 1.25
Location of supplier 6.00 6.00 1.63 1.00 1.50

217
 Suppliers’ past records 8.00 8.50 1.33 1.00 2.00
Suppliers’ commitment 8.30 8.00 0.48 0.30 1.00
Cooperation from suppliers 8.20 9.00 1.32 0.80 1.25
Suppliers’ orientation 7.30 7.00 0.82 0.50 1.00
M=Median; =Mean; =Standard Deviation (SD); IQD=Interquartile Deviation

From the nine (9) identified attributes under the third sub-attribute (Client Focus and
Involvement), the experts reached consensus on five attributes with IQD cut-off (IQD ≤1)
score set to achieve consensus (Table 9.8). Five (5) attributes had strong consensus, while
the remaining five (5) had good consensus. Two (2) of the attributes had a very high impact
(VHI: 9-10). The remaining seven (7) other attributes had a high impact (HI: 7-8.99) on
TQM in the construction industry (Table 9.8).

Table 9.8: Client Focus and Involvement

Client Focus and Involvement

M SD IQD
Client brief/ input 8.50 8.50 1.08 1.50
Client complaint information/feedback 8.70 9.00 0.82 1.00
Market investigation 7.90 8.00 0.32 0.00
Client satisfaction survey 8.70 9.00 0.82 1.00
Quality warranty 8.00 8.00 1.15 2.00
Client information system 7.50 7.00 0.97 1.25
Client services 7.30 7.00 0.95 0.50
Client cooperation 8.00 8.00 0.82 2.00
Client orientation 7.00 7.00 0.82 0.25
M=Median; =Mean; =Standard Deviation (SD); IQD=Interquartile Deviation

Experts reached consensus on all twelve attributes identified under the fourth sub-attribute
(Company Quality System Evaluation) with IQD cut-off (IQD ≤1) score set to achieve
consensus (Table 9.9). Eight (8) attributes had strong consensus and other four had good
consensus (Table 9.9). The eight (8) attributes had a very high impact (VHI: 9-10), while
the remaining four (4) of the attributes had a high impact (HI: 7-8.99).

218
Table 9.9: Company Quality System Evaluation

Company Quality System Evaluation M SD IQD

Evaluation of strategy 8.20 8.00 0.79 0.25
Evaluation of overall company
8.70 9.00 0.67 0.25
performance
Evaluation of departments’ performance 8.50 9.00 0.71 1.00
Evaluation of employees’ performance 7.90 8.00 0.57 0.25
Evaluation of quality costs 7.90 8.00 0.57 0.25
Evaluation of quality manual 8.20 8.00 0.63 1.00
Evaluation of quality control system 8.70 9.00 0.67 0.25
Evaluation of quality system procedures 8.70 9.00 0.82 1.00
Evaluation of end results 8.60 9.00 0.70 1.00
Quality audit 9.00 9.00 0.47 0.00
Quality benchmarking 8.80 9.00 0.42 0.25
Quality information system 8.70 9.00 0.48 1.00
M=Median; =Mean; =Standard Deviation (SD);; IQD=Interquartile Deviation

Experts reached consensus on five (5) attributes from the fifth sub-attribute (Company
Vision and Plan Statement) identified with IQD cut-off (IQD ≤1) score set to achieve
consensus (Table 9.10). One (1) of the nine (9) attributes had strong consensus, while the
remaining eight (8) had good consensus (Table 9.10). One (1) of the nine (9) attributes
(Quality Improvement Plan) had a very high impact (VHI: 9-10), while the remaining eight
(8) had a high impact (HI: 7-8.99). These representations are given in Table 9.10.

Table 9.10: Company Vision and Plan Statement

Company Vision and Plan Statement M SD MAD IQD

Vision statement 7.50 8.00 0.97 0.70 1.25
Quality policy 8.30 8.50 1.06 0.90 2.00
Overall business performance plan 7.70 8.00 0.82 0.50 1.00
Product quality goal 8.10 8.00 0.88 0.50 1.00
Quality improvement plan 8.80 9.00 0.63 0.20 0.00
Formulation of vision and plan 7.60 8.00 1.07 0.80 1.50
Concreteness of future plan 7.40 8.00 1.35 1.00 2.25
Employees’ contribution to the vision 7.80 8.00 0.63 0.20 0.00
Involvement of employees in the
7.50 8.00 0.97 0.50 1.00
development of vision statement
M=Median; =Mean; =Standard Deviation (SD); IQD=Interquartile Deviation

219
From the fifteen (15) attributes under the sixth sub-attribute (Product Selection and Design
Management) identified, the experts reached consensus on eleven attributes with IQD cut-
off (IQD ≤1) score set to achieve consensus (Table 9.11). Seven (7) of the attributes were
considered by the experts to have a very high impact (VHI: 9-10) under the median score.
The remaining eight (8) attributes were considered by the experts to have a high impact (HI:
7-8.99) on TQM in the construction industry (Table 9.11).

Table 9.11: Product Selection and Design Management

Product Selection and Design

M SD IQD
Management
Client brief/input 8.90 9.00 0.88 0.50
Cost implication 8.00 8.00 0.82 0.25
Environmental issues 6.56 7.00 2.07 2.50
Socio cultural issues 7.40 8.00 2.37 0.25
Appearance/finishes required 8.40 9.00 1.17 1.25
Strength required 8.20 9.00 1.69 1.25
Intended purpose of the material 7.60 9.00 2.55 2.25
Where the material to be used 7.20 7.00 0.42 0.25
Design methods/techniques 8.10 8.00 0.99 0.25
Design detailing 8.60 9.00 0.84 1.00
Concurrent engineering 7.90 8.00 0.99 0.25
Design experiments/testing 7.90 8.00 0.99 0.25
Quality function deployment 8.70 9.00 1.06 0.25
Value engineering 8.70 9.00 1.06 0.25
Computer-aided design 7.00 7.00 1.49 0.50
M=Median; =Mean; =Standard Deviation (SD); IQD=Interquartile Deviation

From the fifteen (15) attributes under the seventh sub-attribute (Construction Process
Management and Improvement) identified, the experts reached consensus on twelve
attributes with IQD cut-off (IQD ≤1) score set to achieve consensus (Table 9.12). Twelve
(12) attributes had strong consensus, with the remaining three (3) having good consensus.
Four (4) of the attributes were considered by the experts to have a very high impact (VHI:
9-10) under the median score. The remaining eleven (11) of the attributes were considered
by the experts to have a high impact (HI: 7-8.99) on TQM in the construction industry
(Table 9.12).

220
Table 9.12: Construction Process Management and Improvement

Construction Process Management

M SD IQD
and Improvement
Project monitoring and control
8.90 9.00 0.74 0.00
improvement
Equipment maintenance and innovation 8.10 8.00 0.88 0.50
Inventory management 7.30 7.00 0.67 0.25
Use of quality materials 8.70 9.00 1.06 1.50
Use of quality tools 8.00 8.00 1.05 0.50
Use of quality manual 8.00 8.00 0.94 0.00
Understanding of work
8.70 9.00 1.16 1.25
instructions/quality manual
Quality system procedures improvement 8.20 8.00 0.79 0.25
Obtaining ISO 9000 certification for
8.00 8.00 0.82 0.25
operation
Understanding of quality control system 8.20 8.00 0.63 0.00
Continuity of quality control circle
8.60 9.00 0.84 1.00
activities
Appropriate use of system structure and
8.00 8.00 0.94 0.00
standards
Targets and priority measures 7.80 8.00 0.79 0.25
Safe working environment 8.30 8.50 1.16 1.25
Utilization of analysis results 7.90 8.00 0.74 0.00
M=Median; =Mean; =Standard Deviation (SD); IQD=Interquartile Deviation

From the fourteen (14) attributes under the eighth sub-attribute (Construction Employees
Involvement and Motivation) identified, the experts reached consensus on thirteen attributes
with IQD cut-off (IQD ≤1) score set to achieve consensus (Table 9.13). Thirteen (12)
attributes had strong consensus, with the remaining one (1) having good consensus. Only
one (1) of the attributes was considered by the experts to have a very high impact (VHI: 9-
10) under the median score. The remaining thirteen (13) of the attributes were considered
by the experts to have a high impact (HI: 7-8.99) on TQM in the construction industry
(Table 9.13).

221
Table 9.13: Construction Employees’ Involvement and Motivation

Construction Employees Involvement

M SD IQD
and Motivation
Education and training 8.70 9.00 0.48 1.00
Salary promotion 7.70 8.00 0.67 0.25
Position promotion 7.50 8.00 0.97 1.25
Employee rewards 7.60 8.00 0.70 1.00
Bonus scheme 7.80 8.00 0.79 0.25
Conducive working environment 7.89 8.00 0.78 0.00
Involvement of employees within-
8.10 8.00 0.57 0.25
functional team
Employee participation 8.20 8.00 0.42 0.25
Employees’ suggestion 8.10 8.00 0.32 0.00
Improving employee commitment 8.20 8.00 0.42 0.25
Job rotation 7.90 8.00 0.88 0.50
Involvement in quality control circle 7.90 8.00 0.74 0.00
Employee involvement in cross-
7.80 8.00 0.42 0.25
functional team
Employee recognition 8.10 8.00 0.32 0.00
M=Median; =Mean; =Standard Deviation (SD); IQD=Interquartile Deviation

DSO 2: To determine the factors that enable construction companies to implement

TQM

From the twenty-two (22) identified factors that were considered to enable construction
companies to implement TQM, the experts reached consensus on sixteen factors with an
IQD cut-off (IQD ≤1) score on TQM in the construction industry (Table 9.14). Eight (8) of
the factors had a very high impact (VHI: 9-10), while the remaining fourteen (14) of the
factors had a high impact (HI: 7-8.99) on TQM in the construction industry (Table 9.14).

Table 9.14: Factors that Enable Construction Companies to Implement TQM

Factors that enable Construction

Companies to Implement TQM M SD IQR

Top management support 9.10 9.00 0.74 1.25

Good quality driven agenda 8.90 9.00 0.57 0.25
Commitment to people 8.30 8.00 0.48 1.00
Training and re-training of staff 8.70 9.00 0.67 0.25
Institution of continuous improvement 8.60 9.00 0.70 1.00
222
 process within the company
Systematic TQM framework 8.20 8.00 0.42 0.25
Efficient management system 8.50 9.00 0.71 1.00
Proper understanding of TQM among
8.80 9.00 0.79 0.25
construction professionals
Availability of resources 7.50 8.00 1.51 1.00
Good communication in organisation 7.60 8.00 1.26 1.50
Good quality policy 7.90 8.00 0.74 0.00
Changes in company structure 7.40 8.00 1.26 1.00
Good organizational and quality culture 8.60 9.00 0.70 1.00
Teamwork 8.00 8.00 0.47 0.00
Availability of quality information 8.10 8.00 0.57 0.25
Employee involvement 7.70 8.00 0.67 1.00
Good climatic condition 6.80 7.00 0.42 0.25
Good socio-economic environment 6.50 7.00 1.90 1.50
Proper project conceptualization 6.60 7.00 1.71 0.50
Project manager competence 7.20 8.00 1.62 1.25
Clarity of goals and objectives 8.00 9.00 1.49 2.25
Adequate TQM experts 7.50 8.00 0.85 1.25
M=Median; =Mean; =Standard Deviation (SD); IQR=Interquartile Deviation

DSO 3: To Identify the Factors that Affect Construction Companies for not
Implementing TQM in Ghana
From the twenty (20) identified factors that were considered to affect construction
companies in Ghana for not implementing TQM, the experts reached consensus on only
nine factors with an IQD cut-off (IQD ≤1) score on TQM in construction (Table 9.15).
Eleven (11) of the factors had a very high impact (VHI: 9-10) on TQM in the construction
industry. Nine (9) factor had a medium impact (HI: 7-8.99) on TQM in the construction
industry (Table 9.15).

Table 9.15: Factors that Affect Construction Companies for not Implementing TQM
in Ghana

Factors that Affect Construction

M SD IQR
Companies if not Implementing TQM
Lack of efficient TQM management
8.50 9.00 1.08 2.00
system
Lack of or limited company resources 7.30 8.00 2.06 1.00
Unavailable TQM policy 8.50 9.00 1.08 2.00
223
 Lack of or limited knowledge of TQM 8.50 9.00 0.85 1.25
Inability to employ TQM personnel 7.20 8.00 1.87 1.00
Absent of systematic TQM framework 8.10 8.00 0.57 0.25
Lack of and inability to train and educate
8.30 9.00 1.70 1.25
employees on TQM
Lack of understanding among
construction professionals in applying 8.40 9.00 0.84 1.25
TQM
Lack of coordination of the
implementation of TQM policy within 8.40 9.00 1.17 1.25
the organisation
Limited access to body responsible for
7.40 8.00 2.32 0.50
the implementation of TQM policy
Lack of TQM expert 8.00 9.00 1.89 1.25
Reluctance to change old management
8.90 9.00 0.57 0.25
technique
Complex nature of TQM technique 7.50 8.00 2.37 1.25
Lack of commitment from management 9.22 9.00 0.67 1.00
Perception that TQM may not yield any
7.10 8.00 2.33 1.50
better results
TQM technique is time consuming 8.00 8.00 1.05 0.50
TQM technique is costly 8.10 8.00 0.74 1.25
Lack of finance in the management of
8.10 8.00 0.57 0.25
TQM experts
Lack of enforcement from the legislative
bodies overseeing the implementation of 7.90 9.00 2.18 1.25
TQM
Lack of interest in the application of
8.80 9.00 1.14 0.50
TQM
M=Median; =Mean; =Standard Deviation (SD); IQR=Interquartile Deviation

DSO 4: To evaluate organizational factors and issues that affects the implementation of
TQM

From the twenty (20) organizational factors and issues that affect the implementation of
TQM, the experts reached consensus on twelve factors with IQD cut-off (IQD ≤1) score on
TQM implementation in the construction industry (Table 9.16). Six (6) of the factors had a
very high impact (VHI: 9-10) on TQM implementation in the construction industry. The
remaining fourteen (14) factors had a high impact (HI: 7-8.99) on TQM implementation in
the construction industry (Table 9.16).

224
Table 9.16: Organizational Factors that Affect Construction Companies in the
Implementation of TQM

Organizational Factors that Affect

Construction Companies in the M SD IQR
Implementation of TQM

Leadership and top management

9.40 9.00 0.52 1.00
commitment
Top management support 9.30 9.00 0.48 1.00
Project manager competence 8.80 9.00 0.79 0.25
Quality policy 8.90 9.00 0.88 0.50
Supplier management 8.00 8.00 0.82 0.25
Limited cash flow to manage TQM 7.70 8.00 1.64 0.75
Change in management
7.90 8.00 1.73 1.50
behaviour/attitude
Change in status quo 7.80 8.00 1.69 1.50
Personnel to manage and monitor TQM
7.50 8.00 1.58 1.25
application
Education and training 9.00 9.00 0.94 1.25
Employee relations 8.20 8.00 0.92 1.25
Employee involvement 8.78 9.00 0.83 0.50
Lack of interest in the application of
8.30 8.00 1.25 1.50
TQM
Client involvement 8.30 8.00 0.82 1.00
Teamwork 8.50 8.00 1.08 2.00
Feedback by project participants 8.50 8.00 0.71 1.00
Design quality management 8.30 8.00 0.82 1.00
Strategic quality management 8.30 8.00 0.82 1.00
Quality data and reporting 8.40 8.00 0.84 1.00
Analysis of quality information 8.50 8.00 0.97 1.25
M=Median; =Mean; =Standard Deviation (SD); IQR=Interquartile Deviation

DSO5. To identify the effects of implementation of TQM in the construction industry

From the seventeen (17) identified effects of implementation of TQM in the construction
industry, the experts reached consensus on fourteen of the effects with IQD cut-off (IQD
≤1) score on TQM in the construction industry (Table 9.17). Eleven (11) of the effects had
a very high impact (VHI: 9-10) on TQM in the construction industry. The remaining six (6)
effects had a high impact (HI: 7-8.99) on TQM in the construction industry (Table 9.17).

225
Table 9.17: Effects of Implementing TQM in the Construction Industry

Effects of Implementing TQM in the

M SD MAD IQR
Construction Industry
Reduction in construction costs 9.20 9.00 0.63 0.40 1.00
Higher client satisfaction 9.20 9.00 0.63 0.40 1.00
Improve employee job satisfaction 9.10 9.00 0.57 0.30 0.25
Improve schedule performance 8.60 9.00 0.84 0.60 1.00
Elimination/Reduction of reworks 9.80 10.00 0.42 0.20 0.25
Defect-free product at first attempt 8.90 9.00 0.74 0.30 0.00
Improve safety 8.40 8.00 0.70 0.40 1.00
Higher productivity 9.00 9.00 0.47 0.20 0.00
Lower employee turnover 8.30 8.00 0.95 0.70 1.25
Speed up construction work 8.30 8.00 0.82 0.50 1.00
Improved relationships with
8.20 8.00 0.79 0.40 0.25
subcontractors
Improved methods of working 8.20 8.00 0.79 0.40 0.25
Better control over the construction
8.70 9.00 0.82 0.50 1.00
process
Increased profitability 9.00 9.00 0.82 0.60 2.00
Gaining competitive advantage over
9.30 9.00 0.67 0.50 1.00
other companies
Decreasing waste 8.70 9.00 0.95 0.70 1.25
Better coordination of activities 8.20 8.00 0.79 0.40 0.25
M=Median; =Mean; =Standard Deviation (SD); IQR=Interquartile Deviation

9.5 DISCUSSIONS FROM THE DELPHI STUDY RESULTS

DSO 1: To identify the main and sub-attributes that determine TQM implementation in
Ghanaian construction industry.
A set of main attributes and sub-attributes that are relevant to TQM was emphasised
through a comprehensive review of literature. Although the reviewed literature was based
on studies from the developed countries, they were collectively used to examine the
attributes that determine TQM in the Ghanaian construction industry. The influence of the
attributes on TQM was obtained as a product of the impact on the construction industry in
Ghana. The main attributes were based on the level of influence as categorized on the
questionnaire. This was established by assessing the extent to which the listed attributes
will determine the TQM implementation in Ghanaian construction industry. The impact of
the sub-attributes in determining TQM was assessed if they were present or lacking. The
226
rating was based on an ordinal scale of one to ten with one being ‘low influence’ or ‘no
impact’ and ten being ‘high influence’ or ‘very high impact’. The levels of influence and
impact were obtained as a product of the consensus achieved.

The first objective of the Delphi study was to identify the main and sub-attributes that
determine TQM in the Ghanaian construction industry. Findings emanating from the survey
revealed that the attributes that bring about TQM in the Ghanaian construction industry are
similar to those in other cultural contexts. In determining whether an attribute reached
consensus or not, consensus was reached on four (4) attributes (Leadership and Top
Management, Company Quality System Evaluation, Product Selection and Design
Management, Construction Process Management and Improvement) under the main
attributes that determine TQM in the Ghanaian construction industry, with an IQD score of
1.00. Six (6) of the attributes had strong consensus, with the exception of Client Focus and
Involvement, and Product Selection and Design Management, which had good consensus.
Seven (7) attributes from the eight (8) identified main attributes that determine TQM in the
construction industry were considered by the experts to have a very high influence (VHI: 9-
10), with the exception of one (1) attribute (Product Selection and Design Management)
which had a high influence (HI: 7-8.99). These indicate that all the attributes have high
levels of influence on TQM implementation in the Ghanaian construction industry.

The assessment of the sub-attributes of eight (8) major determinants of TQM in the
construction industry showed that the sub-attributes that determine TQM in the Ghanaian
construction industry are identical to other cultural contexts. From the ninety-eight (98)
sub-attributes identified under the (8) major determinants, only thirty-six (36), were found
to be strong determinants of TQM in the construction industry in Ghana. Among these sub-
attributes which were determined by the experts to have reached strong consensus are
Leadership style, Leadership ability in solving quality problem, Leadership initiatives,
Leadership and top management knowledge in TQM, Top management commitment in
TQM, Top management interaction with workers, Top management participation, Top
management encouragement, Supplier performance Evaluation, Supplier quality audit,
Supplier Communication, Supplier knowledge in TQM, Cooperation from suppliers, Client
227
complaint information/feedback, Client satisfaction survey, Evaluation of overall company
performance, Evaluation of departments’ performance Evaluation of quality control system,
Evaluation of quality system procedures, Evaluation of end results, Quality audit, Quality
benchmarking, and Quality information systems. Other sub-attributes which were
considered by the experts to be strong determinants of total quality in the construction
industry are Quality improvement plan, Client brief/input, Appearance/finishes required,
Strength required, Intended purpose of the material, Design detailing, Quality function
Deployment, Value engineering, Project monitoring and control improvement, Use of
quality materials, Understanding of work instructions/quality manual, Continuity of quality
control circle activities, and Education and training.

The assessment of the sub-attributes findings replicated the results of the majority of studies
on TQM in the construction industry in relation to adequate Training and education,
Supplier quality management and in general Leadership and top management commitment
(Saraph et al., 1989; Flynn et al., 1994; Ahire et al., 1996; Imbeah & Dansoh framework,
2011; Adusa-Poku framework, 2014) upon which the current study’s sub-attributes were
also based. Leadership style, Leadership ability in solving quality problem, Leadership
initiatives, Leadership and top management knowledge in TQM, Top management
commitment in TQM, Top management interaction with workers, Top management
participation, Top management encouragement were rated as variables with a very high
impact as these impact on the quality of leadership and top management. Project monitoring
and control improvement, Use of quality materials, Understanding of work
instructions/quality manual, and Continuity of quality control circle activities also have a
very high impact in relation to Construction Process Management and Improvement. These
findings agree with previous findings from the studies done by Saraph et al. (1989), Flynn
et al. (1994), and Ahire et al. (1996). Further findings as related to Supplier performance
Evaluation, Supplier quality audit, Supplier Communication, Supplier knowledge in TQM,
Cooperation from suppliers, Client complaint information/feedback, Client satisfaction
survey, Evaluation of overall company performance, Evaluation of departments’
performance Evaluation of quality control system, Evaluation of quality system procedures,
Evaluation of end results, Quality audit, Quality benchmarking, Quality information
228
system, Quality improvement plan, Client brief/input, Appearance/finishes required,
Strength required, Intended purpose of the material, Design detailing, Quality function
Deployment, and Value engineering imply that the sub-variables play a vital role in TQM in
the construction industry.

In conclusion, the results seem to suggest that the attributes that bring about TQM in the
Ghanaian construction industry are similar to the determinants in other cultural contexts.
Furthermore, TQM in the construction industry is assured if more attention is given to these
attributes in the development of an integrated TQM model for the construction industry in
Ghana. Particular attention should be given to the thirty-six (36) sub-attributes determined
by the experts who have all been described as having a significant influence and very high
impact on determining TQM in the Ghanaian construction industry.

Objective DSO2
The second objective of the Delphi study was to determine the factors that enable
construction ompanies to implement TQM in the construction industry. The assessment of
the twenty two (22) factors of major determinants of TQM in the construction industry
showed that the experts reached consensus on sixteen (16) factors with an IQD score
between 0.00 and one (1). Eight (8) of the enable factors (Top management support, Good
quality driven agenda, Training and re-training of staff, Institution of continuous
improvement process within the company, Efficient management system, Proper
understanding of TQM among construction professionals, Good organizational and quality
culture, Clarity of goals and objectives) had a very high impact (VHI: 9-10) on TQM in the
construction industry. The remaining fourteen (14) of the enable factors had a high impact
(HI: 7-8.99) on TQM in the construction industry. These indicate that all the enabling
factors have a high level of influence on TQM implementation in the Ghanaian construction
industry.

Objective DSO3
The third objective of the Delphi study was to identify the factors that affect construction
companies for not implementing TQM in Ghana. Construction companies’ inability to
229
implement TQM in the Ghanaian construction industry might be due to several factors,
some of which may be beyond their control and others out of their reach owing to limited
resources. The assessment of the twenty (20) factors of major determinants of TQM in the
construction industry showed that experts reached consensus on nine (9) factors with IQD
less or equal to one (1). Eleven (11) of the factors (Lack of efficient TQM management
system, Unavailable TQM policy, Lack of or limited knowledge of TQM, Lack of and
inability to train and educate employees on TQM, Lack of understanding among
construction professionals in applying TQM, Lack of coordination of the implementation of
TQM policy within the organisation, Lack of TQM experts, Reluctance to change old
management technique, Lack of commitment from management, Lack of enforcement from
the legislative bodies overseeing the implementation of TQM, and Lack of interest in the
application of TQM) had a very high impact (VHI: 9-10) on TQM in the construction
industry. The remaining nine (9) of the factors had a high impact (HI: 7-8.99) on TQM in
the construction industry. These indicate that most of the factors have a very high level of
influence on construction companies in Ghana which do not implement TQM.

Objective DSO4
The fourth objective of the Delphi study was to evaluate the organizational factors and
issues that affect the implementation of TQM in the construction industry. The assessment
of the twenty (20) organizational factors and issues as considered by the experts to be the
major determinants of TQM in the construction industry showed that the experts reached
consensus on twelve (12) factors (Leadership and top management commitment, Top
management support, Project manager competence, Quality policy, Supplier management,
Limited cash flow to manage TQM, Employee involvement, Client involvement, Feedback
by project participants, Design quality management, Strategic quality management, Quality
data and reporting) were considered by the experts to have achieved consensus with IQD
less or equal to one (1). Six (6) of the organizational factors and issues (Leadership and top
management commitment, Top management support, Project manager competence, Quality
policy, Education and training, Employee involvement) to have a very high influence (VHI:
9-10) on TQM in the construction industry. The experts reached consensus on the
remaining fourteen (14) of the organizational factors and issues as having a high influence
230
(HI: 7-8.99) on TQM in the construction industry. Their SD ranges from 0.48-1.73 and their
mean ranges from 7.50 – 9.40. These indicate how significant organizational factors and
issues are regarding the TQM implementation in the construction industry.

Objective DSO5
The fifth objective of the Delphi study was to identify the effects of implementation of
TQM in the construction industry. The assessment of the seventeen (17) major effects of the
implementation of TQM in the construction industry showed that the experts reached
consensus on fourteen (14) of the effects with an IQD less or equal to one (1). Eleven (11)
of the effects (Elimination/Reduction of reworks, Reduction in construction costs, Higher
client satisfaction, Improve employee job satisfaction, Improve schedule performance,
Defect-free product at first attempt, Higher productivity, Better control over the
construction process, Increase profitability, Gaining competitive advantage over other
companies, Decreasing waste) had a very high influence (VHI: 9-10) on TQM in the
construction industry, while the remaining six (6) of the effects had a high influence (HI: 7-
8.99). Their SD ranged from 0.42-0.95 and their mean ranged from 8.20 – 9.30. These
indicate how important and influential these effects are on the TQM implementation in the
construction industry.

From all of the above, a number of factors that were considered to be important in
determining TQM in the Ghanaian construction industry have been identified and
amplified by the Delphi study. The factors considered to be paramount determinants of
TQM in the Ghanaian construction industry include Leadership and Top Management,
Company Supplier Quality Management, Client Focus and Involvement, Company Quality
System Evaluation, Company Vision and Plan Statement, Product Selection and Design
Management, Construction Process Management and Improvement, and Construction
Employees’ Involvement and Motivation. These factors have been collectively considered
for the development of a holistic TQM model in this study. Five of the factors have been
previously considered in the development of TQM model in other cultural contexts, but
none of the existing models to date have included Vision and Plan Statement, Product
selection and Design Management, and Company Quality System Evaluation as inclusive
231
factors to develop a model to guide professionals in the construction industry in making
decisions about the criteria to be given priority in TQM issues.

9.6 CONCLUSION

This chapter presented a summary of results and discussions of the results of both the
Delphi rounds (first to the second round). Computation for each question element was made
for the influence and impact of the attributes in determining TQM in the construction
industry and how this will contribute to the improvement of project performance in the
construction industry in Ghana. The influence or impact of the absence or presence of a
particular TQM element on the overall TQM of the other elements was presented. The
chapter concluded with a summative discussion of the findings based on the objectives of
the Delphi study. The findings from the expert participants revealed a consistent discussion
on TQM in Ghanaian construction industry with consensus being reached in most cases and
in others with a discrete conclusion. The result of the Delphi study assisted in the
determination of key factors and constructs that are of critical significance (influence) to
determine TQM in the construction industry. The outcome led to the development of the
holistically integrated TQM conceptual model for the construction industry in Ghana.
Chapter 10 provides the evaluation of these factors and their interrelationships.

232
 CHAPTER TEN
10 THE CONCEPTUAL INTEGRATED TOTAL QUALITY MANAGEMENT
MODEL
10.1 INTRODUCTION
This chapter presents a discussion of the findings from the review of the literature. The
conceptual model theory forms the basis of the discussion. Based on an in-depth review of
the previous models as presented in Chapter 2 of this thesis, the hypothesised integrated
holistic TQM model is presented in this chapter. The five constructs (Leadership and Top
Management, Company Supplier Quality Management, Client Focus and Involvement,
Construction Process Management and Improvement, and Construction Employee
Involvement and Motivation) that influence the implementation of TQM in the construction
industry are explained in detail and presented in this chapter. The additional three
constructs (Company Vision and Plan Statement, Product Selection and Design
Management, and Company Quality System Evaluation) which were identified as gaps in
TQM research have already been explained and discussed in detail in Chapter 3 of this
thesis. Detailed discussions of the holistic integrated model and the variables of the model,
the identification of the model and justification for the selected variables are all given in
this chapter.

10.2 SELECTION OF VARIABLES FOR TQM IMPLEMENTATION

Both objective and subjective attributes have been combined in the TQM study models. For
the assessment of TQM implementation, the research by Sila and Ebrahimpour (2002)
identified twenty-five TQM elements which are commonly used by researchers to measure
TQM implementation. Their research discovered eight common cores of the factors, namely
leadership and top management commitment, employee training, customer focus and
satisfaction, teamwork, employee involvement, continuous improvement and innovation,
and quality information and performance. Literature also discloses that different countries
have adopted related TQM constructs as criteria for quality awards under different titles
(Metri, 2005).

233
The selection of the variables for this study is anchored around system theory in that it takes
all the organizations’ systems for successful implementation of TQM and the organization
performance is largely dependent on its ability to continuously improve on the management
of its systems. Also system theory is based on the principle that each organization is
composed of a system of interrelated processes and people which make up the system’s
components. The system theory was updated by the Deming theory (Deming, 1986).
Hackman and Wageman (1995) systematically reviewed Deming’s theory and proposition
relating to TQM. Their review indicated the following five interventions as the core of
TQM: Explicit identification and measurement of customer needs and wants, the creation of
supplier partnership, the use of functional teams to identify and resolve quality problems,
the use of scientific methods to monitor performance and identify points of high leverage
for performance improvement, and the use of process management heuristics to improve
team effectiveness.

Comparing the conceptual frameworks of Saraph, Benson and Schroeder (1989), Flynn,
Schroeder and Sakakibara (1994), Ahire, Golhar and Waller (1996); Adusa–Poku, (2014)
and that of quality award models i.e the European Quality Award Model (1994), the
Deming Prize Quality Award model (1996), and the Malcolm Baldrige National Quality
Award model (1999), it was found that most of the research findings relating to TQM
implementation shared a similar view with the Deming theory of profound knowledge
which is a management philosophy grounded in systems theory (Deming, 1986).

These frameworks and the three quality award models helped to identify a range of
intangible and tangible processes that influence the firm’s TQM implementation and the
end results. It was obvious that there are some similarities and dissimilarities in their
conceptual frameworks. This study integrated the identified TQM constructs as much as
possible and came up with the following five constructs: Leadership and Top Management,
Company Supplier Quality Management, Client Focus and Involvement, Construction
Process Management and Improvement, and Construction Employee Involvement and
Motivation.

234
Hence, the existing TQM model contains leadership and top management with twelve
variables; Company Supplier Quality Management with twelve variables; Client Focus and
Involvement with nine variables; Construction Process Management and Improvement with
fifteen variables; and Construction Employee Involvement and Motivation with fourteen
variables.

Almost all the TQM studies have these constructs conceptualized on a frequent basis.
However, the current study brings into focus company vision and plan statement which has
nine variables, product selection and design management which has fifteen variables, and
company quality system evaluation which has twelve variables. These three additional
constructs were not found in the identified model and frameworks. The gaps identified in
the literature review are these three additions and were found to be peculiar to both
developed and developing countries.

The five constructs that influence the implementation of TQM in the construction industry
are explained in detail and presented in this chapter. The additional three constructs have
been presented and discussed in detail under a section in Chapter 3 (observed gaps in the
TQM litereature).

10.2.1 Leadership and Top Management (LTM)

Leadership is the capability to inspire confidence and support among those needed to
accomplish organizational objectives (DuBrin, 1995). Anderson et al. (1994a) explained the
concept of leadership as the ability of top management to establish, practise, and lead a
long-term vision for the company, driven by changing customer requirements as opposed to
an internal management control role. Leadership is therefore exemplified by clarity of
vision, long-term orientation, coaching management style, participative change, employee
empowerment, and planning and implementing organizational change. Juran and Gryna
(1993) mentioned that certain roles of top management can be recognized as establishing
quality policies, establishing and deploying quality objectives, providing resources,
providing problem-oriented training, and stimulating improvement. The European Quality
235
Award (1994) and the Malcolm Baldrige Quality Award (1999), on the other hand,
recognize the vital role of leadership in creating the goals, values and systems that guide the
pursuit of continuous performance improvement. Recognition of the crucial role of
leadership and its responsibility in quest of continuous quality improvement supports the
arguments put forward by quality gurus such as Deming (1986), Juran (Juran & Gryna,
1993), and Crosby (1979). Hence, the concept of leadership in this study can be defined as
the ability of top management to lead the company in continuously pursuing long-term
overall business success. This is demonstrated by top management participation, top
management encouragement, employee empowerment, top management learning, top
management commitment to employee education and training, and top management pursuit
of product quality and long-term business success.

A predominant theme in quality management literature is that strong commitment from top
management is very important. Indicating such commitment is therefore a prime leadership
principle for realizing TQM. Brown, Pearl and Akintoye (1994) stated that a lack of top
management commitment is one of the causes for the failure of TQM initiatives. However,
top management commitment itself is not enough. It is more important that top
management personally participates in various quality management activities. Furthermore,
it should strongly encourage employee involvement in quality management activities.
DuBrin (1995) stressed that an important leadership practice is to encourage people to
assess the level of quality.

To become a successful leader in most modern firms, the top manager must continue to
develop and learn. Knowledge of the business and constant learning are important
requirements of efficient leadership (DuBrin, 1995). In order to lead the company
effectively, top management must encourage workers to fix the problems they face. Thus,
workers must have the ability to solve problems and prevent their further occurrence. Also
in order to effectively lead the company, top management must be dedicated to providing
adequate resources for employees’ education and training, building trustworthy
relationships with employees, and regarding them as precious resources of the company.
Top management must be dedicated to assigning adequate resources to prevent, as well as
236
repair, quality problems. Top management should discuss quality frequently, for example,
by giving speeches on the topic and asking questions about quality at every staff meeting.
In fact, people make things happen. Therefore, top management must train and coach
workers to assess, analyze, and improve work processes (Dale & Plunkett, 1990; Deming,
1986).

The research study carried out by Garvin (1986) recommended that high levels of quality
performance were always accompanied by an organizational dedication to that goal; high
product quality did not exist without strong top management commitment. Many such
scientific studies have also found that top management support for quality was a key factor
in quality enhancement. Hence, it is important that top management should focus on
product quality rather than yields only. More significantly, it is essential for the company to
engage long-term business success. Seeking short-term business success places quality
behind yield, costs, and meeting delivery schedules (Zhang, 2000b). This present study
looks at leadership/management that has been hypothesised for the development of a
holistic TQM model and is summarised in Table 10.1.

10.2.2 Company Supplier Quality Management (CSQM)

Supplier quality management can be defined as the set of supplier-related quality
management practices for improving suppliers’ quality of products and services. This is
exemplified by firm-supplier partnerships, product quality as the criterion for supplier
selection, participation in suppliers, communication with suppliers, understanding of
supplier performance, and supplier quality audits (Mann, 1992; Zhang, 2000a).

In modern industrial production, the interdependence of buyers and suppliers has increased
considerably. The supplier becomes an extension of the buyer’s organization to a certain
extent. A revolution in the link between customers and providers appeared in the form of
supplier collaboration (Juran & Gryna, 1993). Hackman and Wageman (1995) stated that
building relationships with suppliers is one of the main TQM implementation practices.
Anderson et al. (1994a) also pointed out that external collaboration between a company and
its suppliers has benefits in the just-in-time buying systems. Working collaboratively with
237
suppliers on a long-term basis is helpful. Deming (1986) strongly suggested working with
the supplier as a partner in a long-term relationship of loyalty and trust to enhance the
quality of incoming materials and decrease costs. A long-term relationship between buyer
and supplier is essential for the best economy.

Deming (1986) and Ishikawa (1985) recommended that companies choose their suppliers
on the basis of quality, rather than solely on price. Deming (1986) added that price has no
meaning without a measure of the quality being bought. Without sufficient measures of
quality, business drifts to the lowest bidder, low quality and high cost being the inevitable
outcomes. The company therefore must change its focus from lowest initial cost of material
bought to lowest total cost. Companies should take part directly in suppliers’ activities
related to quality (Mann, 1992; Zhang, 2000a). Deming (1986) and Ishikawa (1985)
therefore suggested that companies work directly with suppliers to ensure that their
materials are of the highest possible quality. Incoming material control is extremely
important for supplier quality management. Specifications and standards should be
recognized as requirements for approval of raw materials, parts, and components.
Techniques such as acceptance sampling inspection and 100% inspection can be used to
provide acceptance at most economical levels (Feigenbaum, 1991).

Evaluating suppliers is an essential activity to ensure areliable high quality of incoming

materials in the company (Feigenbaum, 1991). Each supplier is measured against another
specific supplier or group of suppliers for price, quality, delivery, and other important
performance measures. Supplier performance evaluation includes objective assessment of
one supplier’s performance, which can feed back to that supplier. In the case of a poor
quality situation, such information can be used by the supplier to come up with corrective
action. Supplier quality evaluation also provides a quantitative summary of supplier quality
over a period of time (Juran & Gryna, 1993).

Companies need to have detailed information about supplier quality information such as
drawings, specifications, and other necessary information. It is also essential to establish a

238
supplier information feedback system which can be used for giving feedback to suppliers
about their products (Feigenbaum, 1991; Juran & Gryna, 1993).

Supplier quality audit, an organized evaluation of supplier abilities to provide materials of

the necessary quality and quantity, is an essential basis for initial supplier selection and
ongoing supplier quality surveillance (Feigenbaum, 1991). Surveillance can take several
forms: Inspection of product, meetings with suppliers to review quality status, audits of
elements of the supplier, review of SPC data, and witnessing of specific operations or tests
(Juran & Gryna, 1993). The supplier quality management that is considered for the present
study is summarised in Table 10.1.

10.2.3 Client Focus and Involvement (CFI)

Client focus can be defined as the extent to which a company continuously satisfies client
or customer desires and expectations. A successful company recognizes the need to put the
customer first in every decision taken (Philips Quality, 1995). The key to quality
management is keeping a close association with the customer in order to fully establish the
customer’s needs, as well as to receive feedback on the degree to which those needs are
being met. The customer should take part in the product design and development process;
with input at every point so that there is less probability of quality problems once full
production begins (Flynn et al., 1994). Deming (1986) recommended that the customer is
the most essential part of the production line and that the product should be aimed at the
needs of the customer.

Receiving customer complaint information is to seek opportunities to improve product and

service quality. Quality complaints have different problems that require different actions.
Based on customer complaint information, it is important to identify the “vital few” serious
complaints that demand in-depth study in order to discover the basic causes and to remedy
those causes (Juran & Gryna, 1993). To improve customer focus efforts, customer
complaints should therefore be treated with top priority. Records and analyses of customer
complaint reports from the field furnish useful product-control information. Such
information reflects the effectiveness of control programmes and highlights those
239
nonconformities for which more aggressive corrective action must be initiated
(Feigenbaum, 1991).

Obtaining customer satisfaction information is essential for pursuing customer focus efforts.
Intensive examination of finished products from the viewpoint of the customer can be a
useful predictor of customer satisfaction. Such information includes data on field failures
and service-call rates, and analysis and reporting of customer attitude trends regarding
product quality. Such information is valuable for new product development (Feigenbaum,
1991). The results of customer satisfaction surveys can be used to take immediate action on
customer complaints, identify problems requiring generic corrective action, and provide a
quantitative measurement of customer satisfaction (Juran & Gryna, 1993). Customer
satisfaction may very well predict the future success or failure of a firm (Kanji & Asher,
1993). Thus, it is very important to achieve customer satisfaction and their perception of
quality. The insights gained can clearly help the firm improve quality.

In-depth marketing research can identify suddenly arising customer needs. The attainment
of quality requires the performance of a wide variety of identification activities of quality
tasks such as the study of customers’ quality needs, design review, and field complaint
analysis (Juran & Gryna, 1993). To achieve quality, it is essential to know what customers
need and then provide products that meet their requirements (Ishikawa, 1985). According to
the review results from Hackman and Wageman (1995), obtaining data about customers is
one of the most commonly used TQM implementation practices. Deming (1986) suggested
that firms understand what the customer needs and wishes now and in the future so that
products and services can be designed to satisfy those needs and wishes.

In order to pursue customer focus, firms should always provide warranties on their products
sold to customers. Thus, customers will reduce their risk in buying products. In addition,
firms should pay sufficient attention to customer services. In a word, pursuing customer
focus efforts should be a long-term business strategy; it is never ending (Juran & Gryna,
1993). The consideration of customer focus and involvement for the present study is
summarised in Table 10.1.
240
10.2.4 Construction Process Management and Improvement (CPMI)
Process refers to certain unique combinations of tools, machines, methods, materials, and
people employed in production. Process control and improvement indicates a set of
methodological and behavioral practices, which are implemented to control and improve
processes that produce products and services (Juran & Gryna, 1993). In fact, process
control and improvement can enable the manufacturing process to function as anticipated,
without breakdowns, missing materials, fixtures, tools, and the like and regardless of
workforce variability (Flynn et al., 1994). A set of practices of process control and
improvement is described in the following paragraphs.

An essential issue in process control and improvement is the maintenance of process

capability to meet production requirements. Process capability is mostly independent of
specification tolerances for parts to be manufactured within the process. It is essential to
establish these capabilities as elemental to product-control standards setting (Feigenbaum,
1991). Process capability study gives a basis for this determination and its associated
assignment of parts to those facilities that can economically maintain the required
tolerances (Gitlow et al., 1989).

Juran and Gryna (1993) stated that another aspect of process control and improvement is
equipment maintenance, which enables variation to be kept within satisfactory limits and
keeps the manufacturing process running efficiently. Feigenbaum (1991) asserted that
manufacturing equipment inevitably wears under continuous usage which would result in
poor-quality products. Hence, a programme of preventive maintenance is an essential
quality management practice since it ensures a frequently scheduled examination of
processing facilities before they break down.

A company should endeavour to design its process to be “foolproof” in order to reduce the
possibility of worker mistake. Foolproofing explains methods, such as poka-yoke, which
enable activities or operations to be carried out in the proper means. Foolproofing
techniques can be divided into two types: Alarms and controls. Alarm devices may light a
red lamp, sound a buzzer, or flash an alarm light if a mistake is detected. Control devices
241
may interrupt work by activating a clamp, stopping a machine, or halting a conveyor if an
error happens so that a defect does not shift on to the next process (Slack et al., 1995).

A number of quality techniques and tools can be applied to control and improve processes.
These techniques comprise the seven QC tools and Plan, Do, Check, and Act (PDCA)
cycle. The PDCA cycle is fundamentally the scientific technique applied to continuous
process improvement (Dale, 1999; Deming, 1986; Mann, 1992). According to Dale (1999),
statistical process control can also be applied to attain process stability and offer guidance
on how the process may be improved. Statistical process control is the application of
statistical techniques to the measurement and analysis of variation in any process (Juran &
Gryna, 1993). Deming (1986) asserted that without statistical control, the process is in
chaos, the noise of which will mask the effect of any effort to bring improvement.

Construction Process Management and Improvement for the current study should also
consider quality system procedures improvement. Quality system procedures improvement
is defined as the organizational structure, procedures, processes and resources required to
apply quality management (ISO 8402, 1994). In 1987, the International Standardization
Organization published the ISO 9000 standards series on quality management and quality
assurance. Applying ISO 9000 is a way in seeking quality system improvement in a
company. In this study, quality system improvement means to set up a quality system
according to the specifications of ISO 9000. Through the execution of ISO 9000, a quality
manual, quality system procedures, and work instructions are established. In the end, a firm
may apply to be registered as having an ISO 9001 (9002 or 9003) quality certificate
(Mirams & McElheron, 1995; Randall, 1995).

A quality manual is a document stating the quality policy and describing the quality system
of an organization (ISO 8402, 1994), and should cover all the applicable elements of the
quality system standard required for an organization. Guidelines for developing quality
manuals (ISO 10013, 1995) can be used for drawing up a quality manual.

242
A procedure is a specified way to perform an activity. A written procedure contains the
reasons and scope of an activity; what shall be done and by whom; when, where and how it
should be done; what materials, equipment and documents shall be used; and how it shall
be managed and documented. Documented quality system documents explain the activities
of individual functional units required to apply the quality system elements (ISO 8402,
1994; ISO 10013, 1995). Work instructions include detailed work documents, which can
guide individuals in performing specific work. It should be noted that drawing up various
work instructions should be based on the existing documents and characteristics of the firm,
and should be presented to different individuals for extensive review. Thus, these work
instructions can be effectively implemented in practice (Mirams & McElheron, 1995;
Randall, 1995).

It should be noted that a quality system should be consistently improved. Quality system
documents should be consistently modified with the change of quality activities within the
firm. It is essential to maintain the quality system’s conformance with the ISO 9000
requirements at the same time (Randall, 1995). With an ISO 9000 quality system in place, a
company may consider becoming ISO 9000 certified. This present study looks at
Construction Process Management and Improvement that has been hypothesised for the
development of a holistic TQM model and are summarised in Table 10.1.

10.2.5 Construction Employee Involvement and Motivation (CEIM)

Every organization is concerned with what should be done to achieve sustained high levels
of performance through its workforce. This means giving close attention to how individuals
can best be motivated through means such as incentives, rewards, and promotions, among
others and the organizational context within which they carry out the work (Armstrong,
2006). The study of motivation is concerned with why people behave in a certain way. In
general it can be described as the direction and persistence of action. Hence, motivation is
defined as the set of processes that determine the choices people make about their
behaviours. It is concerned with why people choose a particular course of action in
preference to others, and why they continue with chosen action, often over a long period,
and in the face of difficulties and problems (Mullins, 2005). Motivation can therefore be
243
said to be at the heart of how innovative and productive things get done within an
organization (Bloisi et al., 2003). Motivation is an abstract term. It imparts incentives that
require a response on the part of someone else to achieve a defined goal. In business,
motivation is not synonymous with salaries; money is a means for accommodating the
economic needs of workers. Motivation means an inner wholesome desire to exert effort
without the external stimulus of money.

Motivating is the ability of indoctrinating the personnel with a unity of purpose and
maintaining a continuing, harmonious relationship among all people. It is a force which
encourages and promotes a willingness of every employee to cooperate with every member
of the team. To maintain it is to create and perpetuate the climate which brings harmony
and equilibrium into the entire work group for the benefit of all who are involved – the
company as a whole. It has been established that motivation is concerned with the factors
that influence people to behave in certain ways. Arnold et al. (1991) established the
following three components of motivation:
Direction: what the person is trying to do,
Effort: how hard a person is trying, and
Persistence: how long a person keeps on trying (Armstrong, 2006).
Since the effective motivation comes from within, by motivating others the manager can do
more than create conducive conditions that cause people to do their work with willingness and
enthusiasm. Motivating is the work managers perform to inspire, encourage and impel people to
take action (Allen, 1986). To motivate the employee, the employee must be reached; to reach
him there must be a completed understanding of the complexity his make-up (Allen, 1986).
Motivation efforts must be directed towards improving company operations. To be effective,
however, they must also be designed to show benefits to the employee. In fact, motivation can
best be accomplished when workers are able to merge their personal ambitions with those of the
company. According to Robin and DeCenzo (1995; 271) motivation is defined as “…the
willingness to exert a high level of effort to reach organizational goals, conditioned by the
effort’s ability to satisfy some individual need”. Campbell and Pritchard (1976) define
motivation as a set of independent and dependant relationships that explain the direction,
amplitude and persistence of an individual’s behaviour, holding constant the effects of aptitude,
skills, understanding of a task and the constraints operating in the work environment.
244
Schrader (1972) linked construction workers’ needs to motivation, and it was subsequently
concluded by Thomas et al. (1990) that there is evidence supporting the existence of a
linkage between employees’ motivational level and their individual performance. Atkinson
(1964) defines it as the contemporary immediate influence on the direction, vigour, and
persistence of action.

The relationship between the employer and employee must be one of understanding in order for
the employee to identify himself with his work and with the business he is working for. Lack of
motivation in return affects TQM. A number of symptoms may point to low morale:
declining productivity, high employee turnover, increasing number of grievances, higher
incidence of absenteeism and tardiness, increasing number of defective products, higher
number of accidents or a higher level of waste materials and scrap (Day, 1978). A
motivated employee is a loyal employee and to be loyal implies that the employee supports
the actions and objectives of the firm. The appearance of the job as a whole has a bearing
on the willingness and quality of an employee’s performance (Bruce, 1962).

According to McClelland (1961), individuals tend to develop certain motivational drives on

the cultural environment in which they live and these drives affect the way people view
their jobs. McClelland suggests that achievement, affiliation, competence and power are
four types of motivational drives that are found in individuals that are self-motivated and
this may be the case for many construction workers. Motivation plays a part in enhancing
construction labour productivity (Smithers & Walker, 2000) and forms the basis for the
identification of the work environment factors. For example, Laufer and Moore (1983)
advocated the use of financial incentive programmes to improve construction labour
productivity, reinforcing Maloney’s (1982) thesis of driving forces that lead to productivity
improvements.

Autonomy and comradeship (Edwards & Eckblad, 1984) are also found to be important
aspects that add to the way construction workers are self-motivated about their work.
However, much work in linking motivation and productivity relied on Hertzberg’s sample
involving mainly white-collar professionals (Mullins, 1996). Furthermore Hofstede, (1980)
245
decried such motivational theories as merely points made about the ad nauseam emphasis
on the managerial perspective in the quest to improve productivity. John Borcherding and
Clarkson Oglesby (1978) posited that a productive job creates high levels of job satisfaction
while a non-productive job (one which falls behind schedule) produces dissatisfaction at all
levels of the management/worker chain. The relationship is believed to be due to the very
nature of construction, thus different from the one found in an office or factory setting
where high levels of job satisfaction lead to greater productivity.

In construction, a worker, through his own efforts, produces a highly visible, physical
structure in which great satisfaction comes from completion. Therefore, jobs that are well-
planned and run smoothly produce great satisfaction while jobs with poor management
(with scheduling and planning problems) create dissatisfaction. This illustrates the
relationship between job satisfaction and productivity since well-managed jobs are
generally more productive. A close review of all theories of human motivation reveals a
common driving principle that people do what they are rewarded for doing. In general, the
theories on motivation can be classified as employees need motivation through goal-setting,
employee reward/incentives and reinforcement.

In order to keep construction workers motivated, their needs must be addressed as project
goals are reached. Satisfying workers’ needs can be viewed as distributing incentives when
certain objectives are achieved. Employees have needs that they want met and employers
have goals that they reach and they can work together as a team to satisfy the wants of both
the employees and their employers (McKenzie & Lee, 1998). Workers who are motivated
to help reach the goal of the employer and do so should be recognized with an incentive or
reward. When considering what type of incentives to use there are two types to be aware of,
namely extrinsic and intrinsic. Extrinsic rewards are external rewards that occur apart from
work, such as money and other material things. On the other hand, intrinsic rewards are
internal rewards that a person feels when performing a job so that there is a direct and
immediate connection between work and reward.

The power of incentives is immense and pervasive, which is all the more reason they
require careful management (McKenzie & Lee, 1998). Heap (1987) presented a

246
summarized list of advantages and disadvantages associated with financial incentives.
Many construction companies have already considered that there can be advantages and
disadvantages of developing an incentive programme. A study by Sanders and Thompson
(1999) showed that those companies that keep their programmes simple with the main
objective of the programe in mind (to benefit the project in reference to cost, schedule,
customer service, environment and quality) are also assured of the success of any incentive
programme. Incentives are usually defined as tangible rewards that are given to those who
perform at a given level. Such rewards may be available to workers, supervisors, or top
managers. Whether the incentive is linked directly to such items as safety, quality or
absenteeism, the reward follows successful performance (MaKenzie & Lee, 1998). Many
companies feel that pocket money is no longer a good motivator. Others contend that small
rewards such as toasters and blenders do not motivate. Many companies therefore offer
profit-sharing plans. Some companies have abandoned monetary rewards and instead offer
lavish trips to such places as Europe and some Caribbean islands. Because of the expense,
these programmes require careful monitoring. Some companies merely reward good
producers with an extra day off with pay (MaKenzie & Lee, 1998). Other concerns reward
top performers with better working conditions. Since incentive programmes aim to increase
workers’ performance levels, the measure used to decide whether a reward has been earned
should be carefully set. The performance level must be attainable or workers will not try to
reach the goal. That fact underscores the usefulness of having workers themselves
contribute their ideas about what constitutes a reasonable level of performance.

An incentive scheme may also fail if the measure of success ignores quality or safety. An
obvious problem exists when an incentive is applied to work that is machine paced.
Incentives should be clearly linked to performance, but not all incentives can be clearly tied
to objective criteria (MaKenzie & Lee, 1998). Some incentive rewards are issued on the
basis of a subjective assessment by a superior on the merit of particular workers. This
method in particular may cause conflicts among workers, especially those who do not win
rewards. This present study looks at construction employee involvement and motivation
that has been hypothesised for the development of a holistic TQM model as is summarised
in Table 10.1.

247
Table 110.1: Conceptual Model Latent Features

Latent Variable Construct Measurement Variables

Leadership and Top Management Leadership style

(TML) Leadership ability in solving quality problems
Leadership skills
Leadership initiatives
Leadership and top management knowledge in
TQM
Top management commitment in TQM
Top management interaction with workers
Top management participation
Top management learning
Top management empowerment
Top management encouragement
Top management’s role model
Company Supplier Quality Partnership with suppliers
Management (CSQM)
Supplier selection criteria

Participation with suppliers

Supplier performance evaluation

Supplier quality audit

Supplier communication
Supplier knowledge of TQM
Location of supplier
Suppliers past records
Suppliers commitment
Cooperation from suppliers

248
 Suppliers orientation
Client Focus and Involvement (CFI)
Client brief/ input

Client complaint information/feedback

Market investigation

Client satisfaction survey

Quality warranty

Client information system

Client services

Client cooperation

Client orientation

Company Quality System Evaluation Evaluation of strategy

(CQSE)
Evaluation of overall company performance

Evaluation of departments’ performance

Evaluation of employee performance

Evaluation of quality costs

Evaluation of quality manual

Evaluation of quality control system

Evaluation of quality system procedures

Evaluation of end results

Quality audit

Quality benchmarking

Quality information system

249
Company Vision and Plan Statement Vision statement
(CVPS)
Quality policy

Overall business performance plan

Product quality goal

Quality improvement plan

Formulation of vision and plan

Concreteness of future plan

Employee contribution to the vision

Involvement of employees in the development of

vision statement

Product Selection and Design Client brief/input

Management (PSDM)
Cost implication

Environmental issues

Socio cultural issues

Appearance/finishes required

Strength of required

Intended purpose of the material

Where the material to be used

Design methods/techniques

Design detailing

Concurrent engineering

Design experiments/testing

Quality function deployment

250
 Value engineering

Computer-aided design

Construction Process Management Project monitoring and control improvement

and Improvement (CPMI)
Equipment maintenance and innovation

Inventory management

Use of quality materials

Use of quality tools

Use of quality manual

Understanding of work instructions/quality

manual

Quality system procedures improvement

Obtaining ISO 9000 certification for operation

Understanding of quality control system

Continuity of quality control circle activities

Appropriate use of system structure and standards

Targets and priority measures

Safe working environment

Utilization of analysis results

Construction Employees’ Education and training

Involvement and Motivation (CEIM)
Salary promotion

Position promotion

Employee rewards

251
 Bonus scheme

Conducive working environment

Involvement of employees within-functional team

Employee participation

Employee suggestions

Improving employee commitment

Job rotation

Involvement in quality control circle

Employee involvement in cross-functional team

Employee recognition

TQM Outcomes (TQMO) Greater reduction of construction costs

Higher product quality

Improved schedule performance

Elimination of reworks

Defect-free product at first attempt

Higher safety standards

Higher productivity

Good methods of working

Higher profitability

Higher reduction of construction waste

252
10.3 MODEL SPECIFICATION AND JUSTIFICATION
This thesis aims to build a conceptual TQM model. The theoretical conceptual framework
for the current research is built on the work of Saraph et al. (1989), Flynn et al. (1994),
Ahire et al. (1996) and Adusa–Poku (2014) whose works were also built on the previous
Deming theory and that of quality award models, as discussed in Chapter 2 (Section 2.6.1 –
2.6.2). The theoretical essence of the Deming (1986) approach to TQM concerns the
creation of an organizational system that fosters cooperation and learning for facilitating the
implementation of process management practices, which, in turn, lead to continuous
improvement of processes, products, and services as well as to employee fulfillment, both
of which are critical to customer satisfaction, and ultimately, to firm survival. Adusa-Poku
(2014) shared a similar view with Deming (1986). Adusa-Poku emphasised that process
management, continuous improvement, employees’ satisfaction/empowerment, supplier
chain management, customer focus, management/leadership, and training are the set of
basic factors forming the building blocks of the proposed TQM framework (Adusa-Poku,
2014). The proposed framework by Adusa-Poku (2014) is based on six criteria which are
also based on European Foundation for Quality Management (EFQM). EFQM is also a
quality model based on nine (9) elements; five (5) enablers and the four (4) elements are
Results as explained in Chapter 2 (refer to section 2.6.1.2). It was argued that three basic
criteria needed for a successful implementation of a TQM framework are process
management, leadership commitment, and customer focus (Adusa-Poku, 2014). These three
basic criteria are the triangular pillars on which TQM derives its support. Process
Management is at the top of the triangle, indicating that without process management the
framework will fail. Customer Focus and Management are the Human Resources who play
their role effectively in managing all necessary processes. The customers include the
internal and external customers. Any one of these criteria that fails leads to a failure of
TQM. Adusa-Poku (2014) further stated that Continuous Improvement, Training, and
Supplier Management are the connectors to the main pillars for an effective and efficient
TQM implementation in the construction industry.

Saraph et al. (1989), Flynn et al. (1994), Ahire et al. (1996) as well as Adusa–Poku (2014)
conceptualised TQM models which consists of the following common basic components:

253
leadership/management, supplier quality management, evaluation, process control and
improvement, product design, employee participation, recognition and reward, education
and training, and customer focus and involvement. The fundamental underpinning of four
models, and the incorporated theoretical perspectives which have been adopted in other
similar studies are therefore useful for conceptualizing the present study as a variety of
TQM studies and TQM being conceptualized within the broader theoretical framework.
Therefore, the conceptual framework for this thesis is primarily based on the approach used
by Saraph et al. (1989), Flynn et al. (1994), Ahire et al. (1996) as well as Adusa–Poku,
(2014). Based on the fundamental factors and constructs associated with all the previous
models as revealed in Table 10.1, the present model or conceptual framework for the study
looks at the leadership and top management, company supplier quality management, client
focus and involvement, construction process management and improvement, and
construction employee involvement and motivation. These factors have been measured in
the majority of the previous studies, but consideration has not been given to company vision
and plan statement, product selection and design management, and company quality system
evaluation which have been classified as the exogenous variables and their role in
predicting overall TQM, which is the endogenous variable. These will in turn, predict the
construction industry implementation of TQM. The study aims to forecast the relative
predictive power of these different variables for TQM in order to test or determine whether
TQM implementation depends on the supposed features of the variables as indicated in the
other frameworks, taking into account the needs of the construction industry in Ghana. It is
apparent that some of the variables discussed above should be measured by objective
means, some by subjective means and some include both forms of measurement. The
reason for combining both objective and subjective indicators within the proposed model is
supported by Campbell, Converse and Rogers (1976) and Falah, Al-Abed and Stan (1995)
who stated that by themselves, objective indicators are often misleading and will remain so
until indicators that human beings attached to them are obtained. Likewise, by themselves,
subjective indicators are insufficient as guides to policy. The conceptual model theorises
that TQM is established by the relationship that exists between the exogenous variables
which include the basic elements by which the subjective and objective measurements are
linked. These variables identified from the review of the literature are considered to be the
254
major determinants of TQM. The determinants identified have been adopted to fit into the
TQM implementation in the Ghanaian construction industry.

10.4 STRUCTURAL COMPONENT OF THE MODEL

The integrated TQM model for the Ghanaian construction industry in the case of
developing countries is derived from Leadership and Top Management (LTM), Company
Supplier Quality Management (CSQM), Company Quality System Evaluation (CQSE),
Client Focus and Involvement (CFI), Company Vision and Plan Statement (CVPS), Product
Selection and Design Management (PSDM), Construction Process Management and
Improvement (CPMI), and Construction Employees’ Involvement and Motivation (CEIM),
in the process of achieving TQM in the construction industry. The TQM model to be tested
in the postulated hypothesis not based on prior study or any examination is composed of
LTM, CSQM, CQSE, CFI, CVPS, PSDM, CPMI, CEIM, and it is a multidimensional
structure. The postulated model is presented in Figure 10.1 (Model 1.0). The conceptualized
model is the notion that TQM implementation is related to the evaluation of many variables,
such as LTM, CSQM, CQSE, CFI, CVPS, PSDM, CPMI, and CEIM. It is therefore
difficult to discuss the principal variable without reference to variables of company vision
and plan statement, product selection and design management, and company quality system
evaluation and the inclusion of the other exogenous variables. In this study, the objective
evaluation of TQM was assessed by measuring the actual condition of the construction
industry that is an exogenous variable in the model.

255
 Leadership and Top Management
(LTM)

Company Supplier Quality

Management (CSQM)

Client Focus and Involvement

(CFI)

Construction Process Management

and Improvement (CPMI)

Total Quality Management

Construction Employees Involvement (TQM)
and Motivation (CEIM)

Company Vision and Plan

Statement (CVPS)

Product Selection and Design

Management (PSDM)

Company Quality System

Evaluation (CQSE)

Figure 110.1: Integrated TQM Conceptualised Model (Model 1.0)

10.5 MEASUREMENT COMPONENT OF THE MODEL DYNAMIC

The measurement component of the hypothesized model comprises the following TQM
implementation factors: LTM = 12 measurement variables; CSQM = 12 measurement
variables; CFI = 9 measurement variables; CEIM = 14 measurement variables; CVPS = 9
measurement variables; PSDM = 15 measurement variables; CPMI = 15 measurement
variables; CQSE = 12 measurement variables and TQMO = 10 manifest measurement
variables. The success for the consideration of TQM implementation for the benefit of the
construction industry has been theorised in the present model.

256
10.6 CONCLUSION
The theorised conceptual model given in this chapter is a multidimensional structure
composed of eight latent variables of Leadership and Top Management, Company Supplier
Quality Management, Client Focus and Involvement, Construction Process Management
and Improvement, Construction Employees’ Involvement and Motivation, Company Vision
and Plan Statement, Product Selection and Design Management, and Company Quality
System Evaluation. These factors were derived from the literature review. The explanation
of the selected variables for the construction of the integrated TQM model has been
highlighted in the theoretical framework in this chapter. Chapter 11 present the discussion
of the survey results.

257
 CHAPTER ELEVEN
11 SURVEY ANALYSIS
11.1 INTRODUCTION
The theoretical conceptual model is presented in Figure 10.1 (Model 1.0). The hypothesized
integrated holistic TQM model theory is based on literature and the views of experts
obtained during the Delphi study. The views of the experts have been described in detail
Chapter 8. This chapter presents the descriptive statistics, inferential statistics and
hypotheses testing results based on the questionnaire analysis. Statistical analyses
techniques used to test the validity and reliability of the measuring instrument have been
discussed in this chapter. Other statistical techniques used are the empirically testing of the
proposed conceptual model presented in Chapter 10. The results obtained from the
quantitative survey are provided and discussed. The data analysis was conducted in two
stages:
i. Descriptive data analysis, and
ii. Multivariate correlational data analysis including exploratory factor analysis (EFA),
confirmatory factor analysis (CFA), and structural equation modelling (SEM) software
EQations (EQS) Version 6.2 for analysis (Bentler, 1999).

A sample size of 536 questionnaires were answered by respondents and returned at the end
of the survey. Greater model fit bias as indicated by Tong (2007) has been attributed to
sample size which will affect the model fit. Therefore, the sample size obtained for the
current study is considered as large (Kline, 2005:15). Harris and Schaubroeck (1990)
indicated that a sample size of fewer than 100 cases will be difficult to analyse when the
analytical tool to be used is SEM. A similar view has been shared by Kline (2005:15).
Sometimes, appropriate sample size may depend on observed variables as indicated by
Tong (2007) and supported by MacCallum, Browne and Sugawara (1996). The minimum
sample of 200 for SEM analysis was used by Bentler and Chou in 1987. The variable ratio
of an ideal SEM model has been suggested by Tong (2007) to be at least 5:1. This implies
that a SEM model with 10 observed variables should have more than 50 respondents. For
the purpose of this study the researcher collected 536 responses. There are 74 hypothesized
observed variables and the ratio to sample size for the current study is 7.24:1. Therefore, the
258
variable ratio to sample size meets the requirement recommendation in literature by Tong
(2007). The sample data (536) was finally taken through random sampling before carrying
out the EFA and CFA. A total of 275 samples were realized for the EFA analysis and 261
samples for the CAF analysis.

11.2 DESCRITIVE STATISTICS

This section provides demographic information on the individual respondents. The analyzed
results for the descriptive data were the respondents’ background information including
their individual information and the company information. Descriptive statistics such as
percentages, means and standard deviation were used in the analysis.

11.2.1 Demographic Profile of the Respondents and Firms

A total of 536 samples were realized for the study. The responses represent 85.1 per cent (N
= 456) males and 14.9 per cent (N = 80) females, as shown in Table 11.1.

Table 11.1: Gender

Gender Frequency Percentage

Male 456 85.1
Female 80 14.9
Total 536 100.0
Source: Fieldwork (2018)
Table 11.2 shows that the majority of the respondents in the age group of 26 to30 years and
36 to 40 years constituted (26.9%; N=144) each of the sample, followed by the age group
of 21 to 25 years (20.5%; N=110 ), 31 to 35years (17.2%; N=92), 41 to 45 years (6.0%:
N=32), 15 to 20years (1.7%; N=9) and the age range 46 years and above which constituted
(0.9%; N=5) of the sample.

259
 Table11.2: Age Category

Age Category Frequency Percentage

15-20 years 9 1.7
21-25 years 110 20.5
26-30 years 144 26.9
31-35 years 92 17.2
36-40 years 144 26.9
41-45 years 32 6.0
46 years and above 5 0.9
Total 536 100.0
Source: Fieldwork (2018)
The highest education level of the majority of the sample respondents was a bachelor’s
degree (first degree) certificate (42.2%; N = 226), followed by postgraduate master’s
degree (MSc., MPhil.) (31.5; N = 169), a national/national certificate (25.45; N = 136) and
the least number of the sample respondents held doctoral (PhD) degrees (0.9%; N=5) as
shown in Table 11.3.

Table 11.3: Highest Qualification

Qualification Frequency Percentage

National/National Certificate 136 25.4
Bachelor’s degree (first degree) 226 42.2

Postgraduate – Master’s degree (MSc., 169 31.5

MPhil.)
Doctoral degree (PhD) 5 0.9
Total 536 100.0
Source: Fieldwork (2018)

Table 11.4 shows that a large number of the respondents (30.0% ; N-161) have worked in
the construction industry for between one to five years, followed by a group with six to ten
years of service (25.2%; N = 135), eleven to fifteen years (23.3%; N = 125) , sixteen to
twenty years (10.3%; N = 55), twenty-one to twenty-five years (6.7%; N = 36), twenty-six
to thirty years (3.2%; N = 17) and the year range 31years and above constituted the least

260
number of years of work experience in the construction industry, namely 1.3 per cent of the
sample.
Table 11.4: Work Experience in the Construction Industry

Work experience Frequency Percentage

1-5 years 161 30.0
6-10 years 135 25.2
11-15 years 125 23.3
16-20 years 55 10.3
21-25 years 36 6.7
26-30 years 17 3.2
31 years and above 7 1.3
Total 536 100.0
Source: Fieldwork (2018)

The majority of the respondents (40.91%; N= 219) held the position of Site Engineer in
their firm, followed by the Site Supervisor position (28.0%; N=150). The position of
Quality Control Officer constituted the least (2.4 %; N=13) of the sample as shown in Table
11.5.
Table 11.5: Position Held in the Firm

Position Frequency Percentage

Project Manager 87 16.2
Site Engineer 219 40.9
Quality Control Officer 13 2.4
Supervisor 150 28.0
Foreman 27 5.0
Others 40 7.5
Total 536 100.0
Source: Fieldwork (2018)

Table 11.6 shows that a large number of the respondents (56.0%; N=300) have worked in
their current firm for between one to five years, followed by a group with six to ten years,
11-15 years, 16-20 years, 21-25 years, and 26-30years representing (19.6%; N=105),
(14.6%; N=78), (5.4%; N=29), (4.3%; N=23), (0.2%: N=1) respectively of the sample. The
ranges of year thirty-one years and above constituted 0.0%.

261
 Table 11.6: Tenure of Current Position

Tenure Frequency Percentage

1-5 years 300 56.0
6-10 years 105 19.6
11-15 years 78 14.6
16-20 years 29 5.4
21-25 years 23 4.3
26-30 years 1 0.2
31 or more years 0 0.0
Total 536 100.0
Source: Fieldwork (2018)

The majority of the respondents (29.3%; N=157) indicated their firms have been in
existence in the range of six to ten years, followed by the range of eleven to fifteen years
(17.4%; N=93). The firms with the year range of sixteen to twenty years constituted the
least (5.6 %; N=30) of the sample as shown in Table 11.7.

Table 11.7: Existence of Firm

Firm Existence Frequency Percentage

1-5 years 85 15.9
6-10 years 157 29.3
11-15 years 93 17.4
16-20 years 30 5.6
21-25 years 64 11.9
26-30 years 35 6.5
31 or more years 72 13.4
Total 536 100.0
Source: Fieldwork

The majority of the construction projects currently undertaken by the firms were building
construction projects (65.9%; N=353), followed by civil engineering projects (29.1%;
N=156) while both building and civil construction projects constituted the least (5.0%;
N=27) of the sample as shown in Table 11.8.

262
 Table 11.8: Type of Current Construction Projects

Type of construction project Frequency Percentage

Building construction 353 65.9
Civil engineering 156 29.1
Both building and civil works 27 5.0
Total 536 100.0
Source: Fieldwork (2018)

Table 11.9 shows that a large number of the respondents (44.0%; N=236) worked with the
D2/K2 class of contractors, followed by the D1/K1 class (32.6 %; N=175) and the D4/K4
class constituted the least (10.8%; N=58) of the sample.

Table 11.9: Classification of Firm

Classification Frequency Percentage

D1/K1 175 32.6
D2/K2 236 44.0
D3/K3 67 12.5
D4/K4 58 10.8
Total 536 100.0
Source: Fieldwork (2018)

11.2.2 Level of TQM Implementation in the Ghanaian Construction Industry

Table 11.10 indicates the level of TQM implementation in the Ghanaian construction
industry in terms of percentage responses. The respondents were asked to choose between
the following statements; ‘To no extent’, ‘To a small extent’, ‘To a moderate extent’, ‘To a
large extent’, ‘Completely’ to indicate the level of TQM implementation in their firm. The
majority of the respondents (38.1%; N=204) indicated that their firms have implemented
TQM to a moderate extent, followed by 30.4%; N=163 indicating that their firms have
implemented TQM to a large extent and the statement ‘To no extent’ constituted the least
level TQM implementation by the firms, namely 2.4 per cent of the sample. This suggests
that not all the firms under study have implemented TQM.

263
Table 11.10: TQM Implementation in the Ghanaian Construction Industry

Extent of TQM Frequency Percentage

implementation
To no extent 13 2.4
To a small extent 106 19.8
To a moderate extent 204 38.1
To a large extent 163 30.4
Completely 50 9.3
Total 536 100.0
Source: Fieldwork (2018)

11.2.3 Attributes Influencing the Implementation of TQM in the Ghanaian

Construction Industry
Table 11.11 indicates the Leadership and Top management features in terms of percentage
responses on a scale of 1 (Not at all influential) to 5 (Extremely influential), and a MS
ranging between 1.00 and 5.00. All the MSs are above the midpoint score of 3.00, which
indicates that the respondents agreed with the Leadership and Top Management features of
TQM implementation. It is notable that all the Leadership and Top Management features
have a MS > 4.00 ≤ 5.00, which indicates that the respondents perceive the Leadership and
Top Management features to be between ‘very influential’ and ‘extremely influential’. The
relatively high MS = 4.02- 4.24 achieved suggests that these variables are very significant
in driving TQM implementation among construction firms in Ghana.

264
Table 11.11: Leadership and Top Management features

Attributes Not at all MS SD Rank

influential……..Extremely
1 2
influential 3 4 5
Top management 0.00 2.24 15.86 37.69 44.22 4.24 0.80 1
participation in TQM
activities
Top management support of 0.00 4.85 19.96 27.43 47.76 4.18 0.92 2
TQM
Top management 1.12 3.36 21.27 27.05 47.20 4.16 0.95 3
commitment to TQM
Leadership ability in solving 1.31 1.68 19.22 36.38 41.42 4.15 0.88 4
quality-related problems
Top management
knowledge and proper 1.12 4.66 19.40 32.65 42.16 4.10 0.95
5
understanding of TQM
Leadership style of 0.00 4.29 18.84 39.55 37.31 4.10 0.85 5
managing employees
Top management interaction 0.19 2.24 20.34 42.16 35.07 4.10 0.81 5
with workers
Leadership initiatives 0.00 3.54 23.13 35.82 37.50 4.07 0.86 6
towards TQM
Top management 0.00 3.54 25.37 36.38 34.70 4.02 0.86 7
empowerment of employees
Source: Fieldwork (2018)
to solve quality problems

Table 11.12 indicates the Company Supplier Quality Management features in terms of
percentage responses on a scale of 1 (Not at all influential) to 5 (Extremely influential), and a
MS ranging between 1.00 and 5.00. All the MSs are above the midpoint score of 3.00,
which indicates that the respondents agreed with the Company Supplier Quality
Management features of TQM implementation. It is notable that all the Company Supplier
Quality Management features have a MS > 3.50 ≤ 5.00, which indicates that the
respondents perceive the Company Supplier Quality Management features to be ‘very

265
influential’. The relatively high MS = 3.88 - 4.13 achieved suggests that these variables are
very significant in driving TQM implementation among construction firms in Ghana.

Table 11.12: Company Supplier Quality Management Features

Attributes Not at all MS SD Rank

influential……..Extremely
1 2
influential 3 4 5
Supplier quality audit 0.19 2.80 19.40 38.99 38.62 4.13 0.83 1
Supplier knowledge of TQM 0.19 4.29 23.51 29.29 42.72 4.10 0.92 2
Supplier communication 0.19 2.43 21.83 39.37 36.19 4.09 0.83 3
Supplier performance 0.56 5.22 22.20 29.66 42.35 4.08 0.95 4
evaluation
Participation of suppliers in 0.19 8.40 16.23 35.07 40.11 4.07 0.96 5
TQM activities
Suppliers’ commitment to 0.56 4.48 24.81 28.92 41.23 4.06 0.94 6
TQM
Supplier selection criteria 0.00 6.53 16.79 44.22 32.46 4.03 0.87 7
Suppliers’ orientation on 0.19 3.73 26.49 38.06 31.53 3.97 0.86 8
TQM
Partnership with suppliers 0.00 4.66 30.41 37.50 27.43 3.88 0.87 9

Source: Fieldwork (2018)

Table 11.13 indicates the Client Focus and Involvement features in terms of percentage
responses on a scale of 1 (Not at all influential) to 5 (Extremely influential), and a MS
ranging between 1.00 and 5.00. All the MSs are above the midpoint score of 3.00, which
indicates that the respondents agreed with the Client Focus and Involvement features of
TQM implementation. It is notable that all the Client Focus and Involvement features have
a MS > 4.00 ≤ 5.00, which indicates that the respondents perceive the Client Focus and
Involvement features to be between ‘very influential’ and ‘extremely influential’. The
relatively high MS = 4.01- 4.29 achieved suggests that these variables are very significant
in driving TQM implementation among construction firms in Ghana.

266
Table 11.13: Client Focus and Involvement Features

Attributes Not at all MS SD Rank

influential……..Extremely
1 2
influential 3 4 5
Client cooperation 0.37 .19 15.86 37.50 46.08 4.29 0.76 1
Client information system 0.00 5.22 18.28 28.73 47.76 4.19 0.91 2

Quality warrant to client 0.00 1.87 18.10 41.23 38.81 4.17 0.78 3
Client services 0.00 3.54 19.59 39.18 37.69 4.11 0.84 4
Market investigation 0.19 2.99 19.03 41.42 36.38 4.11 0.82 4
Client satisfaction survey 0.00 7.09 15.49 41.42 36.01 4.06 0.89 5
Client compliant 0.00 2.24 25.37 39.18 33.21 4.03 0.82 6
information/feedback
Client brief/input 0.00 3.36 24.25 40.11 32.28 4.01 0.84 7

Source: Fieldwork (2018)

Table 11.14 indicates the Company Quality System Evaluation features in terms of
percentage responses on a scale of 1 (Not at all influential) to 5 (Extremely influential), and a
MS ranging between 1.00 and 5.00. All the MSs are above the midpoint score of 3.00,
which indicates that the respondents agreed with the Company Quality System Evaluation
features of TQM implementation. It is notable that all the Company Quality System
Evaluation features have a MS > 4.00 ≤ 5.00, which indicates that the respondents perceive
the Company Quality System Evaluation features to be between ‘very influential’ and
‘extremely influential’. The relatively high MS = 4.11- 4.29 achieved suggests that these
variables are very significant in driving TQM implementation among construction firms in
Ghana.

267
Table 11.14: Company Quality System Evaluation Features

Attributes Not at all MS SD Rank

influential……..Extremely
1 2
influential 3 4 5
Evaluation of overall 0.00 2.43 11.57 40.86 45.15 4.29 0.76 1
company performance
Evaluation of the quality 0.00 2.61 12.13 42.16 43.10 4.26 0.77 2
strategy framework
Evaluation of quality system 0.00 2.99 15.30 37.69 44.03 4.23 0.81 3
procedures
Evaluation of the quality 1.12 1.49 18.47 35.82 43.10 4.18 0.86 4
control system
Evaluation of the quality 0.37 1.87 20.15 34.51 43.10 4.18 0.84 4
manual
Evaluation of the quality 0.00 5.22 12.50 41.42 40.86 4.18 0.84 4
information system
Evaluation of end results 0.00 1.49 19.78 38.62 40.11 4.17 0.79 5

Evaluation of employee 0.19 4.29 14.93 40.30 40.30 4.16 0.85 6

performance
Evaluation of quality costs 0.00 2.80 18.84 42.54 35.82 4.11 0.80 7

Source: Fieldwork (2018)

Table 11.15 indicates the Company Vision and Plan Statement features in terms of
percentage responses on a scale of 1 (Not at all influential) to 5 (Extremely influential), and a
MS ranging between 1.00 and 5.00. All the MSs are above the midpoint score of 3.00,
which indicates that the respondents agreed with the Company Vision and Plan Statement
features of TQM implementation. It is notable that all the Company Vision and Plan
Statement features have a MS > 4.00 ≤ 5.00, which indicates that the respondents perceive
the Company Vision and Plan Statement features to be between ‘very influential’ and
‘extremely influential’. The relatively high MS = 4.05- 4.28 achieved suggests that these
variables are very significant in driving TQM implementation among construction firms in
Ghana.

268
Table 11.15: Company Vision and Plan Statement Features

Attributes Not at all MS SD Rank

influential……..Extremely
1 2
influential 3 4 5
Good quality policy and 0.00 1.12 13.25 42.54 43.10 4.28 0.73 1
driven agenda
Involvement of employees 0.00 2.80 18.47 34.89 43.84 4.20 0.84 2
in the development of the
Quality improvement plan 0.00 2.61 19.96 34.70 42.72 4.18 0.84 3
vision statement
Clarity of vision and plan 0.00 4.66 18.10 38.81 38.43 4.11 0.86 4
statement
Employee contribution to 0.00 5.97 23.32 27.43 43.28 4.08 0.95 5
the vision
Concreteness of the future 0.00 6.34 21.08 33.77 38.81 4.05 0.92 6
plan
Formulation of vision and 0.00 5.04 20.15 39.74 35.07 4.05 0.87 6
plan
Source: Fieldwork (2018)

Table 11.16 indicates the Product Selection and Design Management features in terms of
percentage responses on a scale of 1 (Not at all influential) to 5 (Extremely influential), and a
MS ranging between 1.00 and 5.00. All the MSs are above the midpoint score of 3.00,
which indicates that the respondents agreed with the Product Selection and Design
Management features of TQM implementation. It is notable that all the Product Selection
and Design Management features have a MS > 3.50 ≤ 5.00, which indicates that the
respondents perceive the Product Selection and Design Management features to be ‘very
influential’. The relatively high MS = 3.83- 4.23 achieved suggests that these variables are
very significant in driving TQM implementation among construction firms in Ghana.

269
Table 11.16: Product Selection and Design Management Features

Attributes Not at all MS SD Rank

influential……..Extremely
1 2
influential 3 4 5
Design methods/techniques 0.00 1.49 15.49 41.98 41.04 4.23 0.76 1
Strength required for the end 0.00 1.12 19.40 36.94 42.54 4.21 0.79 2
product
Appearance/finishes 0.00 3.17 17.72 39.37 39.74 4.16 0.82 3
required
Intended purpose of the 0.00 2.80 19.22 37.87 40.11 4.15 0.83 4
material
Design detailing 0.00 .19 22.39 39.93 37.50 4.15 0.77 4
Where the material to be 0.00 2.99 16.04 45.71 35.26 4.13 0.78 5
used
Environmental issues 0.00 5.97 16.04 36.75 41.23 4.13 0.89 5

Cost of selection and design 0.00 5.41 15.11 42.54 36.94 4.11 0.85 6
of the product
Client brief/input 1.87 4.85 15.67 39.93 37.69 4.07 0.94 7

Socio-cultural issues 0.19 5.04 28.73 44.03 22.01 3.83 0.83 8

Source: Fieldwork (2018)

Table 11.17 indicates the Construction Process Management and Improvement features in
terms of percentage responses on a scale of 1 (Not at all influential) to 5 (Extremely
influential), and a MS ranging between 1.00 and 5.00. All the MSs are above the midpoint
score of 3.00, which indicates that the respondents agreed with the Construction Process
Management and Improvement features of TQM implementation. It is notable that all the
Construction Process Management and Improvement features have a MS > 4.00 ≤ 5.00,
which indicates that the respondents perceive the Construction Process Management and
Improvement features to be between ‘very influential’ and ‘extremely influential’. The
relatively high MS = 4.02 - 4.33 achieved suggests that these variables are very significant
in driving TQM implementation among construction firms in Ghana.

270
Table 11.17: Construction Process Management and Improvement Features

Attributes Not at all MS SD Rank

influential……..Extremely
1 2
influential 3 4 5
Understanding of work 0.19 .56 18.66 26.87 53.73 4.33 0.81 1
instructions/quality manual
Project monitoring and 0.00 4.10 13.43 36.19 46.27 4.25 0.84 2
control improvement
Use of a quality tools 0.00 .19 17.54 39.93 42.35 4.24 0.74 3
Efficient management 0.00 3.17 15.86 35.26 45.71 4.24 0.83 3
system
Use of quality materials 0.00 3.54 14.37 39.37 42.72 4.21 0.82 4
Equipment maintenance and 0.00 .93 15.86 44.40 38.81 4.21 0.74 4
innovation
Quality control system 0.19 2.80 17.16 36.75 43.10 4.20 0.83 5

Use of quality manual 0.00 5.04 13.43 41.60 39.93 4.16 0.84 6
Appropriate use of system 0.00 2.24 19.03 39.93 38.81 4.15 0.80 7
structure and standards
Institution of a continuous
0.00 3.36 20.15 38.81 37.69 4.11 0.84
8
improvement process within
the company
Inventory management 0.00 5.22 17.16 43.47 34.14 4.07 0.85 9

Obtaining ISO 9000 2.05 6.34 17.54 35.26 38.81 4.02 1.00 10
certification for operation
Source: Fieldwork (2018)

Table 11.18 indicates the Construction Employees Involvement and Motivation features in
terms of percentage responses on a scale of 1 (Not at all influential) to 5 (Extremely
influential), and a MS ranging between 1.00 and 5.00. All the MSs are above the midpoint
score of 3.00, which indicates that the respondents agreed with the Construction Employees
Involvement and Motivation features of TQM implementation. It is notable that all the
Construction Employees Involvement and Motivation features have a MS > 4.00 ≤ 5.00,
which indicates that the respondents perceive the Construction Employees Involvement and

271
Motivation features to be between ‘very influential’ and ‘extremely influential’. The
relatively high MS = 4.04- 4.29 achieved suggests that these variables are very significant
in driving TQM implementation among construction firms in Ghana.

Table 11.18: Construction Employees’ Involvement and Motivation Features

Attributes Not at all influential…….. MS SD Rank

…………….Extremely influential
1 2 3 4 5
Teamwork 0.00 .75 12.13 44.78 42.35 4.29 0.70 1

Recognition of employees 0.19 1.31 20.34 34.33 43.84 4.20 0.82 2

Conducive working 0.00 3.92 17.54 37.13 41.42 4.16 0.85 3
environment
Employee involvement in 0.00 3.17 18.10 38.25 40.49 4.16 0.83 3
TQM activities
Education and training/re- 3.36 5.78 12.50 28.17 50.19 4.16 1.07 3
training of staff
Salary/position promotion 0.00 4.48 16.98 39.74 38.81 4.13 0.85 4

Bonus scheme 0.19 7.84 15.11 33.40 43.47 4.12 0.95 5

Employee commitment 0.00 2.80 20.90 39.37 36.94 4.10 0.83 6

Employee reward 1.31 6.16 13.81 38.99 39.74 4.10 0.94 6

Availability of a suggestion 0.19 2.24 26.31 35.82 35.45 4.04 0.85 7
forum for employees
Source: Fieldwork (2018)

11.2.4 Project Outcomes that Motivate Construction Companies to Implement TQM

Table 11.19 indicates the project outcomes that motivate firms to implement TQM in the
Ghanaian construction industry in terms of percentage responses on a scale of 1 (Not a
motivating factor) to 5 (Very high motivating), and a MS ranging between 1.00 and 5.00.
All the MSs are above the midpoint score of 3.00, which indicates that the respondents
agreed with the project outcomes that motivate firms to implement TQM in the Ghanaian
construction industry. It is notable that all the ten ranked motivating factors influencing the

272
implementation of TQM in the construction industry have a MS > 4.00 ≤ 5.00, which
indicates that the respondents perceive the motivating factors influencing the
implementation of TQM in the construction industry to be between ‘high motivating factor’
and ‘very high motivating factor’. All the rankings of motivating factors that motivate the
implementation of TQM in the construction industry indicate MS> 4.08 - 4.35. The
relatively high MS achieved suggests that all the motivating factors that motivate the
implementation of TQM in the construction industry are very significant in driving TQM
implementation among construction firms in Ghana.

Table 11.19: Project Outcomes that Motivate Construction Companies to Implement

TQM

Project Outcomes Not a motivating factor…Very MS SD Rank

high motivating
1 2 3 4 5
Higher productivity 0.00 .93 10.07 41.79 47.20 4.35 0.70 1

Good methods of working 0.00 .75 11.01 46.27 41.98 4.29 0.69 2
Higher product quality 0.00 1.12 15.49 37.50 45.90 4.28 0.76 3

Higher reduction of 1.12 1.31 10.82 42.35 44.40 4.28 0.79 3

construction waste
Higher safety standards 0.00 .93 14.18 42.72 42.16 4.26 0.73 4

Higher profitability 0.00 .19 15.49 42.54 41.79 4.26 0.72 4

Defect-free product at first 0.00 1.87 23.88 28.73 45.52 4.18 0.86 5
attempt
Elimination of reworks 1.12 3.54 17.54 39.18 38.62 4.11 0.89 6

Greater reduction of 3.54 3.17 18.66 30.04 44.59 4.09 1.04 7

construction costs
Improvement schedule 0.00 3.73 20.90 39.37 36.01 4.08 0.84 8
performance
Source: Fieldwork (2018)

273
11.2.5 Reasons for Non-Implementation (Non-Adoption) of TQM in Ghanaian
Construction Industry
Table 11.20 indicates the factors influencing firms’ non-implementation of TQM in the
Ghanaian construction industry in terms of percentage responses on a scale of 1 (strongly
disagree) to 5 (strongly agree), and a MS ranging between 1.00 and 5.00. All the MSs are
above the midpoint score of 3.00, which indicates that the respondents agreed with the
factors influencing firms’ non-implementation of TQM in the Ghanaian construction
industry. It is notable that all the sixteen ranked factors influencing firms’ non-
implementation of TQM in the Ghanaian construction industry have a MS > 3.50 ≤ 5.00,
which indicates that the respondents perceive the factors influencing firms’ non-
implementation of TQM in the Ghanaian construction industry to be between ‘neutral’ and
‘agree’. All the rankings of factors influencing firms’ non-implementation of TQM in the
Ghanaian construction industry indicate MS> 3.79 - 4.33. The relatively high MS achieved
suggests that these variables are the most common reasons why construction companies in
Ghana do not implement or adopt TQM.

Table 11.20: Reasons for Non-Implementation (Non-Adoption) of TQM

Reasons/Factors Strongly disagree…Strongly MS SD Rank

agree
1 2 3 4 5
Lack or limited knowledge 0.00 3.54 12.87 30.60 52.99 4.33 0.83 1
of TQM
Lack of enforcement from
the legislative bodies 0.00 2.43 18.47 33.96 45.15 4.22 0.83
overseeing the
2
implementation of TQM
Lack of efficient TQM 0.93 4.48 15.49 35.63 43.47 4.16 0.91 3
management system
Lack of interest in the 0.37 4.10 16.23 41.23 38.06 4.13 0.85 4
application of TQM
Absent of TQM policy 0.56 6.16 20.15 30.04 43.10 4.09 0.96 5

274
 Reasons/Factors Strongly disagree…Strongly MS SD Rank
agree
1 2 3 4 5
Lack of commitment from 0.37 6.90 17.72 33.40 41.60 4.09 0.95 5
management
Reluctance to change old 1.87 3.73 18.66 36.19 39.55 4.08 0.95 6
management technique
Lack of coordination of the
implementation of TQM 0.00 9.33 13.62 38.25 38.81 4.07 0.95
policy within the 7
organization
Lack of training and
2.05 2.24 19.59 39.74 36.38 4.06 0.91
inability to train and educate 8
employees on TQM
Complex nature of TQM 0.19 3.54 23.13 36.57 36.57 4.06 0.87 8
technique
Lack of TQM experts 0.00 7.46 21.08 32.09 39.37 4.03 0.95 9
Lack of finance in the 1.12 8.21 20.52 33.77 36.38 3.96 1.00 10
management of TQM
Lack
experts of understanding
among construction 0.00 5.78 24.81 38.99 30.41 3.94 0.88
11
professionals in applying
TQM
Perception that TQM may 3.92 5.41 23.32 34.70 32.65 3.87 1.06 12
not yield any better results
TQM technique is costly 1.87 8.21 24.81 32.65 32.46 3.86 1.02 13
TQM technique is time
0.37 10.07 28.17 33.40 27.99 3.79 0.98
consuming 14

Source: Fieldwork (2018)

11.2.6 PERFORMANCE OF CONSTRUCTION COMPANIES IN GHANA

Table 11.21 indicates the performance of construction companies in Ghana with
performance indicators in terms of percentage responses on a scale of 1 (very poor) to 5

275
(excellent), and a MS ranging between 1.00 and 5.00. All the MSs are above the midpoint
score of 3.00, which indicates that the respondents agreed with the performance indicators
which determine the performance of construction companies in Ghana. It is notable that all
the ten ranked performance indicators which determine the performance of construction
companies in Ghana have a MS > 3.50 ≤ 4.50, which indicates that the respondents
perceive the performance indicators which determine the performance of construction
companies in Ghana to be between ‘average’ and ‘good’. All the rankings of performance
indicators which determine the performance of construction companies in Ghana indicate
MS> 3.82 - 4.12. The relatively high MS achieved suggests that the performance of
construction companies in Ghana is good.

Table 11.21: Performance of Construction Companies in Ghana

Performance indicators Very MS SD Rank

poor…………………Excellent
1 2 3 4 5
Product quality 0.00 0.19 15.67 55.78 28.36 4.12 0.66 1

Productivity levels 0.00 0.37 20.15 49.44 30.04 4.09 0.71 2

Amount of waste produced 0.00 5.78 16.42 45.34 32.46 4.04 0.85 3

Degree of control over the 0.00 2.61 16.79 55.60 25.00 4.03 0.72 4
construction process
Construction cost 0.00 2.24 26.12 40.86 30.78 4.00 0.81 5

Safety standards 0.56 5.97 19.40 41.42 32.65 4.00 0.90 5

Methods of working 0.56 2.61 16.98 61.75 18.10 3.94 0.71 6

Schedule performance 0.00 2.80 24.81 48.51 23.88 3.93 0.77 7
Defect-free product at first 0.00 2.99 32.46 39.74 24.81 3.86 0.82 8
attempt
Reworks elimination 0.00 8.40 25.00 42.91 23.69 3.82 0.89 9

Source: Fieldwork (2018)

276
11.3 INFERENTIAL STATISTICS
11.3.1 Conceptual TQM Model Indicator Variables obtained after Exploratory
Factor Analysis
A total of 536 samples were realized for the Exploratory Factor Analysis (EFA). Eight (8)
factors including their indicator variables were presented for the EFA based on the 536
cases as shown in Table 11.22. The factors are as follows: Factor one (F 1) with nine (9)
indicator variables, Factor two (F 2) with nine (9) indicator variables, Factor three (F 3)
with eight (8) indicator variables, Factor four (F 4) with nine (9) indicator variables, Factor
five (F 5) with seven (7) indicator variables, Factor six (F 6) with ten (10) indicator
variables, Factor seven (F 7) with twelve (12) indicator variables, and Factor eight (F 8)
with ten (10) indicator variables.

Table 11.22: Conceptual Model Indicator Variables

Indicator Measurement Variable

Variable
F1 LTM1 Leadership style of managing employees

F1 LTM2
Leadership ability in solving quality-related problems
F1 LTM3
Leadership initiatives towards TQM
F1 LTM4
Top management knowledge and proper understanding of TQM
F1LTM5
Top management commitment to TQM
F1 LTM6
Top management interaction with workers
F1 LTM7

F1 LTM8 Top management participation in TQM activities

Top management empowerment of employees to solve quality

F1 LTM9
problems

Top management support of TQM

277
Indicator Measurement Variable
Variable
F2 CSQM1 Partnership with suppliers
F2 CSQM2 Supplier selection criteria
F2 CSQM3 Participation of suppliers in TQM activities
F2 CSQM4 Supplier performance evaluation
F2 CSQM5 Supplier quality audit
F2 CSQM6 Supplier communication
F2 CSQM7 Supplier knowledge of TQM
F2 CSQM8 Suppliers’ commitment to TQM
F2 CSQM9 Suppliers’ orientation on TQM

F3 CFI 1 Client brief/input

F3 CFI 2
Client compliant information/feedback
F3 CFI 3
F3 CFI 4 Market investigation
F3 CFI 5
Client satisfaction survey
F3 CFI 6
F3 CFI 7 Quality warrant to client
F3 CFI 8
Client information system
F4 CQSE1 Evaluation of overall company performance
Client services
F4 CQSE2 Evaluation of the quality strategy framework
Client cooperation
F4 CQSE3 Evaluation of quality system procedures

F4 CQSE4 Evaluation of the quality control system

F4 CQSE5 Evaluation of the quality manual

F4 CQSE6 Evaluation of the quality information system

F4 CQSE7 Evaluation of end results

F4 CQSE8 Evaluation of employee performance

278
F4 CQSE9

Evaluation of quality costs

Indicator Measurement Variable
Variable
F5 CVPS Clarity of vision and plan statement

F5 CVPS Good quality policy and driven agenda

F5 CVPS Quality improvement plan

F5 CVPS Formulation of vision and plan

F5 CVPS Concreteness of the future plan

F5 CVPS Employee contribution to the vision

F5 C VPS Involvement of employees in the development of the vision statement

F6 PSDM1 Client brief/input
F6 PSDM2
Cost of selection and design of the product
F6 PSDM3
Environmental issues
F6 PSDM4

F6 PSDM5 Socio-cultural issues

F6 PSDM6 Appearance/finishes required

F6 PSDM7 Strength required for the end product

F6 PSDM8 Intended purpose of the material

F6 PSDM9
Where the material to be used
F6 PSDM10
Design methods/techniques

Design detailing

279
 F7 CPMI1 Project monitoring and control improvement

F7 CPMI2 Equipment maintenance and innovation

F7 CPMI3 Inventory management

F7 CPMI4 Use of quality materials

F7CPMI5 Use of a quality tools

F7CPMI6 Use of quality manual

F7CPMI7 Understanding of work instructions/quality manual

F7CPMI8 Obtaining ISO 9000 certification for operation

F7CPMI9 Quality control system

F7CPMI10 Institution of a continuous improvement process within the company

F7CPMI11 Appropriate use of system structure and standards

F7CPMI12 Efficient management system

F8 CEIM1 Education and training/re-training of staff

F8 CEIM2 Teamwork
F8 CEIM3 Salary/position promotion
F8 CEIM4 Employee reward
F8 CEIM5 Bonus scheme
F8 CEIM6 Conducive working environment
F8 CEIM7 Employee involvement in TQM activities
F8 CEIM8 Availability of a suggestion forum for employees
F8 CEIM9 Employee commitment
F8 CEIM10 Recognition of employees

Source: Fieldwork (2018)

280
11.3.2 Results of Exploratory Factor Analysis Data
The theoretical conceptual model elements were analyzed using EFA. The Social Sciences
(SPSS) version 20 software Package was used to evaluate the reliability, discriminant
validity and convergent validity of the instrument. Principal axis factoring with oblimin
rotation (PAF Oblimin) was the method of factor extraction used to determine the
unidimensionality of the elements. The Bartlett’s test of sphericity and the Kaiser-Meyer-
Olkin (KMO) measure of sampling adequacy was based on the method used by Farrington
(2009) to assess the factor analysability of data. Tabachnick and Fidell (2007) posit that the
KMO should range from 0 to 1 and a minimum value of 0.60 was suggested as good for
factor analysis. A value greater than 0.50 as a minimum cut-off value and a desirable cut-
off value of 0.80 or higher was recommended. Furthermore, Hair et al. (1998) suggested a
cut off value of KMO greater than or equal to 0.70. For the purpose of this study, data with
KMOs of ≥0.70 (p<0.05) was considered factor analysable. In addition, eigenvalue was
computed to establish the factors within the items proposed. Minimum eigenvalue 1 was
considered significant and used to explain the variance captured by a factor. Eigenvalues of
less than 1 were considered insignificant and therefore excluded (Hair et al., 1998).

Exploratory Factor Analysis: Dimensionality of TQM Elements

In the following section, the measures of reliability, convergent and discriminant validity
for each of the factors realized through EFA (Table 11.22) are discussed.

Factor one (F1)

EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. There were nine items measuring
factor one (F1). The result of F1 is reported in Table 11.23. The corrected item-total
correlation was greater than the suggested cut-off value of 0.30, suggesting that the items
were good measures of the element and the Cronbach’s alpha was greater than 0.700 at
0.930, indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of
0.919 with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency
with the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
281
conducted with the data. All the nine items (LTM1, LTM2, LTM3, LTM4, LTM5, LTM6,
LTM7, LTM8 and LTM9) are expected to measure TQM loaded together on this factor.
The factor loadings for all items were greater than 0.705 reported in Table 11.23, which
were greater than the recommended value of 0.40 as suggested by Field (2005) and Hair et
al. (1998).

Table 11.23: Factor one (F 1)

Cronbach's
Corrected
Factor Alpha if
Item-Total
Loading Item
Correlation
Item Questions Deleted
Leadership style of managing 0.823 0.917
LTM1 0.858
employees
Leadership ability in solving quality- 0.726 0.923
LTM2 related problems 0.755

LTM3 Leadership initiatives towards TQM 0.752 0.725 0.923

Top management knowledge and 0.797 0.919
LTM4 proper understanding of TQM 0.831

Top management commitment to TQM 0.750 0.922

LTM5 0.780
Top management interaction with 0.676 0.926
LTM6 workers 0.705

Top management participation in TQM 0.682 0.926

LTM7 activities 0.707
Top management empowerment of 0.755 0.921
LTM8 employees to solve quality problems 0.786

LTM9 Top management support of TQM 0.779 0.748 0.922

Source: Fieldwork (2018)

Factor two (F 2)
EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. There were nine items measuring
factor two (F2). The result of F2 is reported in Table 11.24. The corrected item-total
correlation was greater than the suggested cut-off value of 0.30, suggesting that the items

282
were good measures of the element and the Cronbach’s alpha was greater than 0.700 at
0.911, indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of
0.891 with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency
with the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the nine items (CSQM1, CSQM2, CSQM3, CSQM4, CSQM5,
CSQM6, CSQM7, CSQM8 and CSQM9) are expected to measure TQM loaded together on
this factor. The factor loadings for all items were greater than 0.585 reported in Table
11.24, which were greater than the recommended value of 0.40 as suggested by Field
(2005) and Hair et al. (1998).

Table 11.24: Factor two (F 2)

Cronbach's
Corrected
Factor Alpha if
Item-Total
Loading Item
Correlation
Item Questions Deleted
CSQM1 Partnership with suppliers 0.697 0.666 0.903

CSQM2 Supplier selection criteria 0.675 0.648 0.905

Participation of suppliers in TQM 0.734 0.899
CSQM3 0.773
activities
CSQM4 Supplier performance evaluation 0.834 0.791 0.894
CSQM5 Supplier quality audit 0.585 0.560 0.910

CSQM6 Supplier communication 0.686 0.655 0.904

CSQM7 Supplier knowledge of TQM 0.810 0.764 0.896
CSQM8 Suppliers’ commitment to TQM 0.773 0.717 0.900

CSQM9 Suppliers’ orientation on TQM 0.739 0.702 0.901

Source: Fieldwork (2018)

283
Factor three (F 3)
EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. There were eight items measuring
factor three (F3). The result of F3 is reported in Table 11.25. The corrected item-total
correlation was greater than the suggested cut-off value of 0.30, suggesting that the items
were good measures of the element and the Cronbach’s alpha was greater than 0.700 at
0.891, indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of
0.894 with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency
with the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the eight items (CFI1, CFI2, CFI3, CFI4, CFI5, CFI6, CFI7,
and CFI8) are expected to measure TQM loaded together on this factor. The factor loadings
for all items were greater than 0.603 reported in Table 11.25, which were greater than the
recommended value of 0.40 as suggested by Field (2005) and Hair et al. (1998).

Table 11.25: Factor three (F 3)

Cronbach's
Corrected
Factor Alpha if
Item-Total
Loading Item
Correlation
Item Questions Deleted
CFI1 Client brief/input 0.768 0.716 0.872
CFI2 Client compliant information/feedback 0.763 0.711 0.873
CFI3 Market investigation 0.728 0.688 0.875
CFI4 Client satisfaction survey 0.723 0.674 0.877
CFI5 Quality warrant to client 0.651 0.610 0.882
CFI6 Client information system 0.776 0.728 0.871
CFI7 Client services 0.603 0.567 0.887

CFI8 Client cooperation 0.676 0.638 0.880

Source: Fieldwork (2018)

284
Factor four (F 4)
EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. There were nine items measuring
factor four (F4). The result of F4 is reported in Table 11.26. The corrected item-total
correlation was greater than the suggested cut-off value of 0.30, suggesting that the items
were good measures of the element and the Cronbach’s alpha was greater than 0.700 at
0.915, indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of
0.914 with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency
with the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the nine items (CQSE1, CQSE2, CQSE3, CQSE4, CQSE5,
CQSE6, CQSE7, CQSE8 and CQSE9) are expected to measure TQM loaded together on
this factor. The factor loadings for all items were greater than 0.676 reported in Table
11.26, which were greater than the recommended value of 0.40 as suggested by Field
(2005) and Hair et al. (1998).

Table 11.26: Factor four (F4)

Cronbach's
Corrected
Factor Alpha if
Item-Total
Loading Item
Correlation
Item Questions Deleted
Evaluation of the quality strategy 0.684 0.906
CQSE1 0.721
framework
Evaluation of overall company 0.702 0.905
CQSE2 performance 0.739

CQSE3 Evaluation of employee performance 0.772 0.732 0.903

CQSE4 Evaluation of quality costs 0.744 0.709 0.904

Evaluation of the quality manual 0.746 0.902
CQSE5 0.785
Evaluation of the quality control system 0.749 0.901
CQSE6 0.787

Evaluation of quality system 0.724 0.692 0.906

CQSE7
procedures

285
 CQSE8 Evaluation of end results 0.676 0.646 0.909
Evaluation of the quality information 0.656 0.908
CQSE9 0.687
system
Source: Fieldwork (2018)

Factor five (F5)

EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. There were seven items measuring
factor five (F 5). The result of F 5 is reported in Table 11.27. The corrected item-total
correlation was greater than the suggested cut-off value of 0.30, suggesting that the items
were good measures of the element and the Cronbach’s alpha was greater than 0.700 at
0.897, indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of
0.875 with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency
with the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the seven items (CVPS1, CVPS2, CVPS3, CVPS4, CVPS5,
CVPS6 and CVPS7) are expected to measure TQM loaded together on this factor. The
factor loadings for all items were greater than 0.673 reported in Table 11.27, which were
greater than the recommended value of 0.40 as suggested by Field (2005) and Hair et al.
(1998).

Table 11.27: Factor five (F 5)

Cronbach's
Corrected
Factor Alpha if
Item-Total
Loading Item
Correlation
Item Questions Deleted
Clarity of vision and plan statement 0.737 0.878
CVPS1 0.787
Good quality policy and driven agenda 0.651 0.888
CVPS2 0.691

CVPS3 Quality improvement plan 0.673 0.632 0.890

CVPS4 Formulation of vision and plan 0.703 0.666 0.886

286
 Concreteness of the future plan 0.734 0.878
CVPS5 0.779

CVPS6 Employee contribution to the vision 0.829 0.777 0.873

Involvement of employees in the 0.710 0.881
CVPS7 0.755
development of the vision statement
Source: Fieldwork (2018)

Factor six (F6)

EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. At the end of the EFA, F6 had two
factor loadings which were relabelled as F6A – Product Design Management (PDM) and F6
B – Product Selection Management (PSM). There were six and four items measuring F6 A
and F6 B respectively.

Factor six A (F6 A)

The result of F 6A is reported in Table 11.28. The corrected item-total correlation was
greater than the suggested cut-off value of 0.30, suggesting that the items were good
measures of the element and the Cronbach’s alpha was greater than 0.700 at 0.849,
indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of 0.882
with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency with
the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the six items (PDM1, PDM2, PDM3, PDM4, PDM5 and
PDM6) are expected to measure TQM loaded together on this factor. The factor loadings
for all items were greater than 0.460 reported in Table 11.28, which were greater than the
recommended value of 0.40 as suggested by Field (2005) and Hair et al. (1998).

287
Table 11.28: Product Design Management (PDM) - Factor six (F6A)

Corrected Cronbach's
Factor
Item-Total Alpha if
Loading
Relabeled Correlation Item Deleted
Item item Questions
PSDM4 PDM1 Socio-cultural issues 0.460 0.541 0.842
PDM2 Strength required for the
PSDM6 0.755 0.693 0.812
end product
PDM3 Intended purpose of the
PSDM7 0.659 0.651 0.820
material
PDM4 Where the material to be
PSDM8 0.740 0.613 0.828
used
PDM5 Design
PSDM9 0.680 0.632 0.824
methods/techniques
PSDM10 PDM6 Design detailing 0.728 0.668 0.817
Source: Fieldwork (2018)

Factor six B (F6 B)

The result of F6 B is reported in Table 11.29. The corrected item-total correlation was
greater than the suggested cut-off value of 0.30, suggesting that the items were good
measures of the element and the Cronbach’s alpha was greater than 0.700 at 0.858,
indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of 0.882
with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency with
the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the four items (PSM1, PSM2, PSM3 and PSM4) are expected
to measure TQM loaded together on this factor. The factor loadings for all items were
greater than 0.624 reported in Table 11.29, which were greater than the recommended value
of 0.40 as suggested by Field (2005) and Hair et al. (1998) except item PSM3
(Environmental issues) which recorded factor loading of 0.384 below the recommended
value of 0.40.

288
Table 11.29: Product Selection Management (PSM) - Factor six (F6B)

Corrected Cronbach's
Factor
Item-Total Alpha if
Loading
Relabeled Questions Correlation Item Deleted
Item item
PSDM1 PSM1 Client brief/input 0.910 0.744 0.802
Cost of selection and
PSDM2 PSM2 design of the product 0.839 0.776 0.789

PSDM3 PSM3 Environmental issues 0.384 0.572 0.872

PSM4 Appearance/finishes 0.624
PSDM5 0.733 0.808
required
Source: Fieldwork (2018)

Factor seven (F7)

EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. There were twelve items measuring
factor eight (F8). The result of F 8 is reported in Table 11.30. The corrected item-total
correlation was greater than the suggested cut-off value of 0.30, suggesting that the items
were good measures of the element and the Cronbach’s alpha was greater than 0.700 at
0.930, indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of
0.902 with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency
with the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the twelve items (CPMI1, CPMI2, CPMI3, CPMI4, CPMI5,
CPMI6, CPMI7, CPMI8, CPMI9, CPMI10, CPMI11 and CPMI12) are expected to measure
TQM loaded together on this factor. The factor loadings for all items were greater than
0.645 reported in Table 11.30, which were greater than the recommended value of 0.40 as
suggested by Field (2005) and Hair et al. (1998).

289
Table 11.30: Factor seven (F7)

Cronbach's
Corrected
Factor Alpha if
Item-Total
Loading Item
Correlation
Item Questions Deleted
Project monitoring and control 0.689 0.924
CPMI1 0.719
improvement
Equipment maintenance and innovation 0.622 0.926
CPMI2 0.645

CPMI3 Inventory management 0.663 0.638 0.926

CPMI4 Use of quality materials 0.739 0.709 0.923

CPMI5 Use of a quality tools 0.675 0.653 0.925

CPMI6 Use of quality manual 0.708 0.684 0.924

Understanding of work 0.732 0.922
CPMI7 0.762
instructions/quality manual
Obtaining ISO 9000 certification for 0.688 0.925
CPMI8 0.718
operation
CPMI9 Quality control system 0.804 0.774 0.921
Institution of a continuous 0.725 0.922
CPMI10 improvement process within the 0.754
company
Appropriate use of system structure and 0.737 0.922
CPMI11 0.766
standards
CPMI12 Efficient management system 0.752 0.721 0.923
Source: Fieldwork (2018)

Factor eight (F8)

EFA was conducted to assess the unidimensionality and reliability of TQM. PAF Oblimin
was specified as the extraction and rotation method. There were ten items measuring factor
nine (F8). The result of F 8 is reported in Table 11.31. The corrected item-total correlation
was greater than the suggested cut-off value of 0.30, suggesting that the items were good
measures of the element and the Cronbach’s alpha was greater than 0.700 at 0.909,
indicating acceptable internal reliability (Nanually & Bernstein, 1994). The KMO of 0.871
with Bartlett’s test of sphericity of p<0.000 was also obtained, indicating consistency with
290
the recommended KMO cut-off value of 0.70 and Bartlett's test of sphericity of p<0.05
suggested by Hair et al. (2010). These results suggested that factor analysis could be
conducted with the data. All the ten items (CEIM1, CEIM2, CEIM3, CEIM4, CEIM5,
CEIM6, CEIM7, CEIM8, CEIM9 and CEIM10) are expected to measure TQM loaded
together on this factor. The factor loadings for all items were greater than 0.660 reported in
Table 11.31, which were greater than the recommended value of 0.40 as suggested by Field
(2005) and Hair et al. (1998).

Table 11.31: Factor eight (F8)

Cronbach's
Corrected
Factor Alpha if
Item-Total
Loading Item
Correlation
Item Questions Deleted
Education and training/re-training of 0.653 0.902
CEIM1 0.694
staff
CEIM2 Teamwork 0.715 0.683 0.900
CEIM3 Salary/position promotion 0.709 0.674 0.900
CEIM4 Employee reward 0.826 0.789 0.892
CEIM5 Bonus scheme 0.688 0.656 0.901
CEIM6 Conducive working environment 0.737 0.705 0.898
Employee involvement in TQM 0.627 0.902
CEIM7 0.663
activities
Availability of a suggestion forum for 0.654 0.901
CEIM8 0.691
employees
CEIM9 Employee commitment 0.725 0.685 0.899

CEIM10 Recognition of employees 0.660 0.623 0.903

Source: Fieldwork (2018)

11.4 THE STRUCTURAL EQUATION MODELING (SEM)

The reason why SEM is preferred to other statistical approaches such as ANOVA and
regression is because it displays better conditions to demonstrate causality. Three
conditions to demonstrate causality are association, isolation and directionality (Hoyle,
1995:10). SEM is not distinctive in the first aspect for isolating and putative causes, it is
291
more flexible and comprehensive than any univariate or multivariate modelling approaches,
and SEM provides a means of controlling not only for extraneous or confounding variables,
but for measurement error as well. Directionality is often greatly misunderstood. SEM can
be shown with many statistical procedures because it comes from theory (research design)
and sample logic. In the case of a model as a whole, SEM produces a good-fit and the result
greatly supports the individual causal relationships within the model. The current research
has clearly shown that measuring TQM is a complex construct. Therefore, in order to
examine the factors that determine TQM in the Ghanaian construction industry, EQS
Version 6.2 software was used to investigate the measurement model adequacy and
structural model goodness-of-fit.

11.4.1 Structural Equation Modelling Analytic Strategy

This study aims to test a model of TQM for the Ghanaian construction industry. Three steps
were conducted in the analysis which includes EFA, CFA and SEM using EQS software
Version 6.2 for analysis. First, a series of CFA was used to test for the measurement
equivalency for each of the nine latent constructs in Table 11.33 and the manifest or
composite variables of TQM in the hypothesized model of TQM. The CFA results defined
the relations between the observed and unobserved variables. The CFA provided the link
between scores on a measuring instrument and the underlying constructs they are designed
to measure. The CFA was carried out to reaffirm the factor structure of the observed and
unobserved variables, hence the construct validity. Secondly, the fit of the entire
measurement model underlying the hypothesized structural model was tested. The structural
model defined the relationship amongst the different exogenous variables and specified the
manner by which each exogenous variable directly or indirectly influences the changes in
the values of other exogenous constructs in the model.

Thus, the endogenous variables were defined (TQM outcomes). All analyses were
performed using Eqations software (EQS), including the testing of the hypothesized SEM.
In SEM, a covariance matrix generated from a particular sample is compared with the
covariance matrix generated from the hypothesized model and fit statistics are used to
determine the acceptability of the solution obtained. Hu and Bentler (1999) have
292
recommended using a combination of fit statistics to evaluate the fit of models (Table
11.34) as adopted for the current study.

11.4.2 Conceptual TQM Model Indicator Variables obtained after CFA

Table 11.33 gives details of the selected model after series of CFA on the nine (9) factors
realized from the EFA. The model was realized during the CFA through the process of
removing cases to achieve stronger models fit statistics. At the conclusion of the stated
process, a nine-factor model was realized. All the eight original conceptual TQM latent
constructs’ names from Figure 10.1 (Model 1.0) were retained except that the latent
construct with the name ‘Product Selection and Design Management (PSDM)’ had two
factor loadings during EFA and was therefore separated with the following assigned names:
i. Product Selection Management (PSM)
ii. Product Design Management (PDM)
Substituted ‘Product Selection and Design Management (PSDM)’ with the two new latent
constructs (‘Product Selection Management’ and ‘Product Design Management’), the
number of the latent constructs of the original conceptual TQM model in Figure 10.1
(Model 1.0) increased from eight to nine, hence a nine-factor model was realized. The
nine-factor model therefore was deemed fit for the current study and presented in Table
11.32.

Table 11.32: Conceptual Model Indicator Variables

Latent Indicator Measurement Variable Label

Construct Variable
Leadership and F1 LTM1 Leadership style of managing LTM1
Top employees
Management
(LTM) F1LTM4 Top management knowledge and LTM4
proper understanding of TQM

F1LTM8 Top management empowerment of LTM8

employees to solve quality problems

F1LTM9 Top management support of TQM LTM9

293
Latent Indicator Measurement Variable Label
Construct Variable
Company F2CSQM4 Supplier performance evaluation CSQM4
Supplier
Quality F2CSQM7 Supplier knowledge of TQM CSQM7
Management
(CSQM) F2CSQM8 Suppliers’ commitment to TQM CSQM8

Client Focus F3CFI1 Client brief/input CFI1

and
F3CFI2 Client complaint information/feedback CFI2
Involvement
F3CFI4 Client satisfaction survey CFI4
(CFI)

Company F4CQSE1 Evaluation of the quality strategy CQSE1

Quality System framework

Evaluation F4CQSE2 Evaluation of overall company CQSE2

(CQSE) performance
F4CQSE3 CQSE3
Evaluation of employee performance

Company F5CVPS5 Concreteness of the future plan CVPS5

Vision and Plan
F5CVPS6 Employee contribution to the vision CVPS6
Statement
F5CVPS7 Involvement of employees in the CVPS7
(CVPS)
development of the vision statement

Product Design F6A PDM2 Strength required for the end product PDM2
Management
F6A PDM3 Intended purpose of the material PDM3
(PDM)
F6A PDM6 PDM6
Design detailing

Product F6B PSM1 Client brief/input PSM1

Selection
F6B PSM2 Cost of selection and design of the PSM2
Management product
(PSM)
F6B PSM4 Appearance/finishes required PSM4

294
 Construction F7CPMI9 Quality control system CPMI9
Process
F7CPMI10 Institution of a continuous improvement CPMI10
process within the company
Management
F7CPMI11 Appropriate use of system structure and CPMI11
and
standards
Improvement
F7CPMI12 Efficient management system CPMI12
(CPMI)
Construction F8CEIM3 Salary/Position promotion CEIM3
Employees’
F8CEIM4 Employee rewards CEIM4
Involvement
F8CEIM5 Bonus scheme
and Motivation CEIM5
(CEIM)
TQM Outcomes F9TQMO1 Greater reduction of construction costs TQMO1
(TQMO) Elimination of reworks
F9TQMO4 TQMO4
Defect-free product at first attempt
F9TQMO5
TQMO5
F9TQMO6 Higher safety standards
TQMO6
Source: Fieldwork (2018)

11.4.3 The Measurement Model Confirmatory Factor Analysis

After the constructs had demonstrated sufficient evidence of unidimensionality and
reliability using EFA, a CFA was then administered. EQS version 6.2 with maximum
likelihood robust estimator was used to analyze the construct validity of the measurement
models. The types of goodness-of-fit indices and their acceptable cut-off values selected for
this study are shown in Table 11.33.

295
Table 11.33: Cut-off Criteria for Fit Indices

Fit index Acronym Acceptable thresholds Reference

for continuous data
Chi-square test x² test Low x² relative df with Hooper et al.
an (2008);
insignificant p- Hsu et al. (2012)
Normed chi-square value of (x²) to df ≤2 or 3
x²/df ratio Ratio Hsu et al., (2012)
good fit
(p>0.05) Kline (2005)

Ratio of (x²) to df
≤5 acceptable
Root-mean squared- RMSEA Values less than 0.05 Schreiber et
error of with al.
approximatio confidence interval (2006); Hsu et
n (CI) al. (2012
0.00-0.05 “good fit” Values
greater than 0.06 to
0.08 with confidence
interval
0.00-0.05 “acceptable
fit”

Values greater than 0.08

to
1.00 with confidence
interval
0.00-0.05 “mediocre
fit”

296
 Fit index Acronym Acceptable thresholds Reference
for continuous data
Comparative fit CFI Equal or greater than 0.95 Schreiber et al.
index “good fit” (2006)
Hu & Bentler (1999)
Equal or greater than
0.90 “acceptable fit”
Tucker-Lewis index TLI/ Equal or greater than 0.95 Schreiber et al.
NNFI “good fit” or (2006)
Equal or greater than Hu & Bentler (1999)
0.90 “acceptable fit”
Standardized root SRMR Equal or less than 0.05 Schreiber et
mean “good al.(2006)
square fit” Hu & Bentler (1999)
residual
Equal or less than
0.08 “acceptable fit”
Normed fit index NFI Greater than 0.90 “good fit” Bentler and Bonnet
(1980)
Source: Fieldwork (2018)

Statistics on SEM Assumptions - Outliers and Missing Data

Data sets investigation revealed that some data sets had missing values. A detailed
examination of the pattern of missing data revealed that the missing data was missing at
random (MAR) and not missing completely at random (MCAR). “The condition that data
was missing completely at random is a situation where the presence or absence of the
observation is independent of other observed variables and the variable itself” (McDonald
& Ho, 2002:70). The condition of MCAR is a very strict assumption that may be difficult to
justify in practice as suggested by McDonald and Ho (2002). Therefore, the assumption of
the condition of MAR was adopted. Hence, the robust maximum likelihood estimation
solution in EQS was used to address the problem. This method produces better results
compared to other methods (Kline, 2005). The assumption in this method was that the
means, variances and covariances were sufficient statistics. Consequently, cases with
missing variables were skipped and not included in the analysis. Further examination of the
data set revealed that there were a few outliers in the data. The EQS result output included
case numbers with the largest contribution to Yuan, Lambert and Fouladi’s normalized

297
multivariate kurtosis. Examination of these case numbers showed the case numbers that
include outliers and it was upon these inspections that the conclusion was based. It was
concluded that there were a few outliers in the data. The robust maximum likelihood (RML)
was adopted for estimation and it happened to be adequate in addressing the problems of
outliers.

Statistics on SEM Assumptions – Data Distribution Characteristics

The model analysis began after the distribution characteristics of the data had been
established through the maximum likelihood estimation. This method assumes multivariate
normality. The EQS result output included univariate statistics such as mean, skewness,
kurtosis and the respective standard deviations. Similarly, the multivariate kurtosis formed
part of the result output. Analysis of the univariate statistics and Yuan, Lambert and Fouladi
based on normalized multivariate kurtosis suggested non-normality in the sample data set.
The result shows that all the Yuan, Lambert and Fouladi estimates of normalized
multivariate kurtosis were greater than the upper limit value of 3.0 (DeCarlo, 1997:292).
The outcome of the results led to descriptions of the data being highly kurtotic. However,
the adoption of the robust maximum likelihood estimation method of the postulated model
was due to the non-normality of the data. The results in the preceding sections as indicated
by Satorra- Bentler scaled statistics (Satorra & Bentler, 1988) are reported using robust
statistics for the chi-square. The first item of each factor is fixed to establish the factors’
scale in the models. Yuan, Lambert and Fouladi’s coefficient and other univariate statistics
are presented in Table 11.34.

Statistics on SEM Assumptions – Identifiability of the Model

According to Boomsma (2000:466), “…the researcher has the responsibility of examining a
model to ascertain whether it is theoretically identified or not”. Therefore, a SEM analysis
was conducted to identify the structural model. In addition to this, it must fulfill the
conditions of model identification.

298
Table 11.34: Univariate Statistics and Yuan, Lambert and Fouladi’s Normalized
Multivariate Estimates

Latent Indicator Mean Skewness Kurtosis SD Yuan,

(σx)
Lambert
(x ) (G1) (G2)
Construct Variable and
Fouladi’s
Leadership LTM1 4.10 -0.608 -0.414 0.851
Coefficie
and Top
Manageme LTM4 4.10 -0.852 0.141 0.946 nt
nt (LTM)
LTM8
4.02 -0.376 -0.844 0.862
LTM9
4.18 -0.744 -0.581 0.917
Company CSQM4 4.08 -0.686 -0.414 0.949
Supplier
CSQM7 4.10 -0.592 -0.689 0.918
Quality
Management CSQM8 4.06 -0.598 -0.547 .0.942
(CSQM) Focus CFI1
Client 4.01 -0.369 -0.725 0.837
and
CFI2 4.03 -0.305 -0.884 0.822
Involvement
(CFI) CFI4 4.06 -0.725 -0.211 0.892
Company CQSE1 4.26 -0.821 0.197 0.769
Quality System
CQSE2 4.29 -0.864 0.254 0.763
Evaluation
(CQSE) CQSE3 4.16 -0.815 0.127 0.846

299
 Company CVPS5 4.05 -0.585 -0.657 0.923
Vision and Plan CVPS6 4.08 -0.581 -0.838 0.949
Statement CVPS7 4.20 -0.673 -0.491 0.835
(CVPS)
Product Design PDM 2 4.21 -0.524 -0.797 0.788
Management 4.15 -0.589 -0.531 0.826
PDM 3
(PDM) 4.15 -0.281 -1.168 0.765
PDM 6

Product PSM 1 4.07 -1.012 0.811 0.945

Selection PSM 2 4.11 -0.738 -0.081 0.852
Management PSM 4 4.16 -0.639 -0.371 0.824
(PSM)
Construction -0.752 -0.119
CPMI9 4.20 0.833
Process -0.549 -0.546
CPMI10 4.11 0.838
Management -0.543 -0.551
and CPMI11 4.15 0.804
-0.795 -0.21
Improvement CPMI12 4.24 0.830
(CPMI)
Construction CEIM3 4.13 -0.687 -0.266 0.850
Employees’
CEIM4 4.10 -1.008 0.607 0.944
Involvement
CEIM5
and Motivation 4.12 -0.845 -0.216 0.950
(CEIM
TQM outcomes TQMO1 4.09 -1.120 0.844 1.036
(TQMO) 180.239
TQMO4 4.11 -0.891 0.578 0.891

TQMO5 4.18 -0.529 -0.992 0.859

TQMO6 4.26 -0.592 -0.425 0.730

Source: Fieldwork (2018) 3.628

300
A model can be analysed when it has been identified. Therefore, it is necessary to carry out such
an identification process. The section meant for results shows the identification (property of the
model) test. As argued by Kline (2005:105), a model is said to be identified if it is theoretically
possible to derive a unique estimate for each parameter. Kline (2005) further argued that the
sample size does not matter if a model is not identified and it will be difficult to analyze it. A
model can then be said to be identified if it has at least as many observations as free model
parameters (namely, the degree of freedom ≥ 0) and that every unobserved variable must be
assigned a scale (Kline, 2005:105). However, a model could be just-identified, over-identified or
under-identified (Byrne, 2006:31). According to Byrne (2006:31), “…an over-identified model is
one in which the number of parameters to be estimated is less than the number of data variances
and covariances of the observed variables and therefore results in a positive degree of freedom”.
Kline (2005:31) posits that the significance of an over-identified model is that it allows for a
model to be rejected, thereby rendering it of scientific value. A just-identified model cannot be
rejected and it is not possible to obtain a solution for an under-identified model. The EQS result
outputs showed that the lowest value for the degree of freedom was 0.0 and the highest value
was 2.0 for the current study. These results show an over-identified model because the scores
showed a positive value of degree of freedom.

11.4.4 Confirmatory Factor Analysis of the Latent Construct

The first step in assessing measurement invariance as indicated by Byrne (2006) is to conduct
separate CFAs of the latent constructs. Therefore, a CFA was carried out on the exogenous
variables (Leadership and Top Management, Company Supplier Quality Management, Client
Focus and Involvement, Company Quality System Evaluation, Company Vision and Plan
Statement, Product Selection Management, Product Design Management, Construction Process
Management and Improvement and Construction Employees’ Involvement and Motivation) and
the endogenous variable (TQM Outcomes) to determine whether the measures used were
sufficient indicators, to assess the coefficients and to reaffirm the factor structure of each
construct. The fit of the items to the latent variables was further conducted with EQS 6.2
Statistical Software to explore the measurement model. If the fit of each of these models is good
and the item loading is acceptable, it can be assumed that the indicators underlying the factor are
301
tapping into the construct at hand in each of the latent constructs. This is in line with the practice
established by McDonald and Ho (2002) and supported by the SEM experts’ recommendations.
The evaluation of models as attested by the experts should be derived from a large number of
criteria rather than a single “magic index” (Kline, 2005; Byrne, 2006). Various goodness-of-fit
indices were considered in the study to determine the goodness-of fit.

11.4.4.1 Fit Statistics on Measurement Models (CFA)

11.4.4.1.1 Measurement Model for Leadership and Top Management Features (LTM)
Construct
A unidimensional model for Leadership and Top Management (LTM) features is presented
(Table 11.33). From the 536 cases analyzed for this construct, four (4) indicator variables
(F1LTM1, F1LTM4, F1LTM8, and F1LTM9) made up of the same factor were realized and the
name LTM was maintained (Table 11.33). All four (4) indicator variables obtained were used for
the CFA (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a variable to be included in a
CFA, thus enabling the model to be described as well-fitting, the distribution of residuals
covariance matrix should be symmetrical and centred around zero (Byrne, 2006:94; Joreskog &
Sorbom, 1988). The four-indicator model provides good measures of residual matrix and
evidence of convergent validity.

Figure 11.1: Measurement Model of Leadership and Top Management

302
The CFA results further revealed that the Leadership and Top Management features had four (4)
dependent variables, five (5) independent variables and eight (8) free parameters. The number of
fixed non-zero parameters was five (5). These are the four (4) dependent indicator variables for
the Leadership and Top Management: Leadership style of managing employees, Top
management knowledge and proper understanding of TQM, Top management empowerment of
employees to solve quality problems, and Top management support of TQM. These indicator
variables are presented in Table 11.36. The Leadership and Top Management features
measurement model shown in Figure 11.1 was analyzed before it could be included in the full
latent variable model. In order to establish how well the model fit the sample data and the
strength of the hypothesized relationship between the variables, results on residual covariance
matrix (unstandardized and standardized), distribution of standardized residuals, fit statistics and
statistical significance at a probability level of 5 per cent were examined. In addition, the
Cronbach’s alpha and the rho coefficient of internal consistency were examined to determine the
score reliability. Results of these statistics are presented in the next section for the Leadership
and Top Management features.

Table 11.35: Postulated Leadership and Top Management Features Model

Latent constructs Indicator variables Label

Leadership and Top Leadership style of managing LTM1
Management (LTM) employees
Top management knowledge and LTM4
proper understanding of TQM
Top management empowerment of LTM8
employees to solve quality
problems
Top management support of TQM LTM9

Source: Fieldwork (2018)

303
Diagnostic Fit analysis: Analysis of Residual Covariance Estimate
The unstandardized and standardized absolute residual matrix values of the Leadership and Top
Management features are presented in Table 11.36 and 11.37. The result reveals that all the
absolute residual values and the average off-diagonal absolute residual values were close to zero.
The unstandardized average off-diagonal residual was 0.0002 while the standardized average off
diagonal residual was found to be 0.0015. A residual value greater than 2.58 is described as large
(Byrne, 2006:94). The results obtained for the Leadership and Top Management features
measurement model suggested a fairly acceptable fit to the sample data because the absolute
residual were all less than 2.58. In order for a model to be described as well-fitting, the
distribution of standardized residuals should be symmetrical and centred around zero (Byrne,
2006:94).

Table 11.36: Residual Covariance Matrix for Leadership and Top Management Model
(Unstandardized)

Unstandardized Residual Covariance Matrix

LTM4 LTM9 LTM8 LTM1

LTM4 0.000
LTM9 -0.019 0.000
LTM8
0.008 0.008 0.000
LTM1 0.005 0.01 -0.013 0.000
Average absolute residual = 0.0400
Average off-diagonal absolute residual = 0.0002

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

304
Table 11.37: Residual Covariance Matrix for Leadership and Top Management Model

Standardized Residual Covariance Matrix

LTM4 LTM9 LTM8 LTM1

LTM4 0.000
LTM9 0.053 0.000
LTM8 0.007 0.021 0.000
LTM1 -0.019 -0.021 -0.007 0.000

Average absolute residual = 0.0009

Average off-diagonal absolute residual = 0.0015

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit, further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)

The analysis strategy of goodness-of-fit for the Leadership and Top Management followed a
three-statistics strategy of fit indexes as recommended by Hu and Bentler (1999). The sample
data on Leadership and Top Management measurement model yielded the S – Bχ2 of 6.827 with
2 degrees of freedom (df) with a probability of p = 0.033. This chi-square value indicated that
the departure of the sample data from the postulated measurement model was significant and
hence indicative of good fit. The chi-square test is very sensitive to sample size and is used more
as a descriptive index of fit rather than as a statistical test (Kline, 2005:136). The normed chi-
square value is usually adopted by most researchers. The normed chi-square is the procedure of
dividing the chi-square by the degrees of freedom. The normed values of up to 3.0 or even 5.0
are recommended (Kline, 2005:137). Values for NFI range between 0 and 1 with Bentler and
Bonnet (1980) recommended values greater than 0.90 indicating a good fit. Hu and Bentler
305
(1999) have given cut-off criteria of NFI ≥ 0.95. This index is sensitive to sample size,
underestimating fit for samples less than 200 (Mulaik et al., 1989; Bentler, 1990) and is thus not
recommended to be solely relied on (Kline, 2005). The non-normed fit index (NNFI, also known
as the Tucker-Lewis index), is an index that prefers simpler models. The value of the NNFI can
indicate poor fit despite other statistics pointing towards good fit if small samples are used
(Bentler, 1990; Kline, 2005; Tabachnick & Fidell, 2007). Owing to the non-normed nature,
NNFI values can go above 1.0 and can thus be difficult to interpret (Byrne, 1998). The chi-
square and degrees of freedom were found to be 3.414. This ratio was within the limit of 3.00 or
5.0 advocated for by some authors (Kline, 2005:137). The CFI value was found to be 0.996
which was higher than the cut-off limit of 0.95 so the model is described to have a good fit. The
NFI value was 0.995 which is above the given cut-off value of NFI ≥ 0.95 and is shown in Table
11.39. Therefore, the model is acceptable. The NNFI value obtained is 0.988 which is also above
the cut-off value of 0.80. These fit indexes for the Leadership and Top Management model
suggested that the postulated model adequately describes the sample data and could therefore be
included in the full latent variable model analysis (Table 11.38).

Table 11.38: Robust Fit Indexes for Leadership and Top Management Features Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 6.827

df 0≥ 2 Good fit

Normed chi-square ≤2 or 3 good fit 3.414 Good fit

= x²/df ratio ≤5 acceptable
CFI 0.90≥ acceptable 0.996 Good fit

0.95≥ good fit

RMSEA Less than 0.05 with 0.067
confidence interval
(CI) 0.00-0.05 Acceptable
“good fit”
0.95≥ good fit

95% 306
NFI Greater than 0.90 0.995 Good fit
“good fit”

NNFI Greater than 0.988 Good fit

0.80. “good fit”

RMSEA 95% CI 0.016: 0.125 Slightly out

of range

SRMR Equal or less than 0.05 0.009 Good fit

“good
fit”

Equal or less
than 0.08 “acceptable
fit” 0.125

Source: Fieldwork (2018)

0.125
Statistical Significance of Parameter Estimates
Table 11.39 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and
show appropriate signs. The estimates were therefore deemed reasonable as well as
statistically significant. The parameter with the highest standardized coefficient was the
indicator with variable LTM1 and LTM4 (Leadership style of managing employees and
Top management knowledge and proper understanding of TQM) and its parameter
coefficient was 0.847.

307
Table 11.39: Factor Loading and Z-statistics of Leadership and Top Management
Measurement

Indicator Unstandardized Standardized Z- R2 Significant at

Statistics 5% level?
Variable Coefficient (λ) Coefficient
(λ)
LTM1 1.000 0.847 4.132 0.717 Yes

LTM4 1.111 0.847 3.361 0.718 Yes

LTM8 0.988 0.826 3.500 0.682 Yes

LTM9 0.995 0.783 2.417 0.613 Yes

(Robust Statistical Significance at 5% level)

Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Leadership and Top Management features). In addition, the R2 values were
also close to the desired value of 1.00, indicating that the factors explained more of the variance
in the indicator variables. The results therefore suggest that the indicator variables significantly
predict the unobserved construct because all the measured variables are significantly associated
with the Leadership and Top Management features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Leadership and Top Management
features construct were determined from the rho and the Cronbach’s alpha coefficient. According
to Kline (2005:59), the reliability coefficient should fall between zero and 1.00. Values close to
1.00 are desired. The rho coefficient of internal consistency was found to be 0.801. This value
was above the minimum required value of 0.70. Likewise, the Cronbach’s alpha was above the
minimum acceptable value of 0.70. The Cronbach’s alpha was found to be 0.895 (Table 11.40).

308
Both of these values revealed a high level of internal consistency and therefore reliability,
suggesting that the indicator variables represent the same latent construct (Leadership and Top
Management).

Table 11.40: Reliability and Construct Validity of Leadership and Top Management
Feature Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient

Leadership and LTM1 0.847

Top Management
LTM4 0.847
0.895 0.801
LTM8 0.826

LTM9 0.783
*Parameter estimates are based on standardized solutions
Source: Fieldwork (2018)

Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher, and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111). The standardized
parameter coefficient presented in Table 11.40 revealed that all coefficients were significantly
higher with the lowest being 0.783 for Leadership and Top Management features. The magnitude
of the parameter estimate was above the 50 per cent minimum. This indicates a strong
relationship between the indicator variables and the factors of the Leadership and Top

309
Management features construct. Therefore the Leadership and Top Management features
satisfied both internal reliability and the construct criteria. The rho value was above the
minimum value of 0.70 and the magnitude, signs and statistical significance of the parameter
estimates were appropriate (Table 11.41).

Summary on Leadership and Top Management Feature Measurement Model

The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indices met the cut-off index criteria and all the parameter estimates were
statistically significant and feasible. Considering these criteria, the measurement model for the
Leadership and Top Management feature was found to be adequately fit the sample data.
Therefore, there was no need to improve the measurement model before it could be included in
the full latent variable model. Hence, the Leadership and Top Management feature construct was
adequately measured by the indicator variables and could be used in the analysis of the full latent
variable model.

11.4.4.1.2 Measurement Model for Company Supplier Quality Management Features

(CSQM) Construct.
A unidimensional model for Company Supplier Quality Management (CSQM) features is
presented (Table 11.33). From the 536 cases analyzed for this construct, three (3) indicator
variables (F2CSQM4, F2CSQM7, and F2CSQM8) made up of the same factor were realized and
the name CSQM was maintained (Table 11.33). All the three (3) indicator variables obtained
were used for the CFA (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a variable to be
included in a CFA, thus enabling the model to be described as well-fitting, the distribution of
residuals covariance matrix should be symmetrical and centred around zero (Byrne, 2006:94;
Joreskog & Sorbom, 1988). The three-indicator model provides good measures of residual
matrix and evidence of convergent validity.

310
Figure 11.2: Measurement model of Company Supplier Quality Management

The CFA results further revealed that the Company Supplier Quality Management features had
three (3) dependent variables, four (4) independent variables and six (6) free parameters. The
number of fixed non-zero parameters was four (4). These are the three (3) dependent indicator
variables for the Company Supplier Quality Management, namely Supplier performance
evaluation, Supplier knowledge of TQM, and Suppliers’ commitment to TQM. These indicator
variables are presented in Table 11.41. The Company Supplier Quality Management features
measurement model shown in Figure 11.2 was analyzed before it could be included in the full
latent variable model.

In order to establish how well the model fit the sample data and the strength of the hypothesized
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the Company Supplier Quality Management
features.

311
Table 11.41: Postulated Company Supplier Quality Management Features Model

Latent constructs Indicator variables Label

Company Supplier Quality Supplier performance evaluation CSQM4
Management (CSQM)

Supplier knowledge of TQM CSQM7

Suppliers’ commitment to TQM CSQM8

Source: Fieldwork (2018)

Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate

The unstandardized and standardized absolute residual matrix values of the Company Supplier
Quality Management features are presented in Table 11.42 and 11.43. The result reveals that all
the absolute residual values and the average off-diagonal absolute residual values were close to
zero. The unstandardized average off-diagonal residual was 0.000 while the standardized average
off-diagonal residual was found to be 0.000. A residual value greater than 2.58 is described as
large (Byrne, 2006:94). The results obtained for the Company Supplier Quality Management
features measurement model suggested a fairly acceptable fit to the sample data because the
absolute residuals were all less than 2.58. In order for a model to be described as well-fitting, the
distribution of standardized residuals should be symmetrical and centred around zero (Byrne,
2006:94).

Table 11.42: Residual Covariance Matrix for Company Supplier Quality Management
Model (Unstandardized)

Unstandardized Residual Covariance Matrix

CSQM8 CSQM4 CSQM7

CSQM8 0.000
CSQM4 0.000 0.000

312
 CSQM7 0.000 0.000 0.000

Average absolute residual = 0.000

Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%
Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

Table 11.43: Residual Covariance Matrix for Company Supplier Quality Management
Model (Standardized)

Standardized Residual Covariance Matrix

CSQM8 CSQM4 CSQM7

CSQM8 0.000
CSQM4 0.000 0.000
CSQM7 0.000 0.000 0.000

Average absolute residual = 0.000

Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%
Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit, further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

313
Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)
The analysis strategy of goodness-of-fit for the Company Supplier Quality followed a three-
statistics strategy of fit indexes as recommended by Hu and Bentler (1999). The sample data on
Company Supplier measurement model yielded the S – Bχ2 of 0.000 with 0.000 degree of
freedom (df) with a probability of p = 0.000. This chi-square value indicated that the departure of
the sample data from the postulated measurement model was significant and hence indicative of
good fit. The chi-square test is very sensitive to sample size and is used more as a descriptive
index of fit rather than as a statistical test (Kline, 2005:136). The normed chi-square value is
usually adopted by most researchers. The normed chi-square is the procedure of dividing the chi-
square by the degrees of freedom. The normed values of up to 3.0 or even 5.0 are recommended
(Kline, 2005:137).

Values for NFI range between 0 and 1 with Bentler and Bonnet (1980) recommended values
greater than 0.90 indicating a good fit. Hu and Bentler (1999) have given a cut-off criterion of
NFI ≥ 0.95. This index is sensitive to sample size, underestimating fit for samples less than 200
(Mulaik et al., 1989; Bentler, 1990), and is thus not recommended to be solely relied on (Kline,
2005). The non-normed fit index (NNFI, also known as the Tucker-Lewis index), is an index that
prefers simpler models. The value of the NNFI can indicate poor fit despite other statistics
pointing towards good fit if small samples are used (Bentler, 1990; Kline, 2005; Tabachnick &
Fidell, 2007). Owing to their non-normed nature, NNFI values can go above 1.0 and can thus be
difficult to interpret (Byrne, 1998). The chi-square and degrees of freedom were found to be
0.000. This ratio was lower than the limit of 3.00 or 5.0 advocated for by some authors (Kline,
2005:137).

The CFI value was found to be 1.000 which was greater than the cut-off limit of 0.95 so the
model is described to have an acceptable fit. The NFI value was 1.000 which is slightly above
the given cut-off value of NFI ≥ 0.95 is shown in Table 11.44. Therefore, the model is
acceptable. The NNFI value obtained is 1.000 which is also above the cut-off value of 0.80.
These fit indexes for the Company Supplier Quality Management model suggested that the

314
postulated model adequately describes the sample data and could therefore be included in the full
latent variable model analysis (Table 11.44).
Table 11.44: Robust Fit Indexes for Features Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.000

df 0≥ 0.000

Normed chi-square ≤2 or 3 good fit 0.000 Good fit

= x²/df ratio ≤5 acceptable

CFI 0.90≥ acceptable 1.000 Good fit

0.95≥ good fit
RMSEA Less than 0.05 with 0.000
confidence interval Good fit
(CI) 0.00-0.05
95% “good fit”
NFI Greater than 0.90 1.000 Good fit
“good fit”

NNFI Greater than 1.000 Good fit

0.80. “good fit”
SRMR Equal or less than 0.05 0.000 Good fit
“good
fit”

Equal or less
than 0.08 “acceptable
fit”
Source: Fieldwork (2018)

Statistical Significance of Parameter Estimates

Table 11.45 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs. The estimates were therefore deemed reasonable as well as statistically
significant. The parameter with the highest standardized coefficient was the indicator with
variable CSQM7 (Supplier knowledge of TQM) and its parameter coefficient was 0.893.

315
Table 11.45: Factor Loading and Z-statistics of Company Supplier Quality Management
Measurement

Indicator Unstandardized Standardized Z- R2 Significant

Statistics at 5%
Variable Coefficient (λ) Coefficient (λ)
level?
CSQM4 0.895 0.773 4.547 0.598 Yes
Yes
CSQM7 1.000 0.893 2.467 0.798
Yes
CSQM8 0.959 0.834 3.100 0.696

(Robust Statistical Significance at 5% level)

Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Company Supplier Quality Management features). In addition, the R2
values were also close to the desired value of 1.00, indicating that the factors explained more of
the variance in the indicator variables. The results therefore suggest that the indicator variables
significantly predict the unobserved construct because all the measured variables are
significantly associated with the Company Supplier Quality Management features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Company Supplier Quality Management
features construct was determined from the rho and the Cronbach’s alpha coefficient. According
to Kline (2005:59), the reliability coefficient should fall between zero and 1.00. Values close to
1.00 are desired. The rho coefficient of internal consistency was found to be 0.877. This value
was above the minimum required value of 0.70. Likewise, the Cronbach’s alpha was above the
minimum acceptable value of 0.70. The Cronbach’s alpha was found to be 0.871 (Table 11.46).
Both of these values revealed a high level of internal consistency and therefore reliability,
316
suggesting that the indicator variables represent the same latent construct (Company Supplier
Quality Management).

Table 11.46: Reliability and Construct Validity of Company Supplier Quality Management
Feature Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Company CSQM4 0.773
Supplier
Quality CSQM7 0.893
Management 0.871 0.877
CSQM8 0.834

*Parameter estimates are based on standardized solutions

Source: Fieldwork (2018)

Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher, and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111).

The standardized parameter coefficient presented in Table 11.46 revealed that all coefficients
were significantly higher, with the lowest being 0.773 for Company Supplier Quality
Management features. The magnitude of the parameter estimate was above the 50 per cent
minimum. This indicates a strong relationship between the indicator variables and the factors of
the Company Supplier Quality Management features construct. Therefore, the Company Supplier

317
Quality Management features satisfied both internal reliability and the construct criteria. The rho
value was above the minimum value of 0.70, and the magnitude, signs and statistical significance
of the parameter estimates were appropriate (Table 11.46).

Summary on Company Supplier Quality Management Feature Measurement Model

The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indexes met the cut-off index criteria and all the parameter estimates
were statistically significant and feasible. Considering these criteria, the measurement model for
the Company Supplier Quality Management feature was found to adequately fit the sample data.
Therefore, there was no need to improve the measurement model before it could be included in
the full latent variable model. Hence, the Company Supplier Quality Management feature
construct was adequately measured by the indicator variables and could be used in the analysis
of the full latent variable model.

11.4.4.1.3 Measurement Model for Client Focus and Involvement Features (CFI)
Construct.
A unidimensional model for Client Focus and Involvement (CFI) features is presented (Table
11.33). From the 536 cases analyzed for this construct, three (3) indicator variables (F3CFI1,
F3CFI2, and F3CFI4) made up of the same factor were realized and the name CFI was
maintained (Table 11.33). All the three (3) indicator variables obtained were used for the CFA
(Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a variable to be included in a CFA,
thus enabling the model to be described as well-fitting, the distribution of residuals covariance
matrix should be symmetrical and centred around zero (Byrne, 2006:94; Joreskog & Sorbom,
1988). The five-indicator model provides good measures of residual matrix and evidence of
convergent validity.

318
Figure 11.3: Measurement Model of Client Focus and Involvement

The CFA results further revealed that the Client Focus and Involvement features had three (3)
dependent variables, four (4) independent variables and six (6) free parameters. The number of
fixed non-zero parameters was four (4). These are the three (3) dependent indicator variables for
the Client Focus and Involvement: Client brief/input, Client complaint information/feedback, and
Client satisfaction survey. These indicator variables are presented in Table 11.47. The Client
Focus and Involvement features measurement model shown in Figure 11.33 was analyzed before
it could be included in the full latent variable model.

In order to establish how well the model fit the sample data and the strength of the hypothesized
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the Client Focus and Involvement features.

319
Table 11.47: Postulated Client Focus and Involvement Features Model

Latent constructs Indicator variables Label

Client Focus and Client brief/input CFI1
Involvement (CFI)
Client complaint CFI2
CFI2
Client complaint
information/feedback
information/feedback
Client satisfaction survey CFI4
CFI4
Client satisfaction survey
CFI6
Source: Fieldwork (2018)
Client information system
CFI8
Diagnostic Fit Analysis:Client cooperation
Analysis of Residual Covariance Estimate
The unstandardized and standardized absolute residual matrix values of the Client Focus and
Involvement features are presented in Tables 11.48 and 11.49. The result reveals that all the
absolute residual values and the average off-diagonal absolute residual values were close to zero.
The unstandardized average off-diagonal residual was 0.000 while the standardized average off-
diagonal residual was found to be 0.000. A residual value greater than 2.58 is described as large
(Byrne, 2006:94). The results obtained for the Client Focus and Involvement features
measurement model suggested a fairly acceptable fit to the sample data because the absolute
residuals were all less than 2.58. In order for a model to be described as well-fitting, the
distribution of standardized residuals should be symmetrical and centred around zero (Byrne,
2006:94).

Table 11.48: Residual Covariance Matrix for Client Focus and Involvement Model
(Unstandardized)

Unstandardized Residual Covariance Matrix

CFI4 CFI2 CFI1

CFI4 0.000
CFI2 0.000 0.000
CFI1
CFI8 0.000 0.000 0.000

Average absolute residual = 0.000

320
 Average off-diagonal absolute residual = 0.000
% falling between -0.1 +0.1 = 99.99%
Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

Table 11.49: Residual Covariance Matrix for Client Focus and Involvement Model
(Standardized)

Standardized Residual Covariance Matrix

CFI4 CFI2 CFI1

CFI4 0.000
CFI2 0.000 0.000
CFI1
CFI8 0.000 0.000 0.000

Average absolute residual = 0.000

Average off-diagonal absolute residual = 0.000
% falling between -0.1 +0.1 = 99.99%
Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit, further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)

The analysis strategy of goodness-of-fit for the Client Focus and Involvement followed a three-
statistics strategy of fit indexes as recommended by Hu and Bentler (1999). The sample data on
Client Focus and Involvement measurement model yielded the S – Bχ2 of 0.000 with 0.000
321
degree of freedom (df) with a probability of p = 0.0000. This chi-square value indicated that the
departure of the sample data from the postulated measurement model was significant and hence
indicative of good fit. The chi-square test is very sensitive to sample size and is used more as a
descriptive index of fit rather than as a statistical test (Kline, 2005:136). The normed chi-square
value is usually adopted by most researchers. The normed chi-square is the procedure of dividing
the chi-square by the degrees of freedom. The normed values of up to 3.0 or even 5.0 are
recommended (Kline, 2005:137).

Values for NFI range between 0 and 1 with Bentler and Bonnet (1980) recommended values
greater than 0.90 indicating a good fit. Hu and Bentler (1999) have given a cut-off criterion of
NFI ≥ 0.95. This index is sensitive to sample size, underestimating fit for samples less than 200
(Mulaik et al., 1989; Bentler, 1990) and is thus not recommended to be solely relied on (Kline,
2005). The non-normed fit index (NNFI, also known as the Tucker-Lewis index) is an index that
prefers simpler models. The value of the NNFI can indicate poor fit despite other statistics
pointing towards good fit if small samples are used (Bentler, 1990; Kline, 2005; Tabachnick &
Fidell, 2007). Owing to their non-normed nature, NNFI values can go above 1.0 and can thus be
difficult to interpret (Byrne, 1998). The chi-square and degrees of freedom were found to be
0.000. This ratio was lower than the limit of 3.00 or 5.0 advocated for by some authors (Kline,
2005:137).

The CFI value was found to be 1.000 which was greater than the cut-off limit of 0.95 so the
model is described to have an acceptable fit. The NFI value was 1.000 which is within the given
range, but the given cut-off value of NFI ≥ 0.95 is shown in Table 11.50. Therefore, the model is
acceptable. The NNFI value obtained is 1.000 which is also above the cut-off value of 0.80.
These fit indexes for the Client Focus and Involvement model suggested that the postulated
model adequately describes the sample data and could therefore be included in the full latent
variable model analysis (Table 11.50).

322
Table 11.50: Robust Fit Indexes for Client Focus and Involvement Features Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.000

df 0≥ 0.000

Normed chi-square ≤2 or 3 good fit

= x²/df ratio ≤5 acceptable
CFI 0.90≥ acceptable 1.000
0.95≥ good fit Good fit

RMSEA Less than 0.05 with 0.000

confidence interval Good fit
(CI) 0.00-0.05
95% “good fit”
NFI Greater than 0.90 1.000
“good fit” Good fit

NNFI Greater than 1.000

0.80. “good fit” Good fit

SRMR Equal or less than 0.05 0.000 Good fit

“good
fit”

Equal or less
than 0.08 “acceptable
fit”
Source: Fieldwork (2018)

Statistical Significance of Parameter Estimates

Table 11.51 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs except CFI4 the value of which (1.895) was slightly below 1.96. However, the
estimates were deemed reasonable as well as statistically significant. The parameter with the
highest standardized coefficient was the indicator with variable CFI1 (Client brief/input) and its
parameter coefficient was 0.825.

323
Table 11.51: Factor Loading and Z-statistics of Client Focus and Involvement
Measurement

Indicator Unstandardized Standardized Z- R2 Significant

Statistics at 5%
Variable Coefficient (λ) Coefficient (λ)
level?
CFI1 1.000 0.825 4.125 0.714 Yes

CFI2 0.979 0.823 3.532 0.655 Yes

CFI4 0.928 0.718 2.211 0.524 Yes

(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Client Focus and Involvement). In addition, the R2 values were also close
to the desired value of 1.00, indicating that the factors explained more of the variance in the
indicator variables. The results therefore suggest that the indicator variables significantly predict
the unobserved construct because all the measured variables are significantly associated with the
Client Focus and Involvement features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Client Focus and Involvement features
construct were determined from the rho and the Cronbach’s alpha coefficient. According to
Kline (2005:59), the reliability coefficient should fall between zero and 1.00. Values close to
1.00 are desired. The rho coefficient of internal consistency was found to be 0.724. This value
was above the minimum required value of 0.70. Likewise, the Cronbach’s alpha was above the
minimum acceptable value of 0.70. The Cronbach’s alpha was found to be 0.834 (Table 11.53).
Both of these values revealed a high level of internal consistency and therefore reliability,
suggesting that the indicator variables represent the same latent construct (Client Focus and
Involvement).

324
Table 11.52: Reliability and Construct Validity of Client Focus and Involvement Feature
Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Client Focus CFI1 0.825
and CFI2 0.823
Involvement CFI4 0.718
0.834 0.724
*Parameter estimates are based on standardized solutions
Source: Fieldwork (2018)

Furthermore, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher, and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111).

The standardized parameter coefficient presented in Table 11.51 revealed that all coefficients
were significantly higher with the lowest being 0.718 for Client Focus and Involvement features.
The magnitude of the parameter estimate was below the 50 per cent minimum. This indicates a
weak relationship between the indicator variables and the factors of the Client Focus and
Involvement features construct. Therefore the Client Focus and Involvement features satisfied
both internal reliability and the construct criteria. The rho value was above the minimum value of
0.70, and the magnitude, signs and statistical significance of the parameter estimates were
appropriate (Table 11.52).

325
Summary on Client Focus and Involvement Feature Measurement Model
The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indexes met the cut-off index criteria and all the parameter estimates
were statistically significant and feasible. Considering these criteria, the measurement model for
the Client Focus and Involvement feature was found to adequately fit the sample data. Therefore,
there was no need to improve the measurement model before it could be included in the full
latent variable model. Hence, the Client Focus and Involvement feature construct was adequately
measured by the indicator variables and could be used in the analysis of the full latent variable
model.

11.4.4.1.4 Measurement Model for Company Quality System Evaluation features (CQSE)
Construct
A unidimensional model for Company Quality System Evaluation (CQSE) features is presented
(Table 11.33). From the 536 cases analysed for this construct, three (3) indicator variables
(F4CQSE1, F4CQSE2, and F4CQSE3) made up of the same factor were realized and the name
CQSE was maintained (Table 11.33). All the three (3) indicator variables obtained were used for
the CFA (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a variable to be included in a
CFA, thus enabling the model to be described as well-fitting, the distribution of residuals
covariance matrix should be symmetrical and centred around zero (Byrne, 2006:94; Joreskog &
Sorbom, 1988). The three-indicator model provides good measures of residual matrix and
evidence of convergent validity.

Figure 11.4: Measurement Model of Company Quality System Evaluation

326
The CFA results further revealed that the Company Quality System Evaluation features had three
(3) dependent variables, four (4) independent variables and six (6) free parameters. The number
of fixed non-zero parameters was four (4). These are the three (3) dependent indicator variables
for the Company Quality System Evaluation: Evaluation of the quality strategy framework,
Evaluation of overall company performance, and Evaluation of employee performance. These
indicator variables are presented in Table 11.53. The Company Quality System Evaluation
features measurement model shown in Figure 11.4 was analyzed before it could be included in
the full latent variable model.

In order to establish how well the model fit the sample data and the strength of the hypothesised
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the Company Quality System Evaluation
features.

Table 11.53: Postulated Company Quality System Evaluation Features Model

Latent constructs Indicator variables Label

Company Quality Evaluation of the quality strategy CQSE1
System Evaluation framework
(CQSE)
Evaluation of overall company CQSE2
performance
Evaluation of employee performance CQSE3

Source: Fieldwork (2018)

Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate

The unstandardized and standardized absolute residual matrix values of the Company Quality
System Evaluation features are presented in Tables 11.54 and 11.55. The results reveal that all
327
the absolute residual values and the average off-diagonal absolute residual values were close to
zero. The unstandardized average off-diagonal residual was 0.000 while the standardized average
off-diagonal residual was found to be 0.000. A residual value greater than 2.58 is described as
large (Byrne, 2006:94). The results obtained for the Company Quality System Evaluation
features measurement model suggested a fairly acceptable fit to the sample data because the
absolute residuals were all less than 2.58. In order for a model to be described as well-fitting, the
distribution of standardized residuals should be symmetrical and centred around zero (Byrne,
2006:94).

Table 11.54: Residual Covariance Matrix for Company Quality System Evaluation Model
(Unstandardized)

Unstandardized Residual Covariance Matrix

CQSE3 CQSE1 CQSE2

CQSE3 0.000
CQSE1 0.000 0.000
CQSE2 0.000 0.000 0.000
Average absolute residual = 0.000
Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall
between -0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01%
fell outside the -0.1 to 0.1 ranges.

328
Table 11.55: Residual Covariance Matrix for Company Quality System Evaluation Model
(Standardized)

Standardized Residual Covariance Matrix

CQSE3 CQSE1 CQSE2

CQSE3 0.000
CQSE1 0.000 0.000
CQSE2 0.000 0.000 0.000

Average absolute residual = 0.000

Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit, further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)

The analysis strategy of goodness-of-fit for the Company Quality System Evaluation followed a
three-statistics strategy of fit indexes as recommended by Hu and Bentler (1999). The sample
data on Company Quality System Evaluation measurement model yielded the S – Bχ2 of 0.000
with a 0.000 degree of freedom (df) with a probability of p = 0.0000. This chi-square value
indicated that the departure of the sample data from the postulated measurement model was
significant and hence indicative of good fit. The chi-square test is very sensitive to sample size
and is used more as a descriptive index of fit rather than as a statistical test (Kline, 2005:136).
The normed chi-square value is usually adopted by most researchers. The normed chi-square is
the procedure of dividing the chi-square by the degrees of freedom. The normed values of up to
3.0 or even 5.0 are recommended (Kline, 2005:137).

329
Values for NFI range between 0 and 1 with Bentler and Bonnet (1980) recommended values
greater than 0.90 indicating a good fit. Hu and Bentler (1999) have given a cut-off criterion of
NFI ≥ 0.95. This index is sensitive to sample size, underestimating fit for samples less than 200
(Mulaik et al., 1989; Bentler, 1990), and is thus not recommended to be solely relied on (Kline,
2005). The non-normed fit index (NNFI, also known as the Tucker-Lewis index) is an index that
prefers simpler models. The value of the NNFI can indicate poor fit despite other statistics
pointing towards good fit if small samples are used (Bentler, 1990; Kline, 2005; Tabachnick &
Fidell, 2007). Owing to non-normed nature, NNFI values can go above 1.0 and can thus be
difficult to interpret (Byrne, 1998). The chi-square and degrees of freedom were found to be
0.000. This ratio was lower than the limit of 3.00 or 5.0 advocated for by some authors (Kline,
2005:137). The CFI value was found to be 1.000 which was slightly greater than the cut-off limit
of 0.95 so the model is described to have an acceptable fit. The NFI value was 1.000 which is
slightly above the given cut-off value of NFI ≥ 0.95 which is shown in Table 11.56. Therefore,
the model is acceptable. The NNFI value obtained is 1.000 which is also above the cut-off value
of 0.80. These fit indexes for the Company Quality System Evaluation model suggested that the
postulated model adequately describes the sample data and could therefore be included in the full
latent variable model analysis (Table 11.56).

Table 11.56: Robust fit Indexes for Company Quality System Evaluation Features
Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.000

df 0≥ 0.000

Normed chi-square ≤2 or 3 good fit

= x²/df ratio ≤5 acceptable
CFI 0.90≥ acceptable 1.000 Good fit
0.95≥ good fit
RMSEA Less than 0.05 0.000
with confidence Good fit
interval
330
 (CI) 0.00-0.05 “good
95% fit”
NFI Greater than 0.90 1.000 Good fit
“good fit”
NNFI Greater than 0.80. 1.000 Good fit
“good fit”
SRMR Equal or less than 0.05 0.000 Good fit
“good
fit”

Equal or less
than 0.08 “acceptable
fit”

Statistical significance of parameter estimates

Source: Fieldwork (2018)

Table 11.57 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs. The estimates were therefore deemed reasonable as well as statistically
significant. The parameter with the highest standardized coefficient was the indicator with a
variable CQSE2 (Evaluation of overall company performance) and its parameter coefficient was
0.832.

Table 11.57: Factor Loading and Z-statistics of Company Quality System Evaluation
Measurement

Indicator Unstandardize Standardized Z- R2 Significan

d Statistic t at 5%
Variable Coefficient (λ)
Coefficient (λ) s level?
CQSE1 1.000 0.813 4.542 0.660 Yes
Yes
CQSE2 0.984 0.832 4.139 0.692
Yes
CQSE3 0.987 0.741 2.301 0.549 Yes
(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)

331
Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Company Quality System Evaluation). In addition, the R2 values were also
close to the desired value of 1.00, indicating that the factors explained more of the variance in
the indicator variables. The results therefore suggest that the indicator variables significantly
predict the unobserved construct because all the measured variables are significantly associated
with the Company Quality System Evaluation features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Company Quality System Evaluation
features construct were determined from the rho and the Cronbach’s alpha coefficient. According
to Kline (2005:59), the reliability coefficient should fall between zero and 1.00. Values close to
1.00 are desired. The rho coefficient of internal consistency was found to be 0.830. This value
was above the minimum required value of 0.70. On the other hand, the Cronbach’s alpha was
above the minimum acceptable value of 0.70. The Cronbach’s alpha was found to be 0.835
(Table 11.58). The Cronbach’s alpha value revealed a high level of internal consistency and
therefore reliability, suggesting that the indicator variables represent the same latent construct
(Company Quality System Evaluation).
Table 11.58: Reliability and Construct Validity of Company Quality System Evaluation
Feature Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Company CQSE1 0.813
Quality
System CQSE2 0.832
Evaluation
Features CQSE3 0.741
0.835 0.830

*Parameter estimates are based on standardized solutions

Source: Fieldwork (2018)

332
Furthermore, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111).

The standardized parameter coefficient presented in Table 11.62 revealed that all coefficients
were significantly higher with the lowest being 0.741 for Company Quality System Evaluation
features. The magnitude of the parameter estimate was above the 50 per cent minimum. This
indicates a strong relationship between the indicator variables and the factors of the Company
Quality System Evaluation features construct. Therefore the Company Quality System
Evaluation features satisfied both internal reliability and the construct criteria. The rho value was
slightly below the minimum value of 0.70, and the magnitude, signs and statistical significance
of the parameter estimates were appropriate (Table 11.58).

Summary on Company Quality System Evaluation Feature Measurement Model

The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indexes met the cut-off index criteria and all the parameter estimates
were statistically significant and feasible. Considering these criteria, the measurement model for
the Company Quality System Evaluation feature was found to adequately fit the sample data.
Therefore, there was no need to improve the measurement model before it could be included in
the full latent variable model. Hence, the Company Quality System Evaluation feature construct
was adequately measured by the indicator variables and could be used in the analysis of the full
latent variable model.

333
11.4.4.1.5 Measurement Model for Company Vision and Plan Statement features (CVPS)
Construct
A unidimensional model for Company Vision and Plan Statement (CVPS) features is presented
(Table 11.33). From the 536 cases analyzed for this construct, three (3) indicator variables
(F5CVPS5, F5CVPS6 and F5CVPS7) made up of the same factor were realized and the name
CVPS was maintained (Table 11.33). All the three (3) indicator variables obtained were used for
the CFA (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a variable to be included in a
CFA, thus enabling the model to be described as well-fitting, the distribution of residuals
covariance matrix should be symmetrical and centred around zero (Byrne, 2006:94; Joreskog &
Sorbom, 1988). The three-indicator model provides good measures of residual matrix and
evidence of convergent validity.

Figure 11.5: Measurement Model of Company Vision and Plan Statement

The CFA results further revealed that the Company Vision and Plan Statement features had three
(3) dependent variables, four (4) independent variables and six (6) free parameters. The number
of fixed non-zero parameters was four (4). These are the three (3) dependent indicator variables
for the Company Vision and Plan Statement: Concreteness of the future plan, Employee
contribution to the vision and Involvement of employees in the development of the vision
statement. These indicator variables are presented in Table 11.59. The Company Vision and Plan
Statement features measurement model shown in Figure 11.5 was analyzed before it could be
included in the full latent variable model.
334
In order to establish how well the model fit the sample data and the strength of the hypothesized
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the Company Vision and Plan Statement
features.

Table 11.59: Postulated Company Vision and Plan Statement Features Model

Latent constructs Indicator variables Label

Company Vision and Concreteness of the future plan CVPS5
Plan Statement (CVPS)
Employee contribution to the vision CVPS6
Involvement of employees in the CVPS7
development of the vision statement

Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate

The unstandardized and standardized absolute residual matrix values of the Company Vision and
Plan Statement features are presented in Tables 11.60 and 11.61. The results reveal that all the
absolute residual values and the average off-diagonal absolute residual values were close to zero.
The unstandardized average off-diagonal residual was 0.000 while the standardized average off-
diagonal residual was found to be 0.000. A residual value greater than 2.58 is described as large
(Byrne, 2006:94). The results obtained for the Company Vision and Plan Statement features
measurement model suggested a fairly acceptable fit to the sample data because the absolute
residual were all less than 2.58. In order for a model to be described as well-fitting, the
distribution of standardized residuals should be symmetrical and centred around zero (Byrne,
2006:94).

335
Table 11.60: Residual Covariance Matrix for Company Vision and Plan Statement Model
(Unstandardized)

Unstandardized Residual Covariance Matrix

CVPS7 CVPS6 CVPS5

CVPS7 0.000
CVPS6 0.000 0.000
CVPS5 0.000 0.000 0.000
Average absolute residual = 0.000
Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01%
fell outside the -0.1 to 0.1 ranges.

Table 11.61: Residual Covariance Matrix for Company Vision and Plan Statement Model
(Standardized)

Standardized Residual Covariance Matrix

CVPS7 CVPS6 CVPS5

CVPS7 0.000
CVPS6 0.000 0.000
CVPS5 0.000 0.000 0.000
Average absolute residual = 0.000
Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

336
From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit; further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)

The analysis strategy of goodness-of-fit for the Company Vision and Plan Statement followed a
three-statistics strategy of fit indexes as recommended by Hu and Bentler (1999). The sample
data on the Company Vision and Plan Statement measurement model yielded the S – Bχ2 of
0.000 with 0.000 degree of freedom (df) with a probability of p = 0.0000. This chi-square value
indicated that the departure of the sample data from the postulated measurement model was
significant and hence indicative of good fit. The chi-square test is very sensitive to sample size
and is used more as a descriptive index of fit rather than as a statistical test (Kline, 2005:136).
The normed chi-square value is usually adopted by most researchers. The normed chi-square is
the procedure of dividing the chi-square by the degrees of freedom. The normed values of up to
3.0 or even 5.0 are recommended (Kline, 2005:137).

Values for NFI range between 0 and 1 with Bentler and Bonnet (1980) recommended values
greater than 0.90 indicating a good fit. Hu and Bentler (1999) have given a cut-off criterion of
NFI ≥ 0.95. This index is sensitive to sample size, underestimating fit for samples less than 200
(Mulaik et al., 1989; Bentler, 1990), and is thus not recommended to be solely relied on (Kline,
2005). The non-normed fit index (NNFI, also known as the Tucker-Lewis index), is an index that
prefers simpler models. The value of the NNFI can indicate poor fit despite other statistics
pointing towards good fit if small samples are used (Bentler, 1990; Kline, 2005; Tabachnick &
Fidell, 2007). Owing to their non-normed nature, NNFI values can go above 1.0 and can thus be
difficult to interpret (Byrne, 1998). The chi-square and degrees of freedom were found to be
0.000. This ratio was lower than the limit of 3.00 or 5.0 advocated for by some authors (Kline,
2005:137).

337
The CFI value was found to be 1.000 which was slightly greater than the cut-off limit of 0.95 so
this is described to have an acceptable fit. The NFI value was 1.000 which is within the given
range, but the given cut-off value of NFI ≥ 0.95 is shown in Table 11.62. Therefore, the model is
acceptable. The NNFI value obtained is 1.000 which is above the cut-off value of 0.80. These fit
indexes for the Company Vision and Plan Statement model suggested that the postulated model
adequately describes the sample data and could therefore be included in the full latent variable
model analysis (Table 11.62).

Table 11.62: Robust Fit Indexes for Company Vision and Plan Statement Features
Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.000

df 0≥ 0.000

Normed chi-square ≤2 or 3 good fit -

= x²/df ratio ≤5 acceptable
CFI 0.90≥ acceptable 1.000 Good fit
0.95≥ good fit
RMSEA Less than 0.05 with 0.000
confidence interval Good fit
(CI) 0.00-0.05
95% “good fit”
NFI Greater than 0.90 1.000 Good fit
“good fit”
NNFI Greater than 1.000 Good fit
0.80. “good fit”
SRMR Equal or less than 0.000 Good fit
0.05 “good
fit”

Equal or less
than 0.08
“acceptable fit”

Source: Fieldwork (2018)

338
Statistical Significance of Parameter Estimates
Table 11.63 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs. The estimates were therefore deemed reasonable as well as statistically
significant. The parameter with the highest standardized coefficient was the indicator with
variable CVPS6 (Employee contribution to the vision) and its parameter coefficient was 0.850.

Table 11.63: Factor Loading and Z-statistics of Company Vision and Plan Statement
Measurement

Indicator Unstandardize Standardized Z- R2 Significan

d Statistic t at 5%
Variable Coefficient (λ)
Coefficient (λ) s level?
CVPS5 1.000 0.828 3.101 0.685 Yes

CVPS6 1.056 0.850 3.414 0.723 Yes

CVPS7 0.847 0.775 0.601 Yes

2.788
(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Company Vision and Plan Statement). In addition, the R2 values were also
close to the desired value of 1.00, indicating that the factors explained more of the variance in
the indicator variables. The results therefore suggest that the indicator variables significantly
predict the unobserved construct because all the measured variables are significantly associated
with the Company Vision and Plan Statement features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Company Vision and Plan Statement
features construct were determined from the rho and the Cronbach’s alpha coefficient. According

339
to Kline (2005:59), the reliability coefficient should fall between zero and 1.00. Values close to
1.00 are desired. The rho coefficient of internal consistency was found to be 0.816. This value
was above the minimum required value of 0.70. Likewise, the Cronbach’s alpha was above the
minimum acceptable value of 0.70. The Cronbach’s alpha was found to be 0.857 (Table 11.64).
Both of these values revealed a high level of internal consistency and therefore reliability,
suggesting that the indicator variables represent the same latent construct (Company Vision and
Plan Statement).

Table 11.64: Reliability and Construct Validity of Company Vision and Plan Statement
Feature Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Company CVPS5 0.828
Vision and
Plan CVPS6 0.850
Statement
(CVPS CVPS7 0.775 0.857 0.816
*Parameter estimates are based on standardized solutions
Source: Fieldwork (2018)

Furthermore, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111).

340
The standardized parameter coefficient presented in Table 11.64 revealed that all coefficients
were significantly higher with the lowest being 0.775 for Company Vision and Plan Statement
features. The magnitude of the parameter estimate was above the 50 per cent minimum. This
indicates a strong relationship between the indicator variables and the factors of the Company
Vision and Plan Statement features construct. Therefore the Company Vision and Plan Statement
features satisfied both internal reliability and the construct criteria. The rho value was above the
minimum value of 0.70, and the magnitude, signs and statistical significance of the parameter
estimates were appropriate (Table 11.64).

Summary of Company Vision and Plan Statement Feature Measurement Model

The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indexes met the cut-off index criteria and all the parameter estimates
were statistically significant and feasible. Considering these criteria, the measurement model for
the Company Vision and Plan Statement feature was found to adequately fit the sample data.
Therefore, there was no need to improve the measurement model before it could be included in
the full latent variable model. Hence the Company Vision and Plan Statement feature construct
was adequately measured by the indicator variables and could be used in the analysis of the full
latent variable model.

11.4.4.1.6 Measurement Model for Product Design Management features (PDM) Construct
An unidimensional model for Product Design Management (PDM) features was presented (Table
11.33). From the 536 cases analyzed for this construct, three (3) indicator variables (F6APDM2,
F6APDM3, and F6APDM6) made up of the same factor were realized and the name PDM was
maintained (Table 11.33). All the three (3) indicator variables obtained were used for the CFA
(Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a variable to be included in a CFA,
thus enabling the model to be described as well-fitting, the distribution of residuals covariance
matrix should be symmetrical and centred around zero (Byrne, 2006:94; Joreskog & Sorbom,
1988). The three-indicator model provides good measures of residual matrix and evidence of
convergent validity.

341
Figure 11.6: Measurement Model of Product Design Management

The CFA results further revealed that the Product Design Management features had three (3)
dependent variables, four (4) independent variables and six (6) free parameters. There were four
(4) fixed non-zero parameters. The three (3) dependent indicator variables for the Product
Design Management are Strength required for the end product, Intended purpose of the material,
and Design detailing. These indicator variables are presented in Table 11.65. The Product Design
Management features measurement model shown in Figure 11.6 was analyzed before it could be
included in the full latent variable model.

In order to establish how well the model fit the sample data and the strength of the hypothesized
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the Product Design Management features.

342
Table 11.65: Postulated Product Design Management Features Model

Latent Indicator variables Label

Product Design
constructs Strength required for the end product PDM2
Management Intended purpose of the material PDM3
(PDM) PDM6
Design detailing

Source: Fieldwork (2018)

Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate

The unstandardized and standardized absolute residual matrix values of the Product Design
Management features are presented in Tables 11.66 and 11.67. The results reveal that all the
absolute residual values and the average off-diagonal absolute residual values were close to zero.
The unstandardized average off-diagonal residual was 0.000 while the standardized average off-
diagonal residual was found to be 0.000. A residual value greater than 2.58 is described as large
(Byrne, 2006:94). The results obtained for the Product Design Management Features
measurement model suggested a fairly acceptable fit to the sample data because the absolute
residual were all less than 2.58. In order for a model to be described as well-fitting, the
distribution of standardized residuals should be symmetrical and centred around zero (Byrne,
2006:94).

Table 11.66: Residual Covariance Matrix for Product Design Management Model
(Unstandardized)

Unstandardized Residual Covariance Matrix

PDM16 PDM3 PDM2

PDM6 0.000
PDM3 0.000 0.000
PDM2 0.000 0.000 0.000

Average absolute residual = 0.000

343
 Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

Table 11.67: Residual Covariance Matrix for Product Design Management Model
(Standardized)

Standardized Residual Covariance Matrix

PDM6 PDM3 PDM2

PDM6 0.000
PDM3 0.000 0.000
PDM2 0.000 0.000 0.000
Average absolute residual = 0.000
Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit; further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

Goodness-of-fit Statistics – Robust Maximum likelihood (RML)

The analysis strategy of goodness-of-fit for Product Design Management followed a three-
statistics strategy of fit indexes as recommended by Hu and Bentler (1999). The sample data on
Product Design Management measurement model yielded the S – Bχ2 of 0.000 with 0.000
344
degrees of freedom (df) with a probability of p = 0.000. This chi-square value indicated that the
departure of the sample data from the postulated measurement model was significant and hence
indicative of good fit. The chi-square test is very sensitive to sample size and is used more as a
descriptive index of fit rather than as a statistical test (Kline, 2005:136). The normed chi-square
value is usually adopted by most researchers. The normed chi-square is the procedure of dividing
the chi-square by the degrees of freedom. The normed values of up to 3.0 or even 5.0 are
recommended (Kline, 2005:137). Values for NFI range between 0 and 1 with Bentler and Bonnet
(1980) recommended values greater than 0.90 indicating a good fit. Hu and Bentler (1999) have
given a cut-off criterion of NFI ≥ 0.95. This index is sensitive to sample size, underestimating fit
for samples less than 200 (Mulaik et al., 1989; Bentler, 1990) and is thus not recommended to be
solely relied on (Kline, 2005). The non-normed fit index (NNFI, also known as the Tucker-
Lewis index) is an index that prefers simpler models. The value of the NNFI can indicate poor fit
despite other statistics pointing towards good fit if small samples are used (Bentler, 1990; Kline,
2005; Tabachnick & Fidell, 2007). Owing to their non-normed nature, NNFI values can go
above 1.0 and can thus be difficult to interpret (Byrne, 1998). The chi-square and degrees of
freedom were found to be 0.000. This ratio was lower than the limit of 3.00 or 5.0 advocated for
by some authors (Kline, 2005:137). The CFI value was found to be 1.000 which was slightly
greater than the cut-off limit of 0.95 so the model is described to have an acceptable fit. The NFI
value was 1.000 which is slightly greater than the given cut-off value of NFI ≥ 0.95 is shown in
Table 11.68. Therefore, the model is acceptable. The NNFI value obtained is 1.000 which is also
above the cut-off value of 0.80. These fit indexes for the Product Design Management model
suggested that the postulated model adequately describe the sample data and could therefore be
included in the full latent variable model analysis (Table 11.68).

345
Table 11.68: Robust fit indexes for Product Design Management Features Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.000

df 0≥ 0.000

Normed chi-square ≤2 or 3 good fit -

= x²/df ratio ≤5 acceptable

CFI 0.90≥ acceptable 1.000 Good fit

0.95≥ good fit

RMSEA Less than 0.05 with 0.000
confidence interval
(CI) 0.00-0.05 Good fit
“good fit” fit
0.95≥ good
NFI Greater than 0.90 1.000 Good fit
“good fit”
95%
NNFI Greater than 1.000 Good fit
0.80. “good fit”

SRMR Equal or less than 0.05 0.000 Good fit

“good
fit”

Equal or less
than 0.08 “acceptable
fit”
Source: Fieldwork (2018)

Statistical Significance of Parameter Estimates

Table 11.69 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs. The estimates were therefore deemed reasonable as well as statistically

346
significant. The parameter with the highest standardized coefficient was the indicator with
variable PDM2 (Strength required for the end product) and its parameter coefficient was 0.786.

Table 11.69: Factor Loading and Z-statistics of Product Design Management Condition
Measurement

Indicator Unstandardized Standardized Z- R2 Significant

Statistics at 5%
Variable Coefficient (λ) Coefficient
(λ) level?

PDM2 1.000 0.786 3.169 0.600 Yes

PDM3 1.009 0.757 2.980 0.627 Yes

PDM6 0.886 0.718 0.515

2.370 Yes
(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Product Design Management features). In addition, the R2 values were also
close to the desired value of 1.00, indicating that the factors explained more of the variance in
the indicator variables. The results therefore suggest that the indicator variables significantly
predict the unobserved construct because all the measured variables are significantly associated
with the Product Design Management features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Product Design Management features
construct were determined from the rho and the Cronbach’s alpha coefficient. According to
Kline (2005:59), the reliability coefficient should fall between zero and 1.00. Values close to
1.00 are desired. The rho coefficient of internal consistency was found to be 0.762. This value
was above the minimum required value of 0.70. Likewise, the Cronbach’s alpha was above the
minimum acceptable value of 0.70. The Cronbach’s alpha was found to be 0.798 (Table 11.70).

347
Both of these values revealed a high level of internal consistency and therefore reliability,
suggesting that the indicator variables represent the same latent construct (Product Design
Management).

Table 11.70: Reliability and Construct Validity of Product Design Management Feature
Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Product PDM2 0.786
Design
Management PDM3 0.757 0.798 0.762
Features
PDM6 0.718

*Parameter estimates are based on standardized solutions

Source: Fieldwork (2018)

Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111). The standardized
parameter coefficient presented in Table 11.70 revealed that all coefficients were significantly
higher with the lowest being 0.718 for Product Design Management features. The magnitude of
the parameter estimate was above the 50 per cent minimum. This indicates a strong relationship
between the indicator variables and the factors of the Product Design Management features
construct. Therefore, the Product Design Management features satisfied both internal reliability

348
and the construct criteria. The rho value was above the minimum value of 0.70 and the
magnitude, signs and statistical significance of the parameter estimates were appropriate (Table
11.70).

Summary on Product Design Management Feature Measurement Model

The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indices met the cut-off index criteria and all the parameter estimates were
statistically significant and feasible. Considering these criteria, the measurement model for the
Product Design Management feature was found to adequately fit the sample data. Therefore,
there was no need to improve the measurement model before it could be included in the full
latent variable model. Hence, the Product Selection and Design Management feature construct
was adequately measured by the indicator variables and could be used in the analysis of the full
latent variable model.

11.4.4.1.7 Measurement Model for Product Selection Management Features (PSM)

Construct
A unidimensional model for Product Selection Management (PSM) features was presented
(Table 11.33). From the 536 cases analyzed for this construct, three (3) indicator variables
(F6BPSM1, F6BPSM2, and F6BPSM4) made up of the same factor were realized and the name
PSM was maintained (Table 11.33). All the three (3) indicator variables obtained were used for
the CFA (Joreskog & Sorbom, 1988; Byrne, 2006:94). In order for a variable to be included in a
CFA, thus enabling the model to be described as well-fitting, the distribution of residuals
covariance matrix should be symmetrical and centred around zero (Joreskog & Sorbom, 1988;
Byrne, 2006:94). The three-indicator model provides good measures of residual matrix and
evidence of convergent validity.

349
Figure 11.7: Measurement Model of Product Selection Management

The CFA results further revealed that the Product Selection Management features had three
(3) dependent variables, four (4) independent variables and six (6) free parameters. There
were four (4) fixed non-zero parameters. There are three (3) dependent indicator variables
for the Product Selection Management, namely Client brief/input, Cost of selection and
design of the product and Appearance/finishes required. These indicator variables are
presented in Table 11.71. The Product Selection Management features measurement model
shown in Figure 11.7 was analyzed before it could be included in the full latent variable
model.

In order to establish how well the model fit the sample data and the strength of the
hypothesized relationship between the variables, results on residual covariance matrix
(unstandardized and standardized), distribution of standardised residuals, fit statistics and
statistical significance at a probability level of 5 per cent were examined. In addition, the
Cronbach’s alpha and the rho coefficient of internal consistency were examined to
determine the score reliability. Results of these statistics are presented in the next section
for the Product Selection Management features.

350
Table 11.71: Postulated Product Selection Management Features Model

Latent constructs Indicator variables Label

Product Selection Client brief/input PSM1
Management (PSM) Cost of selection and design PSM2
of the product
Appearance/finishes required PSM4

Source: Fieldwork (2018)

Diagnostic Fit analysis: Analysis of Residual Covariance EDstimate

The unstandardized and standardized absolute residual matrix values of the Product Selection
and Design Management features are presented in Tables 11.72 and 11.73. The results reveal that
all the absolute residual values and the average off-diagonal absolute residual values were close
to zero. The unstandardized average off-diagonal residual was 0.000 while the standardized
average off-diagonal residual was found to be 0.000. A residual value greater than 2.58 is
described as large (Byrne, 2006:94). The results obtained for the Product Selection Management
Features measurement model suggested a fairly acceptable fit to the sample data because the
absolute residuals were all less than 2.58. In order for a model to be described as well-fitting, the
distribution of standardized residuals should be symmetrical and centred around zero (Byrne,
2006:94).

Table 11.72: Residual Covariance Matrix for Product Selection Management Model
(Unstandardized)

Unstandardized Residual Covariance Matrix

PSM4 PSM2 PSM1

PSM4 0.000
PSM2 0.000 0.000
PSM1 0.000 0.000 0.000

Average absolute residual = 0.000

351
 Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

Table 11.73: Residual Covariance Matrix for Product Selection Management Model
(Standardized)

Standardized Residual Covariance Matrix

PSM4 PSM2 PSM1

PSM4 0.000
PSM2 0.000 0.000
PSM1 0.000 0.000 0.000

Average absolute residual = 0.000

Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit; further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)

The analysis strategy of goodness-of-fit for Product Selection Management followed a three-
statistics strategy of fit indexes as recommended by Hu and Bentler (1999). The sample data on
Product Selection Management measurement model yielded the S – Bχ2 of 0.000 with 0.000

352
degrees of freedom (df) with a probability of p = 0.0000. This chi-square value indicated that the
departure of the sample data from the postulated measurement model was significant and hence
indicative of good fit. The chi-square test is very sensitive to sample size and is used more as a
descriptive index of fit rather than as a statistical test (Kline, 2005:136). The normed chi-square
value is usually adopted by most researchers. The normed chi-square is the procedure of dividing
the chi-square by the degrees of freedom. The normed values of up to 3.0 or even 5.0 are
recommended (Kline, 2005:137). Values for NFI range between 0 and 1 with Bentler and Bonnet
(1980) recommended values greater than 0.90 indicating a good fit. Hu and Bentler (1999) have
given a cut-off criterion of NFI ≥ 0.95. This index is sensitive to sample size, underestimating fit
for samples less than 200 (Mulaik et al., 1989; Bentler, 1990), and is thus not recommended to
be solely relied on (Kline, 2005). The non-normed fit index (NNFI, also known as the Tucker-
Lewis index) is an index that prefers simpler models. The value of the NNFI can indicate poor fit
despite other statistics pointing towards good fit if small samples are used (Bentler, 1990; Kline,
2005; Tabachnick & Fidell, 2007). Owing to their non-normed nature, NNFI values can go
above 1.0 and can thus be difficult to interpret (Byrne, 1998). The chi-square and degrees of
freedom were found to be 0.000. This ratio was lower than the limit of 3.00 or 5.0 advocated for
by some authors (Kline, 2005:137). The CFI value was found to be 1.000 which was slightly
greater than the cut-off limit of 0.95 so the model is described to have an acceptable fit. The NFI
value was 1.000 which is slightly greater than the given cut-off value of NFI ≥ 0.95 which is
shown in Table 11.74. Therefore, the model is acceptable. The NNFI value obtained is 1.000
which is above the cut-off value of 0.80. These fit indexes for the Product Selection Management
model suggested that the postulated model adequately describes the sample data and could
therefore be included in the full latent variable model analysis (Table 11.74).

353
Table 11.74: Robust Fit Indexes for Product Selection and Design Management Features
Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.000
df 0≥ 0.00
0
Normed chi-square ≤2 or 3 good fit -
= x²/df ratio ≤5 acceptable
CFI 0.90≥ acceptable 1.000 Good fit

0.95≥ good fit

RMSEA Less than 0.05 with 0.000
confidence interval
(CI) 0.00-0.05 Good fit
“good fit” fit
0.95≥ good
NFI Greater than 0.90 1.000 Good fit
“good fit”
95%
NNFI Greater than 1.000 Good fit
0.80. “good fit”
SRMR Equal or less than 0.05 0.000 Good fit
“good
fit”

Equal or less
than 0.08 “acceptable
fit”
Source: Fieldwork (2018)

Statistical Significance of Parameter Estimates

Table 11.75 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and
show appropriate signs. The estimates were therefore deemed reasonable as well as
statistically significant. The parameter with the highest standardized coefficient was the
indicator with variable PSM1 (Client brief/input) and its parameter coefficient was 0.893.

354
Table 11.75: Factor Loading and Z-statistics of Product Selection and Design Management
Condition Measurement

Indicator Unstandardized Standardized Z- R Significant

2
Statistics at 5%
Variable Coefficient (λ) Coefficient (λ)
level?

PSM1 1.000 0.893 4.934 0.797 Yes

PSM2 0.851 0.843 4.014 0.710 Yes

PSM4 0.752 0.771 0.594

2.804 Yes
(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)
Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Product Selection Management features). In addition, the R2 values were
also close to the desired value of 1.00, indicating that the factors explained more of the variance
in the indicator variables. The results therefore suggest that the indicator variables significantly
predict the unobserved construct, because all the measured variables are significantly associated
with the Product Selection Management features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Product Selection Management features
construct were determined from the rho and the Cronbach’s alpha coefficient. According to
Kline (2005:59), the reliability coefficient should fall between zero and 1.00. Values close to
1.00 are desired. The rho coefficient of internal consistency was found to be 0.766. This value
was above the minimum required value of 0.70. Likewise, the Cronbach’s alpha was above the
minimum acceptable value of 0.70. The Cronbach’s alpha was found to be 0.872 (Table 11.76).
Both of these values revealed a high level of internal consistency and therefore reliability,
suggesting that the indicator variables represent the same latent construct (Product Selection
Management).

355
Table 11.76: Reliability and Construct Validity of Product Selection Management Feature
Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Product PSM1 0.893
Selection and 0.872 0.766
Design 0.843
Management PSM2
Features 0.771
PSM4

*Parameter estimates are based on standardized solutions

Source: Fieldwork (2018)

Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher, and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111). The standardized
parameter coefficient presented in Table 11.75 revealed that all coefficients were significantly
higher with the lowest being 0.771 for Product Selection Management features. The magnitude
of the parameter estimate was above the 50 per cent minimum. This indicates a strong
relationship between the indicator variables and the factors of the Product Selection Management
features construct. Therefore the Product Selection Management features satisfied both internal
reliability and the construct criteria. The rho value was above the minimum value of 0.70 and the
magnitude, signs and statistical significance of the parameter estimates were appropriate (Table
11.76).

356
Summary on Product Selection Management Feature Measurement Model
The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indices met the cut-off index criteria and all the parameter estimates were
statistically significant and feasible. Considering these criteria, the measurement model for the
Product Selection Management feature was found to adequately fit the sample data. Therefore,
there was no need to improve the measurement model before it could be included in the full
latent variable model. Hence, the Product Selection Management feature construct was
adequately measured by the indicator variables and could be used in the analysis of the full latent
variable model.

11.4.4.1.8 Measurement Model for Construction Process Management and Improvement

Features (CPMI) Construct
A unidimensional model for Construction Process Management and Improvement (CPMI)
features was presented (Table 11.33). From the 536 cases analyzed for this construct, four (4)
indicator variables (F7CPMI9, F7CPMI10, F7CPMI11, and F7CPMI12) made up of the same
factor were realized and the name CPMI was maintained (Table 11.33). All four (4) indicator
variables obtained were used for the CFA (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order
for a variable to be included in a CFA, thus enabling the model to be described as well-fitting,
the distribution of residuals covariance matrix should be symmetrical and centred around zero
(Byrne, 2006:94; Joreskog & Sorbom, 1988). The five-indicator model provides good measures
of residual matrix and evidence of convergent validity.

357
Figure 11.8: Measurement Model of Construction Process Management and Improvement

The CFA results further revealed that the Construction Process Management and Improvement
features had four (4) dependent variables, five (5) independent variables and eight (8) free
parameters. The number of fixed non-zero parameters was five (5). The four (4) dependent
indicator variables for the Construction Process Management and Improvement are Quality
control system, Institution of a continuous improvement process within the company,
Appropriate use of system structure and standards, and Efficient management system. These
indicator variables are presented in Table 11.77. The Construction Process Management and
Improvement features measurement model is shown in Figure 11.8 as analyzed before it could be
included in the full latent variable model.

In order to establish how well the model fit the sample data and the strength of the hypothesized
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the Construction Process Management and
Improvement features.

358
Table 11.77: Postulated Construction Process Management and Improvement Features
Model

Latent constructs Indicator variables Label

Construction Process Quality control system CPMI9
Management and Institution of a continuous CPMI10
Improvement (CPMI) improvement process within the
company
Appropriate use of system CPMI11
structure and standards
Efficient management system CPMI12
Source: Fieldwork (2018)

Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate

The unstandardized and standardized absolute residual matrix values of the Construction Process
Management and Improvement features are presented in Tables 11.78 and 11.79. The results
reveal that all the absolute residual values and the average off-diagonal absolute residual values
were close to zero. The unstandardized average off-diagonal residual was 0.0000 while the
standardized average off-diagonal residual was found to be 0.0003. A residual value greater than
2.58 is described as large (Byrne, 2006:94). The results obtained for the Construction Process
Management and Improvement Features measurement model suggested a fairly acceptable fit to
the sample data because the absolute residuals were all less than 2.58. In order for a model to be
described as well-fitting, the distribution of standardized residuals should be symmetrical and
centred around zero (Byrne, 2006:94).

Table 11.78: Residual Covariance Matrix for Construction Process Management and
Improvement Model (Unstandardized)

Unstandardized Residual Covariance Matrix

CPMI11 CPMI10 CPMI9 CPMI12

CPMI11 0.000
CPMI10 -0.005 0.000

359
 CPMI9 0.004 0.001 0.000
CPMI12 -0.001 0.003 -0.004 0.000
Average absolute residual = 0.0000
Average off-diagonal absolute residual = 0.0000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

Table 11.79: Residual Covariance Matrix for Construction Process Management and
Improvement Model (Standardized)

Standardized Residual Covariance Matrix

CPMI11 CPMI10 CPMI9 CPMI12

CPMI11 0.000
CPMI10 -0.141 0.000
CPMI9 0.124 0.020 0.000
CPMI12 -0.015 0.093 -0.109 0.000
Average absolute residual = 0.0002 0
Average off-diagonal absolute residual = 0.0003 .
0
% falling between -0.1 +0.1 = 99.99% 0
0
Source: Fieldwork (2018)
From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit, further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

360
Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)
The analysis strategy of goodness-of-fit for the Construction Process Management and
Improvement followed a three-statistics strategy of fit indexes as recommended by Hu and
Bentler (1999). The sample data on safe act of workers measurement model yielded the S – Bχ2
of 0.630 with 2 degrees of freedom (df) with a probability of p = 0.0000. This chi-square value
indicated that the departure of the sample data from the postulated measurement model was
significant and hence indicative of good fit. The chi-square test is very sensitive to sample size
and is used more as a descriptive index of fit rather than as a statistical test (Kline, 2005:136).
The normed chi-square value is usually adopted by most researchers. The normed chi-square is
the procedure of dividing the chi-square by the degrees of freedom. The normed values of up to
3.0 or even 5.0 are recommended (Kline, 2005:137). Values for NFI range between 0 and 1 with
Bentler and Bonnet (1980) recommended values greater than 0.90 indicating a good fit. Hu and
Bentler (1999) have given a cut-off criterion of NFI ≥ 0.95. This index is sensitive to sample
size, underestimating fit for samples less than 200 (Mulaik et al., 1989; Bentler, 1990), and it
does not recommended to be solely relied on (Kline, 2005). The non-normed fit index (NNFI,
also known as the Tucker-Lewis index) is an index that prefers simpler models. The value of the
NNFI can indicate poor fit despite other statistics pointing towards good fit if small samples are
used (Bentler, 1990; Kline, 2005; Tabachnick & Fidell, 2007). Owing to their non-normed
nature, NNFI values can go above 1.0 and can thus be difficult to interpret (Byrne, 1998). The
chi-square and degrees of freedom were found to be 0.315. This ratio was lower than the limit of
3.00 or 5.0 advocated for by some authors (Kline, 2005:137). The CFI value was found to be
1.000 which was slightly greater than the cut-off limit of 0.95 so the model is described to have
an acceptable fit. The NFI value was 0.999 which is slightly above the given cut-off value of NFI
≥ 0.95 is shown in Table 11.80. Therefore, the model is acceptable. The NNFI value obtained is
1.000 which is above the cut-off value of 0.80. These fit indexes for the Construction Process
Management and Improvement model suggested that the postulated model adequately describes
the sample data and could therefore be included in the full latent variable model analysis (Table
11.80).

361
Table 11.80: Robust fit Indexes for Construction Process Management and Improvement
Features Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.630

df 0≥ 2 Good fit

Normed chi-square ≤2 or 3 good fit 0.31

= x²/df ratio ≤5 acceptable
5
CFI 0.90≥ acceptable 1.000 Good fit

0.95≥ good fit

RMSEA Less than 0.05 with 0.000
confidence interval
(CI) 0.00-0.05 Good fit
“good fit” fit
0.95≥ good
NFI Greater than 0.90 0.999 Good fit
“good fit”
95%
NNFI Greater than 1.000 Good fit
0.80. “good fit”

SRMR Equal or less than 0.05 0.003 Good fit

“good
fit”

Equal or less
than 0.08 “acceptable
fit”

Source: Fieldwork (2018)

Statistical Significance of Parameter Estimates

Table 11.81 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs. The estimates were therefore deemed reasonable as well as statistically

362
significant. The parameter with the highest standardized coefficient was the indicator with
variable CPMI12 (Efficient management system) and its parameter coefficient was 0.827.

Table 11.81: Factor Loading and Z-statistics of Construction Process Management and
Improvement Measurement

Indicator Unstandardized Standardized Z- R2 Significant

Statistics at 5%
Variable Coefficient (λ) Coefficient
(λ) level?

CPMI9 1.366 0.798 3.046 0.622 Yes

CPMI10 1.369 0.795 3.034 0.627 Yes

CPMI11 1.310 0.793 3.346 0.632 Yes

CPMI12 1.411 0.827 4.034 0.701 Yes

(Robust Statistical Significance at 5% level)

Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Construction Process Management and Improvement features). In addition,
the R2 values were also close to the desired value of 1.00, indicating that the factors explained
more of the variance in the indicator variables. The results therefore suggest that the indicator
variables significantly predict the unobserved construct because all the measured variables are
significantly associated with the Construction Process Management and Improvement features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the Construction Process Management and
Improvement features construct was determined from the rho and the Cronbach’s alpha
coefficient. According to Kline (2005:59), the reliability coefficient should fall between zero and
1.00. Values close to 1.00 are desired. The rho coefficient of internal consistency was found to
be 0.824. This value was above the minimum required value of 0.70. Likewise, the Cronbach’s
363
alpha was above the minimum acceptable value of 0.70. The Cronbach’s alpha was found to be
0.875 (Table 11.82). Both of these values revealed a high level of internal consistency and
therefore reliability, suggesting that the indicator variables represent the same latent construct
(Construction Process Management and Improvement).

Table 11.82: Reliability and Construct Validity of Construction Process Management and
Improvement Feature Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Construction CPMI9 0.798
Process
Management CPMI10 0.795
and 0.879 0.824
Improvement CPMI11 0.793
Features
CPMI12 0.827

*Parameter estimates are based on standardized solutions

Source: Fieldwork (2018)

Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111). The standardized
parameter coefficient presented in Table 11.81 revealed that all coefficients were significantly
higher with the lowest being 0.793 for Construction Process Management and Improvement

364
features. The magnitude of the parameter estimate was above the 50 per cent minimum. This
indicates a strong relationship between the indicator variables and the factors of the Construction
Process Management and Improvement features construct. Therefore the Construction Process
Management and Improvement features satisfied both internal reliability and the construct
criteria. The rho value was above the minimum value of 0.70 and the magnitude, signs and
statistical significance of the parameter estimates were appropriate (Table 11.82).

Summary on Construction Process Management and Improvement feature Measurement

Model
The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indices met the cut-off index criteria and all the parameter estimates were
statistically significant and feasible. Considering these criteria, the measurement model for the
Construction Process Management and Improvement feature was found to adequately fit the
sample data. Therefore, there was no need to improve the measurement model before it could be
included in the full latent variable model. Hence, the Construction Process Management and
Improvement feature construct was adequately measured by the indicator variables and could be
used in the analysis of the full latent variable model.

11.4.4.1.9 Measurement Model for Construction Employees’ Involvement and Motivation

Features (CEIM) construct
A unidimensional model for Construction Employees Involvement and Motivation (CEIM)
features was presented (Table 11.33). From the 536 cases analyzed for this construct, three (3)
indicator variables (F8CEIM3, F8CEIM4, and F8CEIM5) made up of the same factor were
realized and the name CEIM was maintained (Table 11.33). All the three (3) indicator variables
obtained were used for the CFA (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a
variable to be included in a CFA, thus enabling the model to be described as well-fitting, the
distribution of residuals covariance matrix should be symmetrical and centred around zero
(Byrne, 2006:94; Joreskog & Sorbom, 1988). The three-indicator model provides good measures
of residual matrix and evidence of convergent validity.

365
Figure 11.9: Measurement Model of Construction Employees Involvement and Motivation

The CFA results further revealed that the Construction Employees Involvement and Motivation
features had three (3) dependent variables, four (4) independent variables and six (6) free
parameters. There werefour (4) fixed non-zero parameters. The three (3) dependent indicator
variables for the Construction Employees Involvement and Motivation are Salary/Position
promotion, Employee rewards, and Bonus scheme. These indicator variables are presented in
Table 11.84. The Construction Employees Involvement and Motivation features measurement
model shown in Figure 11.9 was analyzed before it could be included in the full latent variable
model.

In order to establish how well the model fit the sample data and the strength of the hypothesized
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the Construction Employees Involvement and
Motivation features.

366
Table 11.83: Postulated Construction Employees Involvement and Motivation

Features Model

Latent constructs Indicator variables Label

Construction Salary/Position promotion CEIM3
Employees
Involvement and Bonus scheme CEIM4
Motivation (CEIM)
Employee rewards CEIM5

Source: Fieldwork (2018)

Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate

The unstandardized and standardized absolute residual matrix values of the Construction
Employees Involvement and Motivation features are presented in Tables 11.84 and 11.85. The
result reveals that all the absolute residual values and the average off-diagonal absolute residual
values were close to zero. The unstandardized average off-diagonal residual was 0.000 while the
standardized average off-diagonal residual was found to be 0.000. A residual value greater than
2.58 is described as large (Byrne, 2006:94). The results obtained for the Construction Employees
Involvement and Motivation Features measurement model suggested a fairly acceptable fit to the
sample data because the absolute residuals were all less than 2.58. In order for a model to be
described as well-fitting, the distribution of standardized residuals should be symmetrical and
centred around zero (Byrne, 2006:94).

Table 11.84: Residual Covariance Matrix for Construction Employees Involvement and
Motivation Model (Unstandardized)

Unstandardized Residual Covariance Matrix

CEIM5 CEIM4 CEIM3

CEIM5 0.000
CEIM4 0.000 0.000
367
 CEIM3 0.000 0.000 0.000

Average absolute residual = 0.000

Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

Table 11.85: Residual Covariance Matrix for Construction Employees Involvement and
Motivation Model (Standardized)

Standardized Residual Covariance Matrix

CEIM5 CEIM4 CEIM3

CEIM5 0.000
CEIM4 0.000 0.000
CEIM3 0.000 0.000 0.000
Average absolute residual = 0.000
Average off-diagonal absolute residual = 0.000

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit, further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

368
Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)
The analysis strategy of goodness-of-fit for the Construction Employees Involvement and
Motivation measurement followed a three-statistics strategy of fit indexes as recommended by
Hu and Bentler (1999). The sample data on Construction Employees’ Involvement and
Motivation measurement model yielded the S – Bχ2 of 0.000 with 0.000 degrees of freedom (df)
with a probability of p = 0.000. This chi-square value indicated that the departure of the sample
data from the postulated measurement model was significant and hence indicative of good fit.
The chi-square test is very sensitive to sample size and is used more as a descriptive index of fit
rather than as a statistical test (Kline, 2005:136). The normed chi-square value is usually adopted
by most researchers. The normed chi-square is the procedure of dividing the chi-square by the
degrees of freedom. The normed values of up to 3.0 or even 5.0 are recommended (Kline,
2005:137). Values for NFI range between 0 and 1 with Bentler and Bonnet (1980) recommended
values greater than 0.90 indicating a good fit. Hu and Bentler (1999) have given a cut-off
criterion of NFI ≥ 0.95. This index is sensitive to sample size, underestimating fit for samples
less than 200 (Mulaik et al., 1989; Bentler, 1990), and is thus not recommended to be solely
relied on (Kline, 2005). The non-normed fit index (NNFI, also known as the Tucker-Lewis
index) is an index that prefers simpler models. The value of the NNFI can indicate poor fit
despite other statistics pointing towards good fit if small samples are used (Bentler, 1990; Kline,
2005; Tabachnick & Fidell, 2007). Owing to their non-normed nature, NNFI values can go
above 1.0 and can thus be difficult to interpret (Byrne, 1998). The chi-square and degrees of
freedom were found to be 0.000. This ratio was lower than the limit of 3.00 or 5.0 advocated for
by some authors (Kline, 2005:137). The CFI value was found to be 1.000 which was slightly
greater than the cut-off limit of 0.95 so the model is described to have an acceptable fit. The NFI
value was 1.000 which is slightly above the given cut-off value of NFI ≥ 0.95 is shown in Table
11.86. Therefore, the model is acceptable. The NNFI value obtained is 1.000 which is also above
the cut-off value of 0.80. These fit indexes for the Construction Employees Involvement and
Motivation model suggested that the postulated model adequately describes the sample data and
could therefore be included in the full latent variable model analysis (Table 11.86).

369
Table 11.86: Robust Fit Indexes for Construction Employees Involvement and Motivation
Features Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 0.000

df 0≥ 0.00 Good fit

0
Normed chi-square ≤2 or 3 good fit -
= x²/df ratio ≤5 acceptable
CFI 0.90≥ acceptable 1.000 Good fit

0.95≥ good fit

RMSEA Less than 0.05 with 0.000
confidence interval
(CI) 0.00-0.05 Good fit
“good fit”
0.95≥ good fit
NFI Greater than 0.90 1.000 Good fit
“good fit”
95%
NNFI Greater than 1.000 Good fit
0.80. “good fit”

RMSEA 95% CI 0.000: 0.076 Acceptabl

e range

SRMR Equal or less than 0.05 0.000 Good fit

“good
fit”

Equal or less
than 0.08 “acceptable
fit”

Source: Fieldwork (2018)

370
Statistical Significance of Parameter Estimates
Table 11.87 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs except CEIM5 the value of which (1.454) was slightly below 1.96. However,
the estimates were deemed reasonable as well as statistically significant. The parameter with the
highest standardized coefficient was the indicator with variable CEIM4 (Bonus scheme) and its
parameter coefficient was 0.932.

Table 11.87: Factor Loading and Z-statistics of Construction Employees Involvement and
Motivation Measurement

Indicator Unstandardize Standardized Z- R2 Significan

d Statistic t at 5%
Variable Coefficient (λ)
Coefficient (λ) s level?
CEIM3 1.000 0.706 2.102 0.500 Yes

CEIM4 1.590 0.932 18.390 0.911 Yes

CEIM5 1.100 0.695 1.454 0.482

Yes

(Robust Statistical Significance at 5% level)

Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (Construction Employees Involvement and Motivation features). In
addition, the R2 values were also close to the desired value of 1.00, indicating that the factors
explained more of the variance in the indicator variables except CEIM5 the value of which
(0.482) was slightly below 0.50. Although the CEIM5 value was slightly below 0.50, the results
still suggest that the indicator variables significantly predict the unobserved construct because all
the measured variables are significantly associated with the Construction Employees
Involvement and Motivation features.

371
Internal Reliability and Validity of Scores
The internal consistency and reliability of scores for the Construction Employees Involvement
and Motivation features construct was determined from the rho and the Cronbach’s alpha
coefficient. According to Kline (2005:59), the reliability coefficient should fall between zero and
1.00. Values close to 1.00 are desired. The rho coefficient of internal consistency was found to
be 0.838. This value was above the minimum required value of 0.70. Likewise, the Cronbach’s
alpha was above the minimum acceptable value of 0.70. The Cronbach’s alpha was found to be
0.837 (Table 11.88). Both of these values revealed a high level of internal consistency and
therefore reliability, suggesting that the indicator variables represent the same latent construct
(Construction Employees Involvement and Motivation).

Table 11.88: Reliability and Construct Validity of Construction Employees Involvement

and Motivation Condition Feature Model

Factor Indicator Factor Cronbach’s Rho

Variable Loading Alpha Coefficient
Construction CEIM3 0.706
Employees.
Involvement CEIM4 0.932 0.837 0.838
and
Motivation CEIM5 0.695
Features
*Parameter estimates are based on standardized solutions
Source: Fieldwork (2018)

Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).

372
Hence, a parameter coefficient should be 0.5 or higher and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111). The standardized
parameter coefficient presented in Table 11.87 revealed that all coefficients were significantly
higher with the lowest being 0.695 for Construction Employees Involvement and Motivation
features. The magnitude of the parameter estimate was above the 50 per cent minimum. This
indicates a strong relationship between the indicator variables and the factors of the Construction
Employees Involvement and Motivation features construct. Therefore the Construction
Employees Involvement and Motivation features satisfied both internal reliability and the
construct criteria. The rho value was above the minimum value of 0.70 and the magnitude, signs
and statistical significance of the parameter estimates were appropriate (Table 11.88).

Summary of Construction Employees Involvement and Motivation Feature Measurement

Model
The CFA revealed that the residual covariance estimates fell within the acceptable range.
Likewise, the robust fit indices met the cut-off index criteria and all the parameter estimates were
statistically significant and feasible. Considering these criteria, the measurement model for the
Construction Employees Involvement and Motivation feature was found to adequately fit the
sample data. Therefore, there was no need to improve the measurement model before it could be
included in the full latent variable model. Hence, the Construction Employees Involvement and
Motivation feature construct was adequately measured by the indicator variables and could be
used in the analysis of the full latent variable model.

11.4.4.1.10 Measurement Model for TQM features (TQM) Construct

A unidimensional model for TQM (TQM) features was presented (Table 11.33). From the 536
cases analyzed for this construct, four (4) indicator variables (F9TQMO1, F9TQMO4,
F9TQMO5, and F9TQMO6) were realized (Table 11.33). All the four (4) indicator variables
obtained were used for the CFA (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a
variable to be included in a CFA, thus enabling the model to be described as well-fitting, the
distribution of residuals covariance matrix should be symmetrical and centred around zero

373
(Byrne, 2006:94; Joreskog & Sorbom, 1988). The four-indicator model provides good measures
of residual matrix and evidence of convergent validity.

Figure 11.10: Measurement model of TQM

The CFA results further revealed that the TQM features had four (4) dependent variables, five
(5) independent variables and 8 free parameters. The number of fixed non-zero parameters was
five (5). The four (4) dependent indicator variables for the TQM are Greater reduction of
construction costs, Elimination of reworks, Defect-free product at first attempt, and Higher
safety standards. These indicator variables are presented in Table 11.89. The TQM features
measurement model shown in Figure 11.10 was analyzed before it could be included in the full
latent variable model.

In order to establish how well the model fit the sample data and the strength of the hypothesized
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardized residuals, fit statistics and statistical significance at a
probability level of 5 per cent were examined. In addition, the Cronbach’s alpha and the rho
coefficient of internal consistency were examined to determine the score reliability. Results of
these statistics are presented in the next section for the TQM features.

374
Table 11.89: Postulated TQM Features Model

Latent constructs Indicator variables Label

TQM Outcomes Greater reduction of construction TQMO1
(TQMO) costs

Elimination of reworks TQMO4

Defect-free product at first attempt TQMO5

Higher safety standards TQMO6

Source: Fieldwork (2018)

Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate

The unstandardized and standardized absolute residual matrix values of the TQM features are
presented in Tables 11.90 and 11.91. The result reveals that all the absolute residual values and
the average off-diagonal absolute residual values were close to zero. The unstandardized average
off-diagonal residual was 0.0007 while the standardized average off-diagonal residual was found
to be 0.0012. A residual value greater than 2.58 is described as large (Byrne, 2006:94). The
results obtained for the TQM features measurement model suggested a fairly acceptable fit to the
sample data because the absolute residuals were all less than 2.58. In order for a model to be
described as well-fitting, the distribution of standardized residuals should be symmetrical and
centred around zero (Byrne, 2006:94).

Table 11.90: Residual Covariance Matrix for TQM Model (Unstandardized)

Unstandardized Residual Covariance Matrix

TQMO6 TQMO5 TQMO4 TQMO1

TQMO6
D19.4 0.000
TQMO5 0.008 0.000
0.000
TQMO4 0.000 -0.010

375
 0.020 0.000
TQMO1
-0.017 0.000

Average absolute residual = 0.0400

Average off-diagonal absolute residual = 0.0007

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.

Table 11.91: Residual Covariance Matrix for TQM Model (Standardized)

Standardized Residual Covariance Matrix

TQMO6 TQMO5 TQMO4 TQMO1

TQMO6 0.000
TQMO5 0.246 0.000
-1.242 0.000
TQMO4 0.010
TQMO1 -0.456 0.005 0.430 0.000

Average absolute residual = 0.0002

Average off-diagonal absolute residual = 0.0012

% falling between -0.1 +0.1 = 99.99%

Source: Fieldwork (2018)

From this information, the results suggested a measurement model that was well fitting despite a
minimal discrepancy in fit between the hypothesized model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit, further tests of goodness-of-fit were
possible to make a conclusive decision on the fit and appropriateness of the measurement model.

376
Goodness-of-fit Statistics – Robust Maximum Likelihood (RML)
The analysis strategy of goodness-of-fit for the TQM followed a three-statistics strategy of fit
indexes as recommended by Hu and Bentler (1999). The sample data on TQM measurement
model yielded the S – Bχ2 of 4.926 with a 2 degree of freedom (df) with a probability of p =
0.0000. This chi-square value indicated that the departure of the sample data from the postulated
measurement model was significant and hence indicative of good fit. The chi-square test is very
sensitive to sample size and is used more as a descriptive index of fit rather than as a statistical
test (Kline, 2005:136). The normed chi-square value is usually adopted by most researchers. The
normed chi-square is the procedure of dividing the chi-square by the degrees of freedom. The
normed values of up to 3.0 or even 5.0 are recommended (Kline, 2005:137).

Values for NFI range between 0 and 1 with Bentler and Bonnet (1980) recommended values
greater than 0.90 indicating a good fit. Hu and Bentler (1999) have given a cut-off criterion of
NFI ≥ .95. This index is sensitive to sample size, underestimating fit for samples less than 200
(Mulaik et al., 1989; Bentler, 1990), and is thus not recommended to be solely relied on (Kline,
2005). The non-normed fit index (NNFI, also known as the Tucker-Lewis index) is an index that
prefers simpler models. The value of the NNFI can indicate poor fit despite other statistics
pointing towards good fit if small samples are used (Bentler, 1990; Kline, 2005; Tabachnick &
Fidell, 2007). Owing to their non-normed nature, NNFI values can go above 1.0 and can thus be
difficult to interpret (Byrne, 1998). The chi-square and degrees of freedom were found to be
2.463. This ratio was lower than the limit of 3.00 or 5.0 advocated for by some authors (Kline,
2005:137).

The CFI value was found to be 0.997 which was slightly greater than the cut-off limit of 0.95 so
the model is described to have an acceptable fit. The NFI value was 0.995 which is slightly
above the given cut-off value of NFI ≥ 0.95 and is shown in Table 11.92. Therefore, the model is
acceptable. The NNFI value obtained is 0.992 which is above the cut-off value of 0.80. These fit
indexes for the TQM model suggested that the postulated model adequately describes the sample
data and could therefore be included in the full latent variable model analysis (Table 11.92).

377
Table 11.92: Robust Fit Indexes for TQM Features Construct

Fit Index Cut-off value Estimate Comment

S – Bχ2 4.926

df 0≥ 2

Normed chi-square ≤2 or 3 good fit 2.463 Good fit

= x²/df ratio ≤5 acceptable
CFI 0.90≥ acceptable 0.997 Good fit
0.95≥ good fit
RMSEA Less than 0.05 with 0.052
confidence interval Good fit
(CI) 0.00-0.05
95% “good fit”
NFI Greater than 0.90 0.995 Good fit
“good fit”
NNFI Greater than 0.992 Good fit
0.80. “good fit”
RMSEA 95% CI 0.000 – 0.113 Slightly out
of range

SRMR Equal or less than 0.05 0.009 Good fit

“good
fit”

Equal or less than

0.08 “acceptable fit”
Source: Fieldwork (2018)

Statistical Significance of Parameter estimates

Table 11.93 shows the correlation values, standard errors and the test of statistics. All the
correlation values were less than 1.00 and all the Z-statistics were greater than 1.96 and show
appropriate signs except TQMO1 the value of which (1.803) was slightly below 1.96. However,
the estimates were deemed reasonable as well as statistically significant. The parameter with the
highest standardized coefficient was the indicator with variable TQMO5 (Defect-free product at
first attempt) and its parameter coefficient was 0.868.

378
Table 11.93: Factor Loading and Z-statistics of TQM Measurement

Indicator Unstandardize Standardized Z- R2 Significan

d Statistic t at 5%
Variable Coefficient (λ)
Coefficient (λ) s level?
TQMO1 1.421 0.768 2.017 0.602 Yes

TQMO4 1.228 0.771 2.394 0.594 Yes

TQMO5 1.333 0.868 0.748 Yes

4.667
TQMO6 1.000 0.766 0.583 Yes
3.450
(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)

Most of the parameter estimates had high correlation values close to 1.00. The high correlation
values suggest a high degree of linear association between the indicator variables and the
unobserved variable (TQM). In addition, the R2 values were also close to the desired value of
1.00, indicating that the factors explained more of the variance in the indicator variables. The
results therefore suggest that the indicator variables significantly predict the unobserved
construct because all the measured variables are significantly associated with the TQM features.

Internal Reliability and Validity of Scores

The internal consistency and reliability of scores for the TQM features construct were
determined from the rho and the Cronbach’s alpha coefficient. According to Kline (2005:59), the
reliability coefficient should fall between zero and 1.00. Values close to 1.00 are desired. The
rho coefficient of internal consistency was found to be 0.785. This value was above the
minimum required value of 0.70. Likewise, the Cronbach’s alpha was above the minimum
acceptable value of 0.70. The Cronbach’s alpha was found to be 0. 866 (Table 11.94). Both of
these values revealed a high level of internal consistency and therefore reliability, suggesting that
the indicator variables represent the same latent construct (TQM).

379
Table 11.94: Reliability and Construct Validity of TQM Feature Model

Factor Indicator Factor Cronbach’ Rho

Variable Loading s Coefficient
Alpha
TQM TQMO1 0.768
Outcomes
Features TQMO4 0.771
0.866 0.785
TQMO5 0.868

TQMO6 0.766

*Parameter estimates are based on standardized solutions

Source: Fieldwork (2018)

Furthermore, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25 per
cent of the total variance in the indicator variable being explained by the latent variable (factor).
Hence, a parameter coefficient should be 0.5 or higher and ideally 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111).

The standardized parameter coefficient presented in Table 11.94 revealed that all coefficients
were significantly higher with the lowest being 0.766 for TQM features. The magnitude of the
parameter estimate was above the 50 per cent minimum. This indicates a strong relationship
between the indicator variables and the factors of the TQM features construct. Therefore the
TQM features satisfied both internal reliability and the construct criteria. The rho value was
above the minimum value of 0.70, and the magnitude, signs and statistical significance of the
parameter estimates were appropriate (Table 11.94).

380
Summary of TQM Measurement Model
The CFA analysis revealed that the residual covariance estimates fell within the acceptable
range. Likewise, the robust fit indexes met the cut-off index criteria and all the parameter
estimates were statistically significant and feasible. Considering these criteria, the measurement
model for the TQM feature was found to adequately fit the sample data. Therefore, there was no
need to improve the measurement model before it could be included in the full latent variable
model. Hence the TQM feature construct was adequately measured by the indicator variables and
could be used in the analysis of the full latent variable model.

11.4.5 Structural Model – Testing of the Hypothesized SEM Model

Each of the nine latent factors to be included in the full latent model was working very well
based on the test indexes and the statistically significances of the parameter estimates. Then the
full structural model was tested, which included all nine factors (with tested indicator variables)
and the TQM outcome (manifest) variables for the study.

Once again, the CFA measurement model for latent constructs was tested in order to confirm
whether the indicators that have been used to measure one or more latent factors hold. Thus,
loadings of the indicators on the specific factors were examined to see how well each factor has
been specified in the context of the others.

Table 11.95: Reliability and Construct Validity of the Latent Variables

Latent No. of Indicator Parameter Cronbach’s Rho

Indicator Variable Coefficient Alpha Coefficient
(Exogenous Variables
Leadership
Factors) and LTM1 0.847
Top Management
LTM4 0.847
(LTM)
4 LTM8 0.826

LTM9 0.783

381
Company CSQM4 0.773
Supplier Quality
CSQM7 0.893
Management
3 CSQM8 0.834 0.963 0.751
(CSQM)

Client Focus and CFI1 0.825

Involvement
CFI2 0.823
(CFI)
3 CFI4 0.718

Company Quality CQSE1 0.813

System
CQSE2 0.832
Evaluation 3
CQSE3 0.741
(CQSE)
Company Vision CVPS5 0.828
and Plan 3 CVPS6 0.850
Statement
CVPS7 0.775
(CVPS)
Product Design PDM2 0.786
Management
PDM3 0.757
(PDM) 3
PDM6 0.718

Product Selection PSM1 0.893

Management
PSM2 0.843
3
(PSM)
PSM4 0.771

382
 Construction CPMI9 0.798
Process
CPMI10 0.795
Management and
Improvement 4 CPMI11 0.793
(CPMI)
CPMI12 0.827

Construction CEIM3 0.706

Employees’
3 CEIM4 0.932
Involvement and
Motivation CEIM5 0.695
TQM
(CEIM)Outcomes TQMO1 0.768
0.866 0.785
(TQMO)
TQMO4 0.771

4 TQMO5 0.868

TQMO6 0.766

Source: Fieldwork (2018)

Covariances between the latent factors are added to the model for any relationship that will be
examined when the structural model is tested. Also, covariances between the latent factors and
outcome variables are also added to rule out the possibility that any of them may serve as an
indicator of any of the proposed factors. As already indicated above (analysis of the
measurement models), the measurement models indicated that the models (latent variables
CFAs) worked well and it was therefore feasible to test the full latent variable model. The
question of whether measurement models should be checked before analysing the full SEM is
simply a strategy a researcher adopts (Hayduk & Glaser, 2000:122). Similarly, the question of
how many factors a construct should have is also debatable (Bollen, 1989; Hayduk & Glaser,
2000:122). However, assessing the measurement models first has an advantage. The first merit of
analyzing the latent variable measurement models separately before analyzing the full SEM
model is that the research is assured of a proper working measurement model before analyzing

383
the full SEM latent model. Hence, the researcher avoids the frustration of re-specifying the full
model if a solution cannot be obtained. According to Herting and Costner (2000:100), “…if a
CFA model cannot be satisfactorily fitted, moving to the structural model will provide no
additional guidance or benefit”. However, these observations as presented in the current study
were a pure confirmatory analysis and therefore recommendations were based on whether the
postulated priori model fit the sample data. Hence, not all the initially derived indicator variables
from the literature which were on the questionnaires were tested in the CFA, as the preliminary
residual covariance matrix (factor loadings) for some indicator variables of some latent
constructs were more than the recommended value. A residual covariance matrix value greater
than 2.58 is described as large (Byrne, 2006:94). In order for a model to be described as well
fitting, the distribution of the residuals should be symmetrical and centred around zero (Byrne,
2006:94; Joreskog & Sorbom, 1988).

11.4.6 Hypothesized Relation for the Structural Model

The hypothesized model (Model 2.0) was tested in which leadership and top management,
company supplier quality management, client focus and involvement, company quality system
evaluation, company vision and plan statement, product design management, product selection
management, construction process management and improvement and construction employees’
involvement and motivation were expected to define TQM. The hypothesized model was fitted
to the data for the entire sample and, as is the norm, covariances for all the exogenous factors and
variables were specified. The nine-factor model was fitted to the data with the RML method of
EQS and the model converged. As with all of the analyses presented in this study, the testing of
this model was based on the robust ML estimation and robust statistics were used to ascertain the
fit of the model. The robust solution adjusts for non-normality in the data. As is the norm in SEM
analyses (Kline, 2005), one variable loading per latent factor was set equal to 1.0 in order to set
the metric for that factor.

384
 Leadership and Top Management
(LTM)

Company Supplier Quality

Management (CSQM)

Client Focus and Involvement

(CFI)

Construction Process Management

and Improvement (CPMI)

Total Quality Management

Construction Employees Involvement (TQM)
and Motivation (CEIM)

Company Vision and Plan

Statement (CVPS)

Product Selection and Design

Management (PSDM)

Company Quality System

Evaluation (CQSE)

Figure 11.11: Hypothesized Model of TQM (Model 1.0)

11.4.7 Fit Statistics on the Structural Model

A CFA of the full latent model was conducted. The full structural model hypothesized that
leadership and top management features, company supplier quality management features, client
focus and involvement features, company quality system evaluation features, company vision
and plan statement features, product design management features, product selection management
features, construction process management and improvement features and construction
employees’ involvement and motivation features define TQM in the Ghanaian construction

385
industry. The SEM model is presented in Figure 11.12 (Model 2.0). Model 2.0 is founded on the
general hypothesis for the study, which is based on the fact that overall TQM is directly related
to the influence of the exogenous variables in predicting or determining overall TQM practices.
The theory and basis of the model were presented in Chapter 10 of the thesis. The number of
cases that were analyzed for the full latent variable Model 2.0 was 536. Out of the total sample
size, all the 536 cases had positive weights. The model had 33 dependent variables and 43
independent variables. It also had 66 free parameters and 43 numbers of fixed non-zero
parameters. The covariance matrix of the model was analyzed using the RML estimation
method. Raw data was used for the analysis.

The raw data was not transformed since data transformation can provide an incorrect
specification (Shook, Ketchen, Hult & Kacmar, 2004: 399). One alternative to transformation is
to use an estimation approach available in EQS (robust maximum likelihood) as already
discussed, which adjusts the model fit chi-square test statistics and standard errors of individual
parameter estimates.

11.4.7.1 Analysis of Residual Covariance Estimate

Investigation of the average absolute residual values of the structural model revealed that all the
absolute residual values and the average off-diagonal absolute residual values were close to zero.
The unstandardized average off-diagonal residual was 0.0814 while the standardized average off-
diagonal residual was found to be 0.0865 (refer to Appendix G). These residual values were
considered small as they were all less than 2.58 (Byrne, 2006:94). In addition, 99.99 per cent of
the residuals fell within the acceptable range of -0.1 and +0.1. The significance of this
distribution is that for a structural model to be described as well-fitting, the distribution of
residuals should be symmetrical and centred around zero (Joreskog &Sorbom, 1988; Byrne,
2006:94), which the analyzed data has displayed. From the above information, the results suggest
that the hypothesized structural model had a good-fit to the sample data. Overall, the model as a
whole appears to be quite well fitting. Therefore, since this initial assessment of the structural
model residuals indicated a good fit, further tests of goodness-of-fit were justified.

386
 11.4.7.2 Structural Model Goodness-of-fit Statistics – Robust Maximum Likelihood
The test of the hypothesis that TQM in the Ghanaian construction industry is a nine-factor
structure as depicted in Figure 11.12 (Model 2.0) via the sample data on the model yielded a
robust likelihood ratio test (S – Bχ2) of 3134.617 with 450 degrees of freedom. The associated p
value was less than 0.001 (p = 0.000) with a sample of 536 cases. The chi-square index
suggested that the difference between the hypothesized model and the sample data matrix was
significant, but not entirely adequate. However, the chi-square test (likelihood ratio test) of fit is
very sensitive and therefore could not be relied upon to determine model fit. The chi-square test
tends to be affected by the sample size with a propensity to reject models if the samples are large
(Joreskog & Sorbom, 1993). Yet, the analysis of covariance structure (SEM) is grounded in large
sample size theory (Byrne, 2006:96). As such, large sample sizes are critical to obtaining precise
parameter estimates, as well as to the tenability of asymptotic distribution approximations
(MacCallum et al., 1996; Byrne, 2006:96). Therefore a normed chi-square value is usually
adopted by most researchers (MacCallum et al., 1996; Bentler, 1999; Kline, 2005:137; Byrne,
2006;). Normed chi-square is the procedure of dividing the S – Bχ2 by the degrees of freedom.
The normed values of up to 3.0 or even 5.0 are recommended (Kline, 2005:137). From the above
chi-square and degrees of freedom values, the ratio was found to be 6.97:1. This ratio was
slightly above the limit of 5.0 but could be acceptable as advocated for by some authors (Kline,
2005:137; Byrne, 2006) and therefore indicative of a reasonable fit of the model.

Other fit indexes such as goodness-of-fit statistics indexes were also applied to determine
goodness-of-fit or acceptable fit of the model to the latent variables. The robust CFI index was
determined and found to be 0.897. The CFI index was equivalent to an upper limit of 0.90, but
was considered because the value was within the range for model acceptance. In addition, the
absolute fit index, NFI, was found to be 0.794 which indicated an adequate fit of the full
structural model to the sample data, while the NNFI was found to be 0.817, indicating a good fit
of the full structural model to the sample data. Another strategic approach is also considered
satisfactory to accept or reject a model (Hu & Bentler, 1999:28). Hence, SRMR statistics were
further used to decide on the acceptability of the model.

387
Table 11.96: Robust Fit Indexes for Structural Model 2.0

Fit Index Cut-off value Estimate Comment

S – Bχ2 3134.617

df 0≥ 450 Good fit

Normed chi-square ≤2 or 3 good fit 6.966 Acceptable

= x²/df ratio ≤5 acceptable range
CFI 0.90≥ acceptable 0.897 Acceptable
0.95≥ good fit

NFI Greater than 0.90 Acceptable

“good fit” 0.794

NNFI Greater than 0.817 Good fit

0.80. “good fit”

SRMR Equal or less than 0.05 0.039 Good fit

“good
fit”

Equal or less
than 0.08 “acceptable
fit”

Source: Fieldwork (2018)

SRMR fit index in this study was found to be 0.039, a value which is lower than a goodness-of-
fit index of 0.05. Hence, SRMR fit index indicated a good fit of the model to the latent variables.
Schreiber et al. (2006) and Hu and Bentler (1999) posit that an SRMR equal to or less than
0.05 indicates “good fit” and also equal to or less than 0.08 indicates “acceptable fit”.
Therefore, the goodness-of-fit statistics indexes (CFI, NFI, NNFI, SRMR and S – Bχ2) met the
condition for model acceptance (Table 11.96) except the RMSEA index @ 95%., which was
found to be just above the upper limit of 0.08 for the model to be described as acceptable. The
robust RMSEA with a 95 per cent confidence interval (lower bound value = 0.102 and the upper
bound value = 0.109) was found to be 0.106. MacCallum et al. (1996) informed that an RMSEA

388
below 0.08 shows a good fit. The model was found to be somewhat unstable with the RMSEA
index because of issues such as the high correlation between factors. However, the Lagrange
multiplier (LM) test conducted on the full latent model sample data did not reveal any significant
indicators of model mis-specification of the hypothesized parameters. In EQS, a model can be
said to be mis-specified if there are any mis-fitting parameters using the LM test (Byrne,
2006:112). The criterion is to identify any significant drop in the χ2 values of parameters. Also,
in univariate and multivariate analysis, the probability that a parameter estimate is equal to zero
should be less than 0.05 in order to be rejected. This is also an indication of mis-specification
according to Byrne (2006:112). Hence, inspection of the LM test output revealed that there were
no significant mis-fitting variables that would have warranted mode re-specification.

11.4.7.3 Internal Reliability and Construct Validity of the SEM Model

The rho coefficient and the Cronbach’s alpha coefficient were examined in order to establish
score reliability for the SEM model. According to Kline (2005:59), the reliability coefficient
should fall between zero and 1.00. Values close to 1.00 are desired. The rho coefficient of
internal consistency was found to be 0.751. This value was above the minimum required value of
0.70. Similarly, the Cronbach’s alpha was above the minimum acceptable value of 0.70 at 0.963.
Both of these values indicated a high degree of internal consistency and homogeneity (Table
11.97). These findings informed that the degree to which responses are consistent across all
indicator variables was statistically significant, indicating that the measures of the latent
variables total scores are the best possible unit of analysis for the exogenous variables which
predictor the endogenous variable (TQM). The construct validity for the SEM model was
determined by examining the magnitude of the parameter coefficients. High parameter
coefficients of greater than 0.5 indicate a close relation between the factor and an indicator
variable. However, a parameter coefficient has to be between 0.5 - 0.7 or greater to explain about
50 per cent of the variance in an indicator variable (Hair et al., 1998:111). Inspection of the
standardized parameter coefficients presented in Table 11.98 shows that they were significantly
high with a maximum of 0.932. The parameter estimate of 0.932 meant that the TQM accounted
for about 93.2 per cent of the variance CEIM 4 and was therefore indicative of a good fit
between the indicator variable and the factor, likewise in the other factors.
389
 11.4.7.4 Structural Model Hypothesis Testing
Besides assessing the goodness-of-fit of the structural model, the feasibility of a model can be
judged by a further inspection of the obtained solution and this involves inspection of the
statistical significance of the parameter estimates, standard errors and the test statistics (Raykov,
1991:501). Therefore, the rejection of the hypothesis depends on how reasonable parameter
estimates were in terms of their magnitude, signs and statistical significance. In addition, if the
output showed estimates that had correlation values greater than 1.00, had negative variances and
the correlation or covariances were not definitely positive, then they were said to be displaying
unreasonable estimates (Byrne, 2006:103). Likewise, the test statistics had to be greater than
1.96 based on the probability level of 5 per cent before the hypothesis can be rejected (Byrne,
2006:103). The test statistics reported was the parameter estimate divided by its standard error
and therefore it functions as a Z-statistics to test that the estimate is statistically different from
zero. Hence, the test was used to evaluate the hypothesis.

Testing the Influence of the Exogenous Variables on overall TQM

It was general hypothesized that TQM is related to the influence of the exogenous variables in
predicting the overall TQM in the construction industry using Ghana as a case study. Results
from the SEM analysis yielded support for the hypothesis. The hypothesized relationships
between all exogenous factors and the endogenous factor were found to be significant and they
all had definite positive directions. Inspection of the correlation values, standard errors and the
test statistic in Table 11.98 revealed that all standardized coefficient correlation values were not
greater than 1.00. All test statistics (Z-values) were greater than 1.96 (p<0.05) and the signs were
appropriate. They all have positive values (refer to Tables 11.98 and 11.99), suggesting that all
latent variables measured the overall TQM. The estimates were therefore reasonable as well as
statistically significant. Therefore, the general hypothesis that TQM is directly related to the
influence of the exogenous variables in predicting overall TQM in the Ghanaian construction
industry could not be rejected. The relationship between TQM and TQM indicators was found to
be the most significant. In order to determine whether each exogenous variable considerably
predicted the endogenous construct, an inspection of the interfactor correlation (R2) values were
examined, thus establishing the exogenous variables’ direct influence on the dependent variable
390
(presented in the subsequent sections). However, the overall results therefore suggest that the
exogenous variables considerably predict the endogenous variable (TQM). Further assessment of
the outcome variables of overall TQM revealed that all standardized factor loadings’ values were
generally large and statistically significant (values ranged from 0.695 to 0.932). However, the
interfactor correlation (R2) values were all statistically significant (values ranged from 0.482 to
0.911) as shown in Table 11.99. The variance accounted for in each measure by the endogenous
variable revealed that the scores were significant at 5 per cent level. The score results suggested
that the interfactor relationship between the manifest variables is strong and has significant levels
of correlations.

Testing the Direct Influence of Leadership and Top Management Features on Overall
TQM
Results from the CFA of the full structural model presented in Tables 11.97 and 11.98 yield
support for the general hypothesis. The relationship between the factors and the endogenous
variable (dependent variable) was found to be statistically significant at 5 per cent probability
level. On the other hand, all standardized parameter estimates showed high correlations values
close to 1.00, suggesting a high degree of linear association between the indicator variables and
the endogenous construct. Inspection of the R2 values for the Leadership and Top Management
variables revealed that the values were above 0.50. LTM1 (R2 = 0.717), LTM4 (R2 = 0.718),
LTM8 (R2 = 0.682) and LTM9 (R2 = 0.613) and close to the desired value of 1.00. Hence, the
direct influence of Leadership and Top Management factor on overall TQM is statistically
significant as shown in Table 11.99.

391
Table 11.97: Model 2.0 Factor Loadings and Z-statistics

Indicator Unstandardized Standardized (Z-values) Significant

Variable at 5%
Coefficient (λ) Coefficient (λ)
level?
LTM1 1.000 0.847 4.132 Yes

LTM4 1.111 0.847 3.361 Yes

LTM8 0.988 0.826 3.500 Yes

LTM9 0.995 0.783 2.417 Yes

CSQM4 0.895 0.773 4.547 Yes

CSQM7 1.000 0.893 2.467 Yes

0.834 3.100 Yes

CSQM8 0.959
CFI1 1.000 0.825 4.125 Yes

CFI2 0.979 0.823 3.532 Yes

CFI4 0.928 0.718 2.211 Yes

CQSE1 1.000 0.813 4.542 Yes

CQSE2 0.984 0.832 4.139 Yes

0.987 0.741 2.301

CQSE3 Yes
CVPS5 1.000 0.828 3.101 Yes
CVPS6 1.056 0.850 3.414 Yes
CVPS7
0.847 0.775
2.788 Yes
PDM 2 1.000 0.786 3.169 Yes

PDM 3 1.009 0.757 2.980 Yes

PDM 6 0.886 0.718
2.370 Yes

392
 PSM 1 1.000 0.893 4.934 Yes
PSM 2 0.851 0.843 4.014 Yes
PSM 4
0.752 0.771
2.804 Yes
CPMI9 1.366 0.798 3.046 Yes

CPMI10 1.369 0.795 3.034 Yes

CPMI11 1.310 0.793 3.346 Yes

CPMI12 1.411 0.827 4.034 Yes

CEIM3 1.000 0.706 2.102 Yes

CEIM4 1.590 0.932 18.390 Yes

CEIM5 1.100 0.695 1.982

Yes
TQMO1 1.421 0.768 2.017 Yes

TQMO4 1.228 0.771 2.394 Yes

TQMO5 1.333 0.868

4.667 Yes
TQMO6 1.000 0.766
3.450 Yes
(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)

** SEM Analysis Norm (Kline, 2005) - One variable loading per latent factor is set equal to 1.0
in order to set the metric for that factor

Testing the Direct Influence of Company Supplier Quality Management Features on

Overall TQM
Inspection of the R2 values for the Company Supplier Quality Management indicators revealed
that all the three indicator variables that were used to measure the latent factor had values close
to the desired value of 1.00. All the three variables, namely CSQM4 (R2 = 0.598), CSQM7 (R2 =
0.798) and CSQM8 (R2 = 0.696) were strong in predicting the endogenous variable (Table
393
11.99). This suggests that the interfactor relationship of these variables and other indicators in
determining the overall TQM is major. Furthermore, assessment of the variance accounted for in
each measure by the endogenous variable revealed that all scores were significant at 5 per cent
level. The reported parameter coefficient explained more than 25 per cent of the variance in the
latent variable, which was indicative of an adequate fit between the latent variables and the
endogenous construct. Thus, the score results suggested that the influence of this latent factor on
the endogenous variable was direct and significant.

Testing the Direct Influence of Client Focus and Involvement Features on Overall TQM
Inspection of the score values for this factor revealed that all standardized factor loadings were
generally large and statistically significant (values ranged from 0.718 to 0.825). The interfactor
correlation (R2) values were also moderately large and statistically significant (values ranged
from 0. 516 to 0.680) as shown in Table 11.98. Also, the variances accounted for in each
measure by the endogenous variable revealed that the scores were significant at 5 per cent level.
The values were above the minimum required value of 25 per cent. Hence, the score results
suggested that the influence of the Client Focus and Involvement on the endogenous variable
was direct and statistically significant.

Testing the Direct Influence of Company Quality System Evaluation Features on Overall
TQM
Assessment of the standardized factor loadings revealed that all values were generally large and
statistically significant (values ranged from 0.741 to 0.832). Also, the interfactor correlation (R2)
values were also moderate and statistically significant (values ranged from 0.549 to 0.692) as
shown in Table 11.98. The total variances accounted for in each indicator variable by the
endogenous variable revealed that the scores were significant at 5 per cent level. The score
results suggested that the influence of Company Quality System Evaluation in determining
overall TQM was direct and statistically significant.

394
Table 11.98: Model 2.0 Factor Loadings, Z-statistics, Variance Accounted for & Reliability
and Construct Validity

Indicator Standardized (Z- R2 Cronbach’s Rho

Variable values)
Coefficient (λ) Alpha Coefficient

LTM1 0.847 4.132 0.717

LTM4 0.847 3.361 0.718

LTM8 0.826 3.500 0.682

LTM9 0.783 2.417 0.613

CSQM4 0.773 4.547 0.598

CSQM7 0.893 2.467 0.798

CSQM8 0.834 3.100 0.696

CFI1 0.825 4.125 0.714

CFI2 0.823 3.532 0.655

0.963 0.751
CFI4 0.718 2.211 0.524

CQSE1 0.813 4.542 0.660

CQSE2 0.832 4.139 0.692

CQSE3 0.741 2.301 0.549

CVPS5 0.828 3.101 0.685

CVPS6
0.850 3.414 0.723
CVPS7
0.775 2.788 0.601

PDM 2 0.786 3.169 0.600

PDM 3 0.757 2.980 0.627

PDM 6 0.515
0.718 2.370
395
 PSM 1 0.893 4.934 0.797
PSM 2
0.843 4.014 0.710
PSM 4
0.771 2.804 0.594
CPMI9 0.798 3.046 0.622

CPMI10 0.795 3.034 0.627

0.632
CPMI11 0.793 3.346
0.701
CPMI12 0.827 4.034
CEIM3 0.706 2.102 0.500

CEIM4 0.932 18.390 0.911

CEIM5 0.695 1.982 0.482

TQMO1 0.768 2.017 0.602

TQMO4 0.771 2.394 0.594 0.866 0.785

TQMO5 0.748
0.868 4.667
TQMO6 0.583
0.766 3.450
(Robust Statistical Significance at 5% level)
Source: Fieldwork (2018)

Testing the Direct Influence of Company Vision and Plan Statement Features on Overall
TQM
The inspection of the standardized factor loadings revealed that all values were generally very
large and statistically significant significant (values ranged from 0.775 to 0.850) as shown in
Table 11.99). The R2 values were large and statistically significant (values ranged from 0.601 to
0.723). This suggests that the inter-factor relationship between the variables is significant. The
variance accounted for in each measure by the endogenous variable revealed that the scores were
significant at 5 per cent level. The score results suggested that the direct influence of Company
Vision and Plan Statement in determining overall TQM is statistically significant.

396
Testing the Direct Influence of Product Design Management Features on Overall TQM
Assessment of the standardized factor loadings revealed that all values were generally large and
statistically significant (values ranged from 0.718 to 0.786). Also, the interfactor correlation (R2)
values were also moderate and statistically significant (values ranged from 0.515 to 0.618) as
shown in Table 11.99. The total variances accounted for in each indicator variables by the
endogenous variable revealed that the scores were significant at 5 per cent level. The score
results suggested that the influence of Product Design Management in determining overall TQM
is statistically significant.

Testing the Direct Influence of Product Selection Management Features on Overall TQM
The inspection of the standardized factor loadings revealed that all values were generally very
large and statistically significant (values ranged from 0.771 to 0.893) as shown in Table 11.99.
The R2 values were large and statistically significant (values ranged from 0.594 to 0.797). This
suggests that the inter-factor relationship between the variables is significant. The variance
accounted for in each measure by the endogenous variable revealed that the scores were
significant at 5 per cent level. The score results suggested that the direct influence of Product
Selection Management in determining overall TQM is statistically significant.

Testing the Direct Influence of Construction Process Management and Improvement

Features on Overall TQM
The inspection of the standardized factor loadings revealed that all values were generally large
and statistically significant (values ranged from 0.793 to 0.827) as shown in Table 11.99. The R2
values were large and statistically significant (values ranged from 0.629 to 0.684). This suggests
that the inter-factor relationship between the variables is significant. The variance accounted for
in each measure by the endogenous variable revealed that the scores were significant at 5 per
cent level. The score results suggested that the direct influence of Construction Process
Management and Improvement in determining overall TQM is statistically significant.

397
Testing the Direct Influence of Construction Employees Involvement and Motivation
Features on Overall TQM
Assessment of the standardized factor loadings revealed that all values were generally large and
statistically significant (values ranged from 0.695 to 0.932). On the other hand, the R2 values
CEIM3 (R2 = 0.500) and CEIM4 (R2 = 0.911), were found to be moderate and very largely
statistically significant respectively. The exception is the indicator variable CEIM5 (R2 = 0.482)
which is the weakest among the variables. The result of R2 for CEIM5 suggests that this
indicator variable did not considerably predict the endogenous factor construct. Despite the non-
coherent level of the inter-factor correlation within the indicator variables, the direct influence of
Construction Employees’ Involvement and Motivation factor on overall TQM is statistically
significant because the scores were significant at 5 per cent level as shown in Table 11.98.

11.4.8 Model Explanation

The CFA structures in figure 11.12 comprise nine (9) unobserved latent factors for an integrated
TQM model (Leadership and Top Management (LTM), Company Supplier Quality Management
(CSQM), Client Focus and Involvement (CFI), Company Quality System Evaluation (CQSE),
Company Vision and Plan Statement (CVPS), Product Design Management (PDM), Product
Selection Management (PSD), Construction Process Management and Improvement (CPMI),
and Construction Employees’ Involvement and Motivation (CEIM)). These factors are termed
first order factors and consist of 29 observed variables. Four of them are considered to measure
LTM (LTM1, LTM4, LTM8 and LTM9), and CPMI (CPMI9, CPMI10, CPMI11 and CPMI12)
respectively, while three are considered to measure CSQM (CSQM4, CSQM7 and CSQM8),
CFI (CFI1, CFI2 and CFI4), CQSE (CQSE1, CQSE2 and CQSE3), CPVS (CPVS5, CPVS6 and
CPVS7), PDM (PDM2, PDM3 and PDM6), PSM (PSM1, PSM2 and PSM4), and CEIM
(CEIM3, CEIM4 and CEIM5) respectively. These 29 observed variables function as indicators of
their respective underlying latent factors. Associated with each observed variable is an error
termed (EEI – EE29) with factors being predicted (LTM, CSQM, CFI, CQSE, CPVS, PDM,
PSM, CPMI and CEIM). The error associated with observed variables represents measurement
error which reflects on their adequacy in measuring the related underlying factors (LTM, CSQM,
CFI, CQSE, CPVS, PDM, PSM, CPMI and CEIM).
398
In order to determine whether each exogenous variable considerably predicted the endogenous
construct (dependent variable), an inspection of the interfactor correlation (R2) values, standard
errors and the test statistic was done. The results revealed that all standardized coefficient
correlation values were not greater than1.00, the test statistics (Z-values) were greater than 1.96
(p<0.05), and the signs were appropriate. They all have positive values, suggesting that all latent
variables measured the overall TQM outcomes as shown in Figure 11.13. The relationships
between all exogenous factors and the endogenous factors in Figure 11.13 were found to be
significant and they all had definite positive directions. The overall results therefore suggested
that the exogenous variables considerably predict the endogenous variable (TQM). Hence, the
finalized model (Figure 11.13) revealed that the exogenous variables (LTM, CSQM, CFI, CQSE,
CVPS, PDM, PSD, CPMI, and CEIM) had a statistically significant influence in determining
TQM (TQM) for the Ghanaian construction industry although the exogenous variables such as
CEIM had a weak (indirect) influence on determining TQM for the Ghanaian construction
industry. However, the assessment of the variance accounted for in each measure by the
endogenous variable revealed that the scores were significant as the values were above the
minimum required value. Therefore the weak variable has a good indirect association with the
other latent variables in the prediction of overall TQM in the construction industry.

Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading) which represent the magnitude of correlation or
covariance between an item and a construct. Higher parameter coefficients were achieved which
also shows that the indicator variables have a stronger relationship with a construct and thus
converge at a common point and indicate a close relationship between the construct and an
indicator variable. The lines of covariances (Figure 11.13) indicate that the integrated holistic
influence of the latent variables determines overall TQM because they were all statistically
significant. It can therefore be concluded that the nine-factor model schematically portrayed in
Figures 11.12 and 11.13 (Model 2.0) represents an adequate description of TQM for the
Ghanaian construction industry.

399
Figure 12.12: Model 2.0 – An Integrated TQM Model

Model parameters (from left to right): LTM (4 indicator variables), CSQM (3 indicator
variables), CFI (3 indicator variables), CQSE (3 indicator variables), CVPS (3 indicator
variables), PSM (3 indicator variables), PDM (3 indicator variables), CPMI (4 indicator
variables), and CEIM (3 indicator variables)

400
Figure 11.13: Model 2.0 – An Integrated TQM Model Covariances Association

Covariance relationship (from left to right): LTM (4 indicator variables), CSQM (3 indicator
variables), CFI (3 indicator variables), CQSE (3 indicator variables), CVPS (3 indicator
variables), PSM (3 indicator variables), PDM (3 indicator variables), CPMI (4 indicator
variables), CEIM (3 indicator variables) and TQM (4 indicator variables)

401
11.4.9 Summary of SEM Model
Results from the EQS output revealed that the robust fit indexes, CFI, NFI, NNFI, and the
SRMR values were acceptable and met the cut-off index criteria except RMSEA the value of
which was slightly above the cut-off index criterion. On the other hand, the parameter estimates
were found to be statistically significant and reasonable. The postulated model, which
hypothesized that overall TQM, is directly related to the influence of the exogenous variables in
predicting or determining overall TQM fit the sample data adequately. In view of the fact that the
analysis was both exploratory and confirmatory of a prior model, there was no need to further
improve the structural model. Investigation of alternative models, such as the reduction of latent
variables, could be a matter for further studies as the current study was both an exploratory and
confirmatory analysis of a prior model. However, the Lagrange multiplier test did not unveil
significant indication of model mis-specification to demand a re-specification. Byrne (2006:112)
informs that for most models, model enhancement is purely a process that attempts to fine-tune
small features of the sample and does not essentially add value to an already fitted model, such
as the current model. Likewise, MacCallum et al. (1992:501) cautioned that “…when an initial
model fits well, it is probably unwise to modify it to achieve even better fit because
modifications may simply be fitting idiosyncratic characteristics of the sample”. Hence, the
presented model (Model 2.0) was therefore accepted with its level of fit. The lines of covariances
(Figure 11.13) indicate that the integrated holistic influence of the latent variables determines
overall TQM because they were all statistically significant.

11.5 CONCLUSION
The postulation for the overall model was that the overall TQM practice is directly related to the
influence of the exogenous (latent) variables in predicting or determining overall TQM. The
SEM results of the measurement model were presented in this chapter. These results were
obtained from an analysis of SEM to determine whether the indicator variables (questionnaire
items) actually measured the constructs that they were supposed to measure. In addition, results
were presented in order to establish whether the statistically significant number of factors for the
latent models was feasible. Likewise, the measurement model reliability and construct validity
were also reported. The analysis of the structural model (full latent model-SEM) was conducted,
402
which validates the hypothesized integrated holistically TQM model. The influence of the latent
variables on the endogenous variable was also reported. It was concluded that there is therefore
no need to further improve the fit of the structural model. Further findings from the SEM results
revealed that the exogenous variables influence or determine TQM in the Ghanaian construction
industry. Further, it was found that all the nine exogenous variables have a significant direct
influence on the endogenous variables. It can therefore be concluded that the nine-factor model
schematically portrayed in Figure 11.12 (Model 2.0) represents an adequate description of TQM
in the Ghanaian construction industry.

403
 CHAPTER TWELVE
12 DISCUSSION OF RESULTS
12.1 INTRODUCTION
The quantitative research findings with reference to the descriptive and inferential statistics are
discussed in detail and presented in this chapter. Also, the research hypotheses are tested based
on the SEM result analysis, validating the assumption that TQM is a nine-factor model
schematically portrayed in Figure 11.12 (Model 2.0). The findings from the SEM analysis which
model TQM as a nine-factor model showed that the factors of Leadership and Top Management,
Company Supplier Quality Management, Client Focus and Involvement, Company Quality
System Evaluation, Company Vision and Plan Statement, Product Selection Management,
Product Design Management, Construction Process Management and Improvement, and
Construction Employees’ Involvement and Motivation were found to have a significant influence
in predicting the TQM in the construction industry. However, the Construction Employees’
Involvement and Motivation feature had a weak (indirect) influence in predicting TQM in the
construction industry. Notwithstanding, the covariation with the other exogenous construct to
determine TQM in the construction industry was found to be statistically significant.

12.2 QUESTIONNAIRE SURVEY RESULTS

To successfully implement TQM in construction industry, it is important to identify the factors
required for the implementation process. The questionnaire survey of the current study was
designed purposely to determine the factors or attributes that influence or predict TQM in the
construction industry and to develop a holistic integrated TQM model for the Ghanaian
construction industry. The structural model results of the thesis hypothesis testing revealed that
the general hypothesis, which states that leadership and top management, company supplier
quality management, client focus and involvement, company quality system evaluation,
company vision and plan statement, product design management, product selection management,
construction process management and improvement, and construction employees’ involvement
and motivation jointly predict TQM in the Ghanaian construction industry could not be rejected.
In view of the hypothesis, the discussion section was structured in order to respond to the
research’s sub-questions.
404
Harris and McCaffer (2002) inform that TQM consists of all activities that managers carry out to
improve their quality and strategy such as quality control, quality planning, quality improvement
and quality assurance. It includes continuous improvement, training and re-training of staff,
customers’ satisfaction, top management support, defect-free products at first attempt,
elimination of reworks, and cost effectiveness, amongst others. Tang, Ahmed, Aoieoung and
Poon (2005), on the other hand, indicated that the five core principles of TQM are continuous
improvement, customer focus, employee participation, teamwork and process focus. These five
principles are related to the five drivers of change suggested by the Egan (1998) report, namely
committed leadership, integrated processes and teams, focus on the customer, commitment to
people and a quality-driven agenda. Hence, the need for top management commitment and
leadership is recognised by most prominent writers in the area of quality such as Deming (1986),
Garvin (1988), Aalbregtse et al. (1991), Berry (1991), Juran and Gryna (1993), Oakland (1993),
Bounds et al. (1994), Dale (1994), Dahlgaard et al. (1998), and Yusof (2000) . Many of these
writers have also found that top management support for quality was a key factor in quality
improvement. Therefore, the first observation that must be made is that the respondents in this
study are among the top management who most probably have worked in their respective firms
for a number of years as revealed by findings related to the position held in the firm by the
respondents (Table 11.5) and work experience in the construction industry (Table 11.4). Findings
from these aspect revealed that a combined percentage of 87.5% of the respondents formed part
of the top management (Project Managers = 16.2%, Site Engineers = 40.9% and Quality Control
Officers 2.4%) while a combined percentage of 12.5% were foremen and others. It was also
found that a combined percentage of 70% of the respondents have worked in their respective
firms for more than six years while 30% have worked in the construction industry for less than
six years. Therefore, their opinions could thus be used to make an inference.

Furthermore, Ahire and Ravichandran (2001) informed that the commitment of top management
is generally a preliminary point for implementing and practising TQM in order to enhance the
performance of an organization. Management at the top level of an organization is responsible
for the implementation of TQM. If top management refuses to get involved and support the new

405
philosophy, it will most likely fail. Top management must support this philosophy both
monetarily and morally or TQM will probably not survive (Walton, 1986). Hence, the most
critical factor contributes to a successful TQM programme is top management commitment
(Ramirez & Looney, 1993; Lewis & Smith, 1994).

12.2.1 Leadership and Top Management Features’ Influence on TQM

The research’s sub-question RO1 was put forward to determine the extent to which Leadership
and Top Management features influenced TQM in the construction industry. First, a descriptive
assessment of the Leadership and Top Management features revealed that 100.0 per cent of the
respondents indicated that top management participation in TQM activities, top management
support of TQM, the leadership style of managing employees, leadership initiatives, and top
management empowerment of employees towards TQM will contribute to TQM in the
construction industry. Another 99.8 per cent of the respondents indicated that the top
management interaction with workers will contribute to TQM in the construction industry. Also,
98.9 per cent of the respondents indicated that top management commitment to TQM and top
management knowledge and proper understanding of TQM will contribute to TQM in the
construction industry. A further assessment of the Leadership and Top Management features
revealed that 98.7 per cent of the respondents indicated leadership ability in solving quality-
related problems as a contributor to TQM in the construction industry (Table 11.11).
Furthermore, results from the structural model revealed that the relationship between the
Leadership and Top Management features and the endogenous variable (TQM) was found to be
statistically significant at a 5 per cent probability level.

On the other hand, all standardized parameter estimates showed high correlations values,
suggesting a high degree of linear association between the indicator variables and the
endogenous construct. Also the interfactor values for this variable were considerable, suggesting
that more than 50.0 per cent of the latent variable considerably predicted the endogenous factor
construct. The summarized result for this variable revealed that the latent factor has a direct
influence in determining overall TQM. The results suggest that most variables included in the

406
model have a significant effect on TQM. Further findings suggest that Leadership and Top
Management features are a significant determinant of TQM in the construction industry. In
addition to this statement, Perles (2002) asserted that top management’s role and managerial
leadership is one of the factors that determine the variation in the success rate of TQM
implementation. The Deming Prize, the European Quality Award (EQA), and the Malcolm
Baldrige National Quality Award (MBNQA) also recognize the crucial role of top management
leadership in creating the goals, values, and systems that guide the pursuit of continuous
performance improvement (Das et al., 2011). Although many of the activities occur in the lower
levels of the organization, only the leadership of top management is in a position to create the
necessary organizational culture that is capable of leading and supporting TQM actions among
employees from the lower levels of the organization (Serafimovska & Ristova, 2011). The same
notion is asserted by Flynn et al. (1995), namely that it is impracticable to adopt TQM and
improve performance without strong top management support. Therefore, as seen from the above
opinions, TQM literature widely supports the necessity of management commitment and
leadership for successful TQM implementation (Gonza´lez & Guille´n, 2002).

The implication of these findings is that overall TQM is a product of the direct influence of
Leadership and Top Management features. Therefore, the TQM in the Ghanaian construction
industry can be improved through Leadership and Top Management features established in this
study, namely the leadership style of managing employees, top management knowledge and
proper understanding of TQM, top management empowerment of employees to solve quality
problems and top management support of TQM. The present work concurs with the work of
Ahire and Ravichandran (2001) who argued that the commitment of top management is
generally a preliminary point for implementing and practising TQM in order to enhance the
performance of an organization. Hence, the most critical factor contributing to a successful TQM
programme is top management commitment (Ramirez & Looney, 1993; Lewis & Smith, 1994).

The findings emanating from the Leadership and Top Management features assessment were
therefore significant, and when attention is given to the Leadership and Top Management
features identified in this study, the much desired TQM in the construction industry will be
407
achieved. The study also recognized that management at the top levels of an organization is
responsible for the implementation of TQM. If top management refuses to get involved and
support the new philosophy, it will most likely fail. Hence, top management must support this
philosophy both monetarily and morally or TQM will probably not survive (Walton, 1986).

12.2.2 Company Supplier Quality Management Features’ Influence on TQM

The findings suggested that Supplier Quality Management features have a direct influence on the
prediction of TQM in the construction industry. These findings support the views of Hackman
and Wageman (1995) who stated that building relationships with suppliers is one of the main
TQM implementation practices. Findings from the interfactor relationship revealed that Supplier
Quality Management features had a significant association with the latent variables in predicting
the endogenous variable. From the assessment of the variance accounted for in each measure by
the endogenous variable it was revealed that all scores were statistically significant at 5 per cent
level (Table 11.47). The reported parameter coefficient explained more than the baseline level of
the variance in the latent variable, which was indicative of an adequate prediction of the
endogenous construct. Hence, these results suggested that the influence of this latent factor on
the endogenous variable was direct and statistically significant.

A descriptive assessment of the Supplier Quality Management features also revealed that the
nine indicator variables used in measuring the Supplier Quality Management construct were
highly effective as shown in Table 11.12. Supplier selection criteria and partnership with
suppliers was the most effective item with 100.0 per cent followed by supplier quality audit,
supplier knowledge of TQM, supplier communication, participation of suppliers in TQM
activities, and suppliers’ orientation on TQM with 99.8 per cent. The lowest effective items were
supplier performance evaluation and suppliers’ commitment to TQM with 99.4 per cent. The
findings suggest that the respondents were in agreement with the Supplier Quality Management
features. These findings support the views of Deming (1986) who strongly suggested that
working with the supplier as a partner in a long-term relationship of loyalty and trust enhances
the quality of incoming materials and decreases costs. Mann (1992) and Zhan (2000a) also added
that companies should take part directly in suppliers’ activities related to quality. Deming (1986)
408
and Ishikawa (1985) therefore suggested that companies work directly with suppliers to ensure
that their materials are of the highest possible quality.

The implication of these findings is that overall TQM is a product of the direct influence of
Supplier Quality Management features and that TQM in the construction industry is paramount.
The findings originating from the Supplier Quality Management features’ assessment were
therefore significant. Anderson et al. (1994a) pointed out that external collaboration between a
company and its suppliers has benefits in the just-in-time buying systems and in the case of a
poor quality situation such information can be used by the supplier and the company to come up
with corrective action.

12.2.3 Client Focus and Involvement Features Influence on TQM

The finding was that Client Focus and Involvement features have a direct positive influence on
TQM in the construction industry. The SEM analysis of the study established three indicator
variables for Client Focus and Involvement. These indicator variables are Client brief/input,
Client complaint information/feedback, and Client satisfaction survey. The SEM analysis for
these variable indicators revealed that all the three indicators were closely associated with the
dependent variable. This is apparent from the standardized factor loadings and interfactor
correlations with other indicator variables, which were statistically significant (Table 11.52).
Also, the variances accounted for in each measure by the endogenous variable revealed that the
scores were statistically significant and the values were above the minimum required value of 25
per cent to be ascribed as an influence on TQM. The findings suggested that Client Focus and
Involvement features have a direct positive influence on the overall TQM. Similarly, a
descriptive assessment of the Client Focus and Involvement features also revealed that the
indicator variables used in measuring the Client Focus and Involvement construct were highly
effective as shown in Table 11.13. The finding was consistent with that of Ishikawa (1985) who
stated that to achieve quality, it is essential to know what customers need and to provide products
that meet their requirements. Hackman and Wageman (1995) also added that obtaining data
about customers is one of the most commonly used TQM implementation practices.

409
The implication of these findings is that overall TQM is a product of the direct influence of
Client Focus and Involvement features. The finding was significant in that it will provide the
firms with a full understanding of what the customer needs and wishes now and in the future so
that products and services can be designed to satisfy those needs and wishes. The overall results
suggested that the influence of Client Focus and Involvement in determining TQM is direct and
statistically more significant than any other factor. Similarly, the findings are significant because
the key to quality management is keeping a close connection with the customer in order to fully
establish the customer’s needs, as well as to receive feedback on the degree to which those needs
are being met. Client Focus and Involvement will help to define the extent to which a company
continuously satisfies client or customer desires and expectations. A successful company
recognizes the need to put the customer first in every decision taken (Philips Quality, 1995).
Therefore the customer should be closely involved in the product design and development
process with input at every point so that there is less probability of quality problems once full
production begins (Flynn et al., 1994). The insights gained from this finding can clearly help the
firm to improve quality.

12.2.4 Company Quality System Evaluation Features’ Influence on TQM

The concept of Company Quality System Evaluation can be defined as a systematic examination
of the extent to which an entity is capable of fulfilling specified requirements (ISO 8402, 1994).
Juran and Gryna (1993) affirmed that a formal evaluation of quality provides a starting point by
offering an understanding of the size of the quality issue and the areas demanding attention. It is
important to make it known that the Company Quality System Evaluation offers an important
base for the company to enhance its quality management practices. Descriptive assessment of the
Company Quality System Evaluation features as presented in Table 11.14 revealed the
following: Evaluation of overall company performance (100%), Evaluation of the quality
strategy framework (100%), Evaluation of quality system procedures (100%), Evaluation of the
quality information system (100%), Evaluation of end results (100%), and Evaluation of quality
costs (100%). Further findings from the descriptive statistics show the presence of the following
Company Quality System Evaluation: Evaluation of employee performance (99.8%), Evaluation
of the quality manual (99.6%) and Evaluation of the quality control system (98.9%). In addition,
410
the SEM analysis of the study established three indicator variables for Company Quality System
Evaluation. These indicator variables are Evaluation of the quality strategy framework,
Evaluation of overall company performance and Evaluation of employee performance. The SEM
analysis for these variable indicators revealed that all the three indicators were closely associated
with the dependent variable. This is apparent from the standardized factor loadings and
interfactor correlations with other indicator variables, which were statistically significant (Table
11.58). Also, the variances accounted for in each measure by the endogenous variable revealed
that the scores were statistically significant and the values were above the minimum required
value of 25 per cent to be ascribed as an influence on TQM. The findings suggested that
Company Quality System Evaluation features have a direct positive influence on the overall
TQM. Hence, such Company Quality System Evaluation information should be communicated
to workers in order to encourage workers to make things better. The study findings concurred
with the work of Juran and Gryna (1993) and Mann (1992) who opined that in order to have an
effective evaluation, a quality information system is essential as it is an organized method of
collecting, storing, analyzing, and reporting information on quality to assist decision-makers at
all levels.

The implication of these findings is that overall TQM is a product of the direct influence of
Company Quality System Evaluation features. The finding was significant in that based on
Company Quality System Evaluation activities the company can amend its business strategy in
order to keep it dynamic as stated by Mann (1992). It should be noted that evaluating the
situation in a company’s quality management practices offers an important base for the company
to enhance its quality management practices. Furthermore, Company Quality System Evaluation
can help to recognize the difference between actual performance and the goal of the company.

12.2.5 Company Vision and Plan Statement Features’ Influence on TQM

The importance of including this variable in the model is to determine the Company Vision and
Plan Statement features’ influence on TQM in the construction industry. Descriptive assessment
of the Company Vision and Plan Statement features as presented in Table 11.15 revealed that
100.0 per cent of the respondents indicated that all the seven variable indicators are influential in
411
determining TQM in the construction industry. It is notable that all the Company Vision and Plan
Statement features have a MS > 4.00 ≤ 5.00, which indicates that the respondents perceive the
Company Vision and Plan Statement features to be between ‘very influential’ and ‘extremely
influential’. Furthermore, SEM analysis was conducted to further determine statistical fitness of
the Company Vision and Plan Statement features. The following three Company Vision and Plan
Statement features were found to be statistically fit for the overall structural modeling of the
study, namely concreteness of the future plan, employee contribution to the vision, and
involvement of employees in the development of the vision statement. The SEM results for this
exogenous variable revealed that the standardized factor values and interfactor correlations for
the Company Vision and Plan Statement latent factors were large and statistically significant
(Table 11.64). Inspection of the total variances accounted for in each measure by the endogenous
variable revealed that the scores were also significant. This result agreed with the findings of
Kanji and Asher (1993) who stated that all employees should involve themselves in the
development of the vision statement to be able to understand how they can contribute to the
vision.

The relatively high agreement achieved by the respondents with regard to Company Vision and
Plan Statement variables suggests that these variables are very significant in driving TQM
implementation among construction firms in Ghana. The implication of these findings is that
overall TQM is a product of the direct influence of Company Vision and Plan Statement features.
The finding was also significant in that a vision statement explains how a company wants to be
seen in its chosen business. In this regard, it explains standards, principles, and values. In
addition, a vision statement is the advertisement of the intention to change. As such, it drives the
company ahead and acts against complacency. Meredith and Shafer (1999) explained that the
purpose of a vision and plan statement is to communicate the firm’s values, principles,
aspirations and purpose so that workers can make decisions that are consistent with and
supportive of these objectives.

412
12.2.6 Product Design Management Features’ Influence on TQM
This research’s sub-question was put forward to determine the extent to which Product Design
Management features influenced TQM in the construction industry. First, a descriptive
assessment of the Product Design Management features revealed that all the MSs are above the
midpoint score of 3.00, which indicates that the respondents agreed with the Product Design
Management features of TQM implementation. It is notable that all the product design
management features have a MS > 3.50 ≤ 5.00, which indicates that the respondents perceive the
Product Design Management features to be ‘very influential’. The relatively high MS = 3.83-
4.23 achieved suggests that these variables are very significant in driving TQM implementation
among construction firms in Ghana (see Table 11.16). Furthermore, results from the structural
model revealed that the relationship between the Product Design Management features and the
endogenous variable (TQM) was found to be statistically significant at 5 per cent probability
level.

On the other hand, all standardized parameter estimates showed high correlations values,
suggesting a high degree of linear association between the indicator variables and the
endogenous construct. Also the interfactor values for this variable were considerable, suggesting
that more than 50.0 per cent of the latent variable considerably predicted the endogenous factor
construct. The summarized result for this variable revealed that the latent factor has a direct
influence in determining overall TQM. The results suggest that most variables included in the
model have a significant effect on TQM. Further findings suggest that Product Design
Management features are a significant determinant of TQM in the construction industry. The
study established the following three variables (strength required for the end product, intended
purpose of the material and design detailing) as the most significant determinants of Product
Design Management which in effect influenced the overall TQM model. These findings concur
with the work of Juran and Gryna (1993) who asserted that sound product design can add to the
improvement of product quality. The findings emanating from the Product Design Management
features’ assessment were significant, and when attention is given to the Product Design
Management features identified in this study, the much desired TQM in the construction industry
will be achieved.

413
The implication of these findings is that overall TQM is a product of the direct influence of
Product Design Management features. Therefore, the TQM in the Ghanaian construction
industry can be improved through Product Design Management features established in this study,
namely strength required for the end product, intended purpose of the material and design
detailing.

12.2.7 Product Selection Management Features’ Influence on TQM

The importance of including this variable in the model is to determine the Product Selection
Management features influence on TQM in the construction industry. A descriptive assessment
of the Product selection Management features as presented in Table 11.16 revealed that that all
the MSs were above the midpoint score of 3.00, which indicates that the respondents agreed with
the Product Selection Management features of TQM implementation. It is notable that all the
Product Selection Management features have a MS > 3.50 ≤ 5.00, which indicates that the
respondents perceive the Product Selection Management features to be ‘very influential’. The
relatively high MS = 3.83- 4.23 achieved suggests that these variables are very significant in
driving TQM implementation among construction firms in Ghana (see Table 11.16). On the
other hand, SEM analysis was conducted to further determine statistical fitness of the Product
Selection Management features. The following three Product Selection Management features
were found to be statistically fit for the overall structural modelling of the study: client
brief/input, cost of selection and design of the product, and appearance/finishes required. The
SEM results for this exogenous variable revealed that the latent factors were large and
statistically significant (Table 11.76). Inspection of the total variances accounted for in each
measure by the endogenous variable revealed that the scores were also significant. These
findings concur with the work of Juran and Gryna (1993) who asserted that good product
selection can add to the improvement of product quality. The findings emanating from the
Product Selection Management features assessment were significant, and when attention is given
to the Product Selection Management features identified in this study, the much desired TQM in
the construction industry will be achieved.

414
The relatively high agreement achieved by the respondents with regard to Product Selection
Management variables suggests that these variables are very significant in driving TQM
implementation among construction firms in Ghana. The implication of these findings is that
overall TQM is a product of the direct influence of Product Selection Management features.
Therefore, the TQM in the Ghanaian construction industry can be improved through Product
Selection Management features established in this study, namely client brief/input, cost of
selection and design of the product, and appearance/finishes required.

12.2.8 Construction Process Management and Improvement features influence on TQM

This research’s sub-question was put forward to determine the extent to which Construction
Process Management and Improvement features influenced TQM in the construction industry.
First, a descriptive assessment of the Construction Process Management and Improvement
features revealed that all the MSs were above the midpoint score of 3.00, which indicates that the
respondents agreed with the Construction Process Management and Improvement features of
TQM. It is notable that all the Construction Process Management and Improvement features have
a MS > 4.00 ≤ 5.00, which indicates that the respondents perceive the Construction Process
Management and Improvement features to be between ‘very influential’ and ‘extremely
influential’. The relatively high MS = 4.02 - 4.33 achieved suggests that these variables are very
significant in driving TQM implementation among construction firms in Ghana (Table 11.11).
Furthermore, results from the structural model revealed that the relationship between the
Construction Process Management and Improvement features and the endogenous variable
(TQM) was found to be statistically significant at 5 per cent probability level.

On the other hand, all standardized parameter estimates showed high correlations values,
suggesting a high degree of linear association between the indicator variables and the
endogenous construct. Also the interfactor values for this variable were considerable, suggesting
that more than 50.0 per cent of the latent variable considerably predicted the endogenous factor
construct. The summarized result for this variable revealed that the latent factor has a direct
influence in determining overall TQM. The results suggest that most variables included in the
model have a significant effect on TQM. Further findings suggest that Construction Process

415
Management and Improvement features are a significant determinant of TQM in the construction
industry. The study established the following four variables, namely quality control system,
institution of a continuous improvement process within the company, appropriate use of system
structure and standards and efficient management system, as the most significant determinants of
Construction Process Management and Improvement which in effect influenced overall TQM
model. These findings concur with the work of Mann (1992) who posited that institution of a
continuous improvement process within the company and the appropriate use of a system
structure improved TQM. The findings emanating from the Construction Process Management
and Improvement features assessment were significant, and when attention is given to the
Construction Process Management and Improvement features identified in this study, the much
desired TQM in the construction industry will be achieved.

The implication of these findings is that overall TQM is a product of the direct influence of
Construction Process Management and Improvement features. Therefore, the TQM in the
Ghanaian construction industry can be improved through Construction Process Management and
Improvement features established in this study, namely quality control system, institution of a
continuous improvement process within the company, appropriate use of system structure and
standards and efficient management system.

12.2.9 Construction Employees’ Involvement and Motivation Features’ Influence on TQM

The SEM analysis for this construct Construction Employees’ Involvement and Motivation
revealed that only two indicator variables were closely associated with the dependent variable.
The other variable was weak in predicting the endogenous variable. The statistical assessment
suggests that the direct influence of these factors on the endogenous variable was weak
(indirect). However, a further assessment of the internal consistency and reliability of scores for
the Construction Employees’ Involvement and Motivation features construct was determined
from the rho and the Cronbach’s alpha coefficient. The assessment of the variance accounted for
in each measure by the endogenous variable revealed that the scores were significant as the
values were above the minimum required value. Therefore the weak variable has a good indirect

416
association with the other latent variables in the prediction of overall TQM in the construction
industry. Further, construct validity was determined from the magnitude and reasonableness of
the parameter coefficients (factor loading) (Table 11.88). The parameter coefficients represent
the magnitude of correlation or covariance between an item and a construct. Higher parameter
coefficients were achieved which also show that the indicator variables have a stronger
relationship with a construct and thus converge at a common point. Parameter coefficients were
found to be greater than 0.5 indicate a close relationship between the construct and an indicator
variable.

Similarly, descriptive assessment of the Construction Employees’ Involvement and Motivation

features revealed that all the MSs were above the midpoint score of 3.00, which indicates that the
respondents agreed with the Construction Employees Involvement and Motivation features of
TQM implementation. It is notable that all the Construction Employees’ Involvement and
Motivation features have a MS > 4.00 ≤ 5.00, which also indicates that the respondents perceive
the Construction Employees Involvement and Motivation features to be between ‘very
influential’ and ‘extremely influential’. The relatively high MS = 4.04- 4.29 achieved suggests
that these variables are very significant in driving TQM implementation among construction
firms in Ghana. These findings agree with the work of Armstrong (2006) who posits that close
attention should be given to individuals and they should best be motivated through means such
as incentives, rewards, and promotions among others and the organizational context within
which they carry out the work. The findings also supported work of Laufer and Moore (1983)
who advocated the use of financial incentive programmes to improve construction labour
productivity, reinforcing Maloney’s (1982) thesis of driving forces that lead to productivity
improvements. On the other hand, the findings disagreed with the work of McClelland (1961)
who suggested that achievement, affiliation, competence and power are four types of
motivational drives that are found in individuals who are self-motivated and this may be the case
for many construction workers.

417
The implication of these findings is that overall TQM is a product of the direct influence of
Construction Employees’ Involvement and Motivation features. The finding is significant in that
lack of motivation in turn affects TQM. The finding is also significant in that Construction
Employees’ Involvement and Motivation features will define the set of processes that determine
the choices people make about their behaviour towards TQM implementation in construction
industry. According to McClelland (1961), individuals tend to develop certain motivational
drives in the cultural environment in which they live and these drives affect the way people view
their jobs and TQM. Hence, the variables established under this construct will serve as standards
for managers of construction firms who are ready to implement TQM for their projects.

12.2.10 Other Questionnaire survey results

Level of TQM implementation in the Ghanaian construction industry
This research’s sub-question was put forward to determine the level of TQM implementation in
the Ghanaian construction industry in terms of percentage responses. Descriptive assessment of
this sub-question revealed that the majority of the respondents (38.1%) indicated that their firms
have implemented TQM to a moderate extent, followed by (30.4%) indicating that their firms
have implemented TQM in a large extent and the statement ‘To no extent”’constituted the least
level TQM implementation by the firms, namely 2.4 per cent of the sample. This suggests that
not all the firms under study have implemented TQM. This finding is in line with the work of
McIntyre and Kirschenman (2000) that surveyed the acceptance of TQM in upper Midwestern in
the United States. Their study revealed that 27.5% of the respondents did not employed TQM
practices.The finding also partially supports the work of Ansah, Aigbavboa and Thwala (2015)
who stated that the construction industry which includes Ghanaian construction industry is slow
in implementing the TQM concept despite its identified potential benefits. Agha (2011) opines
that the reason for the late arrival of the construction industry to the TQM approach is that
construction professionals are not familiar with its principles and techniques. However, Boaden
and Dale (1992) note that the relative immaturity of the construction industry in adopting TQM
may be an advantage if companies learn from their mistakes and the best practices of other
organizations.

418
Project Outcomes that Motivate Construction Companies to Implement TQM
This research’s sub-question was put forward to determine the project outcomes that motivate
construction companies to implement TQM. A descriptive assessment of this sub-question
revealed that all the MSs were above the midpoint score of 3.00, which indicates that the
respondents agreed with the project outcomes that motivate firms to implement TQM in the
Ghanaian construction industry. It is notable that all the ten ranked motivating factors
influencing the implementation of TQM in the construction industry have a MS > 4.00 ≤ 5.00,
which indicates that the respondents perceive the motivating factors influencing the
implementation of TQM in the construction industry to be between ‘high motivating factor’ and
‘very high motivating factor’. All the rankings of motivating factors that motivate the
implementation of TQM in the construction industry indicate MS> 4.08 - 4.35. The relatively
high MS achieved suggests that all the motivating factors that motivate the implementation of
TQM in the construction industry are very significant in driving TQM implementation among
construction firms in Ghana. Among the ten factors identified, higher productivity was ranked
first, good methods of working was ranked second, higher product quality and higher reduction
of construction waste were ranked third, higher safety standards and higher profitability were
ranked fourth, followed by defect-free product at first, elimination of reworks, greater reduction
of construction costs and improvement schedule performance which was ranked last as shown in
Table 11.19.

The literature indicates that the motivating factors that influence the implementation of TQM are
also the benefits that TQM brings at the end of project execution. The benefits include higher
customer satisfaction, reduction in construction costs, improved employee job satisfaction,
improved schedule performance, improved relationships with subcontractors, reduced rework,
improved safety, higher productivity, lower employee turnover, speeding up construction work,
improved methods of working, better control over the construction process, gaining competitive
advantage, increase profitability, decreasing waste and rework, better coordination of activities
and being more customer focused (Hassin et al., 2006; Love et al., 2000; McIntyre &
Kirschenman, 2000; Al-Momani, 2007; Khadour & Darkwa, 2008; Chini & Valdez, 2003; Love
et al., 2004). In considering the application of TQM to construction project schedule, Kiwus and
419
Williams (2001) also conclude that TQM techniques may reduce the frequency and severity of
schedule overruns.

Reasons for Non-Implementation (Non-Adoption) of TQM in Ghanaian Construction

Industry

This research’s sub-question was put forward to determine the reasons for the non-
implementation (non-adoption) of TQM in the Ghanaian construction industry. A descriptive
assessment of this sub-question revealed that all the MSs were above the midpoint score of 3.00,
which indicates that the respondents agreed with the factors identified as the reasons for non-
implementation of TQM in the Ghanaian construction industry. It is notable that all the sixteen
ranked factors affecting the implementation of TQM in the construction industry have a MS >
3.50 ≤ 5.00, which indicates that the respondents perceive the factors affecting the
implementation of TQM in the construction industry to be between ‘agree’ and ‘strongly agree’.
All the rankings of factors affecting firms’ TQM implementation indicate MS> 3.79 - 4.33. The
relatively high MS achieved suggests that these variables are the major factors preventing
construction firms from implementing or adopting TQM. Among the sixteen factors identified,
lack of or limited knowledge of TQM was ranked first, followed by lack of enforcement from the
legislative bodies overseeing the implementation of TQM, lack of an efficient TQM management
system, lack of interest in the application of TQM, the absence of a TQM policy, lack of
commitment from management, reluctance to change old management techniques, lack of
coordination of the implementation of TQM policy within the organization, lack of training and
inability to train and educate employees on TQM, the complex nature of the TQM technique,
lack of TQM experts, lack of finance in the management of TQM experts, lack of understanding
among construction professionals in applying TQM, the perception that TQM may not yield any
better results, TQM technique is costly, and TQM technique is time consuming which was
ranked last as shown in Table 11.20.

On the other hand, EFA was conducted to assess the reliability of the identified factors that serve
as the reasons for non-implementation of TQM in the Ghanaian construction industry. The
reliability of the study's instrument was measured using Cronbach’s alpha. The Cronbach’s
420
alpha was found to be greater than 0.700 at 0.944. The lowest Cronbach’s alpha value was 0.938
while the highest value was 0.944, indicating acceptable internal reliability. The corrected item-
total correlation was also conducted and the finding revealed that the corrected item-total
correlation was greater than the suggested cut-off value of 0.30, suggesting that the items were
good reasons for non-implementation of TQM in Ghanaian construction industry. These findings
concur with the work of Whalen and Rahim (1994) who mentioned that some of the reasons for
non-implementation of TQM in construction industry are the lack of management commitment,
lack of appropriate training, the inability to change the organizational philosophy (culture), and
insufficiency of resources. Mosadeghrad (2014) supported this assertion by adding that
ineffective or inappropriate TQM models and ineffective or inappropriate methods for the
implementation of the TQM are the reasons for the non-implementation of TQM in the
construction industry.

12.3 MAJOR FINDINGS FROM DELPHI SURVEY AND QUESTIONNAIRE RESULTS

Findings from the Delphi study were the factors considered to be the principal determinants of
TQM in the Ghanaian construction industry. These include Leadership and Top Management,
Company Supplier Quality Management, Client Focus and Involvement, Company Quality
System Evaluation, Company Vision and Plan Statement, Product Selection and Design
Management, Construction Process Management and Improvement, and Construction
Employees’ Involvement and Motivation. The first objective of the Delphi study was to identify
the main and sub-attributes that determine TQM in the Ghanaian construction industry. In
determination of whether an attribute reached consensus or not by the experts, consensus was
reached for four (4) attributes (Leadership and Top Management, Company Quality System
Evaluation, Product Selection and Design Management, Construction Process Management and
Improvement) under the main attributes that determine TQM in the Ghanaian construction
industry, with an IQD score of 1.00. Six (6) of the attributes had strong consensus, with the
exception of Client Focus and Involvement, and Product Selection and Design Management,
which had good consensus. Seven (7) attributes from the eight (8) identified main attributes that
determine TQM in the construction industry were considered by the experts to have a very high
influence (VHI: 9-10), with the exception of one (1) attribute (Product Selection and Design
421
Management) which had a high influence (HI: 7-8.99). These indicate that all the attributes have
a high level of influence on TQM implementation in the Ghanaian construction industry.

Also, from the ninety-eight (98) sub-attributes identified under the (8) major determinants, only
thirty-six (36) were found to be the strong determinants of TQM in the construction industry in
Ghana. Among these sub-attributes which were determined by the experts to have reached strong
consensus are leadership style, leadership ability in solving quality problem, leadership
initiatives, leadership and top management knowledge in TQM, top management commitment in
TQM, top management interaction with workers, top management participation, top management
encouragement, supplier performance evaluation, supplier quality audit, supplier communication,
supplier knowledge of TQM, cooperation from suppliers, client complaint information/feedback,
client satisfaction survey, evaluation of overall company performance, evaluation of
departments’ performance evaluation of quality control system, evaluation of quality system
procedures, evaluation of end results, quality audit, quality benchmarking, and quality
information system. Other sub-attributes which were considered by the experts to have strong
determinants on total quality in the construction industry are quality improvement plan, client
brief/input, appearance/finishes required, strength required, intended purpose of the material,
design detailing, quality function deployment, value engineering, project monitoring and control
improvement, use of quality materials, understanding of work instructions/quality manual,
continuity of quality control circle activities, and education and training. The assessment of the
sub-attributes findings replicated the results of the majority of studies on TQM in the
construction industry in relation to adequate training and education, supplier quality management
and in general leadership and top management commitment (Saraph et al., 1989; Flynn et al.,
1994; Ahire et al., 1996; Imbeah & Dansoh framework, 2011; Adusa-Poku framework, 2014)
upon which the current study’s sub-attributes were also based.

The second objective of the Delphi study was to determine the factors that enable construction
companies to implement TQM in the construction industry. The assessment of the twenty-two
(22) factors of major determinants of TQM in the construction industry showed that sixteen (16)
factors were considered by the experts to have achieved strong consensus with an IQD score
422
between 0.00 and one (1). Eight (8) of the enabling factors (top management support, good
quality driven agenda, training and re-training of staff, institution of continuous improvement
process within the company, efficient management system, proper understanding of TQM among
construction professionals, good organizational and quality culture, and clarity of goals and
objectives) had a very high impact (VHI: 9-10) on TQM in the construction industry. The
remaining fourteen (14) of the enable factors had a high impact (HI: 7-8.99) on TQM in the
construction industry. These indicate that all the enabling factors have high levels of influence on
TQM implementation in the Ghanaian construction industry.

The third objective of the Delphi study was to identify the factors that affect construction
companies in Ghana that do not implement TQM. The assessment of the twenty (20) factors of
major determinants of TQM in the construction industry showed that nine (9) factors were
considered by the experts to have achieved strong consensus with an IQD less than or equal to
one (1). Eleven (11) of the factors (lack of efficient TQM management system, unavailable TQM
policy, lack or limited knowledge of TQM, lack of training and inability to train and educate
employees on TQM, lack of understanding among construction professionals in applying TQM,
lack of coordination of the implementation of TQM policy within the organisation, lack of TQM
expert, reluctance to change old management technique, lack of commitment from management,
lack of enforcement from the legislative bodies overseeing the implementation of TQM, and lack
of interest in the application of TQM) had a very high impact (VHI: 9-10) on TQM in the
construction industry. The remaining nine (9) factors had a high impact (HI: 7-8.99) on TQM in
the construction industry. These indicate that most of the factors have very high levels of
influence which affect construction companies in Ghana which do not implement TQM.

The fourth objective of the Delphi study was to evaluate the organizational factors that affect the
implementation of TQM in the construction industry. The assessment of the twenty (20)
organizational factors as considered by the experts to be the major determinants of TQM in the
construction industry showed that twelve (12) (leadership and top management commitment, top
management support, project manager competence, quality policy, supplier management, limited
cash flow to manage TQM, employee involvement, client involvement, feedback by project
423
participants, design quality management, strategic quality management, and quality data and
reporting) were considered by the experts to have achieved consensus with IQD less than or
equal to one (1). Six (6) of the organizational factors (leadership and top management
commitment, top management support, project manager competence, quality policy, education
and training, employee involvement) as considered by the experts had very high levels of
influence (VHI: 9-10) on TQM in the construction industry. The remaining fourteen (14) of the
organizational factors as considered by the experts had high levels of influence (HI: 7-8.99) on
TQM in the construction industry. These indicate how significant organizational factors and
issues are towards TQM implementation in the construction industry.

The fifth objective of the Delphi study was to identify the effects of implementation of TQM in
the construction industry. The assessment of the seventeen (17) major effects of the
implementation of TQM in the construction industry showed that fourteen (14) of the effects
were considered by the experts to have achieved consensus with an IQD less than or equal to one
(1). Eleven (11) of the effects (elimination/reduction of reworks, reduction in construction costs,
higher client satisfaction, improved employee job satisfaction, improved schedule performance,
defect-free product at first attempt, higher productivity, better control over the construction
process, increased profitability, gaining competitive advantage over other companies, decreasing
waste) had very high levels of influence (VHI: 9-10) on TQM in the construction industry, while
the remaining six (6) of the effects had high levels of influence (HI: 7-8.99). These indicate how
important the influence is that these effects have on TQM implementation in the construction
industry.

However, the Delphi study was validated by a field questionnaire survey. The results suggested
that the identified factors from the Delphi study have both a direct and indirect influence in
determining TQM. In the questionnaire survey the hypothesis that the exogenous factors had a
direct and positive influence on TQM could not be rejected. The exogenous variables of
Leadership/Top Management, Company Supplier Quality Management, Client Focus and
Involvement, Company Quality System Evaluation, Company Vision and Plan Statement,
Product Selection Management, Product Design Management, Construction Process
424
Management and Improvement, and Construction Employees’ Involvement and Motivation were
found to have a statistically significant influence in predicting TQM in the Ghanaian
construction industry. The findings from both the Delphi and the questionnaire survey therefore
suggested that the exogenous variables influenced the determination of the endogenous variable
(TQM). The merit of using SEM to validate the Delphi findings was that it was possible to
specifically ascertain which of the exogenous factors had a significant influence on TQM.
Therefore, instead of making a general statement that the exogenous variables had an influence
on determining the TQM, it was possible to state specifically that the factors Leadership/Top
Management features, Company Supplier Quality Management features, Client Focus and
Involvement features, Company Quality System Evaluation features, Company Vision and Plan
Statement features, Product Design Management features, Product Selection Management
features, and Construction Process Management and Improvement features had a direct or
stronger statistically significant influence on TQM in construction industry. Construction
Employees’ Involvement and Motivation features had an indirect (weak) influence in
determining TQM in the construction industry. However, the covariation with the other
exogenous construct to determine TQM in the construction industry was found to be statistically
significant. Therefore the weak variables have a good indirect association with the other latent
variables in the prediction of overall TQM in the construction industry.

12.4 CONCLUSION
In conclusion, the findings from the questionnaire survey generally supported the predictions that
were made by the experts from the Delphi study. The validated predictions were those of
leadership and top management, company supplier quality management, client focus and
involvement, company quality system evaluation, company vision and plan statement, product
selection management, product design management, construction process management and
improvement, and construction employees’ involvement and motivation. In addition, the existing
literature lends support to the findings of the current study. The supported findings were that
leadership and top management, company supplier quality management, client focus and
involvement, company quality system evaluation, company vision and plan statement, product
selection management, product design management, construction process management and
425
improvement, and construction employees’ involvement and motivation are fundamental to
TQM in the Ghanaian construction industry.

426
 CHAPTER THIRTEEN
13 CONCLUSION
13.1 INTRODUCTION
The general overall objective of the current study was to develop an integrated TQM model for
the Ghanaian construction industry. Ghana was used as a case study to identify the determinant
attributes which collectively predict TQM in the construction industry. The study adopted a
mixed-method approach to achieve the general objectives of the study. An extensive literature
review, a Delphi survey and a field questionnaire survey were carried out. The field
questionnaire survey was conducted in order to validate findings from the Delphi study with
regard to the factors which predict TQM. The final results were analyzed using SEM.
Conclusions regarding the study are presented relative to the objectives of the study in the next
sections.

13.1.1 Research Objective RO1

The first objective of the study was to establish the factors that determine TQM in the
construction industry, based on a literature review. In order to achieve this objective, a review of
literature was conducted. Findings are that TQM in the construction industry is a product of a
multi-faceted construct. Findings further reveal that TQM in the construction industry is about
incorporating fundamental management techniques, existing improvement efforts, and the
technical tools (US Department of Defense, 1988) under a disciplined approach focused on
continuous improvement. Kanji and Asher (1996) posited that TQM has a set of four principles
(delight the client, management by fact, people-based management, and continuous
improvement) that determine TQM in the construction industry.

Black and Porter (1996) recognized ten significant factors of TQM, namely client, people and
management, supplier partnership, teamwork structure for improvement, communication of
improvement information, client satisfaction orientation, external interface management,
strategic quality management, operational quality planning, quality improvement measurement
systems, and corporate quality culture. On the other hand, Ho and Fung (1994) recognized ten
TQM elements as commitment, leadership total customer satisfaction, continuous improvement,

427
total involvement, training and education, ownership, reward and recognition, error prevention,
and cooperation and teamwork. Research conducted by Powell (1995) also identified the
following components as a TQM framework, namely executive dedication, adopting the
philosophy, nearer to customers, nearer to suppliers, benchmarking, open organization, training,
employee empowerment, zero-defects mentality, flexible manufacturing, process improvement,
and measurement. On the other hand, Waldman (1994) recognized eight key TQM components
as top management commitment to place quality as a top priority, a broad definition of quality as
meeting customers’ expectations, TQM values and vision, the development of a quality culture,
involvement and empowerment of all organizational members in cooperative efforts to achieve
quality improvements, an orientation toward managing-by-fact, the dedication to continuously
improving employees’ capabilities and work processes through training and benchmarking, and
attempts to get external suppliers and customers involved in TQM efforts.

In fact, researchers have different views on the factors that determine TQM in the construction
industry. Nevertheless, most agree with the works of Saraph et al. (1989), Flynn et al. (1989),
Ahire et al. (1996) and Adusa-Poku (2014). The framework of Saraph et al. (1989) identified
factors that affect TQM in the construction industry as the role of divisional top management and
quality policy; role of quality department; training; product/service design; supplier quality
management; process management/operating; quality data and reporting; and employee relations.
The framework of Flynn et al. (1989) also identified factors that affect TQM in the construction
industry as quality leadership, quality improvement rewards, process control, feedback,
cleanliness and organization new product quality, interfunctional design process; selection for
teamwork potential, teamwork, supplier relationship, and customer involvement. The framework
of Ahire et al. (1996) on the other hand, identified factors that affect TQM in the construction
industry as top management commitment, customer focus, supplier quality management, design
quality management, benchmarking, SPC usage, internal quality information usage, employee
empowerment, employee involvement, employee training, product quality, and supplier
performance. Finally, the framework of Adusa-Poku (2014) identified factors that affect TQM in
the construction industry as process management, leadership commitment, customer focus,
continuous improvement, training and development, and supplier management. Findings from
428
the literature were that more research and effort are required to establish or develop an integrated
TQM that includes the most critical factors that determine TQM for the construction industry.

13.1.2 Research Objective RO2

The second objective of the research was to establish the theories and literature that have been
advanced on TQM in the construction industry and to identify the gaps that needed
consideration. A review of literature was carried out to achieve this objective. The extensive
review of TQM literature proposes that TQM includes an enormous variety of topics and
perspectives. Different writers have varying opinions on the concept of TQM. Nevertheless, the
study identified the theories of Deming, Crosby, Joseph Juran, and Ishikawa as theories that have
been advanced on TQM. The theoretical significance of the Deming strategy to TQM concerns
the creation of an organizational system that encourages learning and cooperation for facilitating
the execution of process management practices which, in turn, results in continuous enhancement
of procedures, products, and services as well as employee fulfilment, both of which are
necessary to client care and eventually to company success. Deming’s theory motivated a
thorough strategy to problem solving and promoted the commonly known Plan, Do, Check, Act
(PDCA) cycle. The PDCA cycle is also known as the Deming cycle. The cycle is about studying
and continuous enhancement, studying what works and what does not in a systematic way; and
the cycle repeats; after one cycle has been completed, another is started. Deming’s theory also
placed significance and liability on management, at the individual and company level, knowing
management to be liable for 94% of quality problems. Deming therefore insisted that
management’s responsibility is to build excellent systems that promote high quality
management. On the other hand, Crosby is well-known for the theory of “Quality is Free” and
“Zero Defects”. He emphasized avoidance rather than after-the-event examination, doing things
right the first time, and zero problems. Crosby’s concept relies on four absolutes of quality
management as follows: Quality is defined as to sticking requirements, prevention is the best
way to guarantee quality, zero defects (mistakes) is the performance standard for quality, and
quality is measured by the cost of non-conformity. Juran’s theory also is responsible for what has
become well-known as the "Quality Trilogy". The quality trilogy is made up of quality planning,
quality control and quality improvement. Ishikawa’s theory is focused on how companies should
429
handle their quality improvement tasks. This is the final theory which takes a look at quality
from a human viewpoint. One of the most well-known of these is the Ishikawa (or fishbone or
cause and effect) diagram. Like other tools, it assists groups in quality improvement. It lists the
major and minor causes leading to one effect (or problem), describes the issue, and recognizes
possible and potential causes by figuring out the possible ones. The diagram consistently
symbolizes and analyses the real causes behind a problem or effect.

This study adopted Deming’s theory of profound knowledge which is a management philosophy
grounded in systems theory in order to help to understand and develop a framework for TQM.
System theory is based on the principle that each organization is composed of a system of
interrelated processes and people which make up the system’s components. The three quality
award models identified also provide a framework for identifying a range of intangible and
tangible processes that influence the firm’s TQM implementation and the end results.

The findings revealed three constructs as gaps which have not been evaluated as all-inclusive
TQM constructs in the development of the previous models and theories. The identified gaps
from the extensive literature review were company vision and plan statement, product selection
and design management, and company quality system evaluation. The identified gaps formed the
new constructs in the current study’s conceptual framework (Model 1.0). The current study
offers a synthesized classification of the constructs which should be collectively considered in
order to predict TQM in the Ghanaian construction industry.

13.1.3 Research Objective RO3

The third research objective (RO3) of the study was to determine the main and sub-attributes of
TQM in the Ghanaian construction industry. A Delphi study was conducted in order to achieve
this objective. The findings indicated that a number of factors that were considered to be
important in determining TQM were identified and amplified by the Delphi study. The factors
considered as being paramount and main determinants of TQM were Leadership and Top
Management, Company Supplier Quality Management, Client Focus and Involvement, Company

430
Quality System Evaluation, Company Vision and Plan Statement, Product Selection and Design
Management, Construction Process Management and Improvement, and Construction
Employees’ Involvement and Motivation. In addition to the main attributes, thirty-six (36) sub-
attributes were found to be strong determinants of TQM in the construction industry in Ghana.
Among these sub-attributes which were determined by the experts to have reached strong
consensus are leadership style, leadership ability in solving quality problem, leadership
initiatives, leadership and top management knowledge in TQM, top management commitment in
TQM, top management interaction with workers, top management participation, top management
encouragement, supplier performance evaluation, supplier quality audit, supplier communication,
supplier knowledge of TQM, cooperation from suppliers, client complaint information/feedback,
client satisfaction survey, evaluation of overall company performance, evaluation of
departments’ performance, evaluation of quality control system, evaluation of quality system
procedures, evaluation of end results, quality audit, quality benchmarking, and quality
information system.

Others sub-attributes which were considered by the experts to have strong determinants on total
quality in the construction industry are quality improvement plan, client brief/input,
appearance/finishes required, strength required, intended purpose of the material, design
detailing, quality function deployment, value engineering, project monitoring and control
improvement, use of quality materials, understanding of work instructions/quality manual,
continuity of quality control circle activities, and education and training.

The findings suggested that the attributes that bring about TQM in the Ghanaian construction
industry are similar to the determinants in other cultural contexts. Further, TQM is assured if
there is a consideration of these factors in the development of construction firms in Ghana. These
factors were collectively considered for the development of the all-inclusive (integrated holistic)
TQM model for construction industry.

431
13.1.4 Research Objective RO4
The fourth research objective (RO4) was to evaluate the critical factors that affect construction
firms for not implementing TQM in Ghana. A Delphi study was conducted in order to achieve
this objective. Findings were that a number of factors that were considered to be important in
evaluating the critical factors and issues that affect the construction firms for not implementing
TQM were identified and amplified by the Delphi study. Findings further reveal that the critical
factors which affect TQM implementation in Ghana, and which are equally applicable to other
developing countries is the lack or limited knowledge of TQM. Also, it is revealed that there is a
lack of enforcement from the legislative bodies overseeing the implementation of TQM.
Moreover, it was found that there is a lack of an efficient TQM management system. The
findings also reveal a lack of interest in the application of TQM. Other critical factors identified
in the study are the absence of a TQM policy, a lack of commitment from management,
reluctance to change old management techniques, a lack of coordination of the implementation
of a TQM policy within the organization, a lack of training and an inability to train and educate
employees on TQM, the complex nature of the TQM technique, a lack of a TQM expert, a lack
of finance in the management of TQM experts, a lack of understanding among construction
professionals in applying TQM, the perception that TQM may not yield any better results, TQM
technique is costly, and that the TQM technique is time consuming. It is notable that all the
sixteen identified factors which influencing firms regarding the non-implementation of TQM in
the Ghanaian construction industry were ranked high. The relatively high MS achieved suggests
that these variables are the reason why construction companies in Ghana do not implement or
adopt TQM.

13.1.5 Research Objective RO5

The fifth research objective of the study was to develop an integrated TQM model for the
Ghanaian construction industry based on a literature review, a Delphi study and the
questionnaire survey. A synthesis of the reviewed literature together with the findings from the
Delphi study was used to achieve this objective. The conceptual model theorized that TQM in
the Ghanaian construction industry is an eight-factor construct. These factors were Leadership
and Top Management, Company Supplier Quality Management, Client Focus and Involvement,
432
Company Quality System Evaluation, Company Vision and Plan Statement, Product Selection
and Design Management, Construction Process Management and Improvement, and
Construction Employees’ Involvement and Motivation. This hypothesis factor model was
validated by conducting a questionnaire survey and analyzing it using SEM software, EQS
Version 6.2. The theorised eight-factor model was taken through EFA. A nine-factor model was
realized during the EFA. The latent construct with the name ‘Product Selection and Design
Management (PSDM)’ had two factor loadings during EFA, hence, was separated into two with
the assigned names ‘Product Selection Management (PSM)’, and ‘Product Design Management
(PDM)’. The final nine-factors for the model are Leadership and Top Management, Company
Supplier Quality Management, Client Focus and Involvement, Company Quality System
Evaluation, Company Vision and Plan Statement, Product Design Management, Product
Selection Management, Construction Process Management and Improvement, and Construction
Employees’ Involvement and Motivation, which jointly predict TQM in the Ghanaian
construction industry.

13.1.6 Research Objective RO6

The sixth and final research objective of the study was to test and validate the conceptually
integrated TQM model for the Ghanaian construction industry. A questionnaire survey and
analysis of the results using SEM software, EQS Version 6.2 were conducted in order to
achieve this objective. Findings from the SEM analysis, which models TQM for the
Ghanaian construction industry as a nine-factor model showed that the factors of Leadership
and Top Management, Company Supplier Quality Management, Client Focus and
Involvement, Company Quality System Evaluation, Company Vision and Plan Statement,
Product Design Management, Product Selection Management, Construction Process
Management and Improvement, and Construction Employees’ Involvement and Motivation
were found to have a significant influence in determining the TQM for the Ghanaian
construction industry. However, the Construction Employees’ Involvement and Motivation
feature had a weak (indirect) influence in predicting TQM. Nevertheless, its covariances
with the other exogenous construct to determine TQM for the Ghanaian construction
433
industry was found to be statistically significant. These findings validated the conceptually
integrated holistic model developed from literature and the Delphi study. The nine-factor
model therefore was deemed fit for the current study.

13.2 CONTRIBUTION AND VALUE OF THE RESEARCH

The value and contribution of the current research are described at three levels, namely the
theoretical, methodological and practical levels of the research findings. However, it is pertinent
to note that the outstanding contribution of the study is the revelation of the influence of top
management in the construction industry in predicting TQM in the Ghanaian construction
industry.

13.2.1 Theoretical Contribution

The results of the SEM analysis indicated that an integrated TQM model is a nine-factor model.
This was achieved through the following processes:
i. The compilation of historical documentation on TQM practices, policy and standards in
developed and developing countries;
ii. The use of a mixed-method approach made up of a Delphi study and a questionnaire
survey. The Delphi survey was used to arrive at a theoretical model. Furthermore, a
questionnaire survey was employed to ascertain the variables in the theoretical model.
The SEM software EQS version 6.2 was used to obtain the required variables on the
TQM model;
iii. Finally, an integrated TQM model for construction firms was developed.
Findings from the literature review did not reveal evidence of research conducted on the current
study and this therefore shows that the study on TQM will assist in improving the quality of
work undertaken by construction firms. Moreover, there was no evidence that suggested that a
mixed method using Delphi and SEM had been used by researchers in this field in Ghana.
Therefore, this study may offer a base for other researchers to use for other follow-up studies.
Likewise, the current study modelled TQM as a nine-factor construct. Apart from the study

434
contributing to theoretical knowledge, it also contributes to methodological advance in terms of
the approach used in conducting the research as indicated in the next section.

13.2.2 Methodological Contribution

Most studies have used univariate statistical methods such as ANOVA, MANOVA or regression
modelling to model TQM. However, the current study used SEM, which is more robust and
superior to the methods mentioned in determining causality of factors in a model and their
direction of influence (Kline, 2005; Musonda, 2012:252, Aigbavboa, 2013:248). SEM is most
commonly thought of as a hybrid between some form of analysis of variance (ANOVA) or
regression and some form of factor analysis. In general, it can be noted that SEM allows one to
perform some type of multilevel regression/ANOVA on factors. With SEM analysis, it was
possible to identify the factors of TQM which have a significant effect and hence influence
construction firms in the construction industry as opposed to a general blanket statement that
there are numerous constructs which influence TQM. The questionnaire survey instrument had
high internal reliability values and therefore could be used in similar studies to validate the
current study or for similar purposes. Findings from the Delphi study and the conceptual model
developed from both the literature review and the Delphi study were validated by conducting a
questionnaire survey. Data from the questionnaire survey was analysed using SEM software,
EQS Version 6.2. As a result of this mixed method, a parsimonious model was developed. Aside
from this contribution and value to the body of knowledge in terms of the methodological
approach, a contribution to practice and the housing industry was also achieved.

13.2.3 Practical Contribution and Value

SEM results indicated that leadership or top management in construction firms had an influence
on TQM in the construction industry. Moreover, the knowledge of the influence of the nine-
factor construct could help the leadership or top management to plan, organize, coordinate and
control all aspects relating to the TQM issues. The practical significance of the study is further
elaborated on as follows:

435
Significance to Construction Firms
The output of the study will help construction firms in making decisions about the criteria to be
given priority in TQM issues. The findings will also help the leadership or top management to
plan programmes for their firms, as well as planning effective TQM policies. This will ultimately
enable the construction firms to know how to commit resources for an effective TQM
implementation. It should be noted that the first building block of the successful TQM model is
to understand and adopt its principles. Hence, the Association of Building and Civil Engineering
Contractors of Ghana (ABCECG) should monitor TQM practices in the Ghanaian construction
industry in order to ensure compliance with the laid down principles of TQM. The integrated
holistic TQM model should be used as a guide to ensure that all construction firms abide by it.
The study offers an opportunity for further research to improve the model developed in this study
and probably refine indicator variables to suit specific environments. Therefore the
recommendations and policy implications for the practice of all these areas in which the current
study may add value and contribute are presented below.

13.3 RECOMMENDATIONS
Recommendations are made from the methodological, theoretical and practical points of view.
13.3.1 Methodological
It is recommended that a similar study should be conducted by soliciting views from both top
management and junior staff (all employees) in construction industry. The inclusion of both top
management and junior staff (all employees) will enable researcher/researchers to have general
and solid views from them with regard to TQM practices in Ghana. Also, further research should
be conducted on the indicator variables to establish any improvement in model fit, as the current
study was purely a factor analysis. There is the possibility that TQM could be defined by more
indicator variables. Recognition should be given, however, to the fact that there is no such a
thing as a perfect model. Moreover, there should be a move to try to improve on the current
model rather than invent a new model. The recommended method could commence with a
Delphi study followed by a questionnaire survey or vice versa in order to improve its
generalisability. Most of the studies in Social Science and most especially TQM studies used
standard statistical procedures such as ANOVA or MANOVA and multiple regressions which do
436
not offer an appropriate and a straightforward way to test a hypothesis at a higher level of
abstraction. Therefore, for similar studies as the current one, SEM with EQS is recommended to
be used as the analysis technique for better results and abstraction.

13.3.2 Theoretical
It was observed from the literature that there were still different definitions and varying
understanding of how TQM is formed in construction industry. This has led in the past to a
limited view and narrow conceptualization of TQM in the Ghanaian construction industry.
Besides, there has not been consensus on how TQM in the Ghanaian construction industry
should be measured. However, in the current study, literature was reviewed and synthesized on
the determinants of TQM. In conjunction with the experts’ knowledge obtained through the
Delphi study, an eight-factor TQM model was arrived at for construction firms in Ghana. These
factors were identified as Leadership and Top Management, Company Supplier Quality
Management, Client Focus and Involvement, Company Quality System Evaluation, Company
Vision and Plan Statement, Product Selection and Design Management, Construction Process
Management and Improvement, and Construction Employees’ Involvement and Motivation. The
latent construct with the name ‘Product Selection and Design Management (PSDM)’ had two
factor loadings during EFA, and was separated into two with the following assigned names:
‘Product Selection Management (PSM)’, and ‘Product Design Management (PDM)’. ‘Product
Selection and Design Management (PSDM)’ was substituted with the two new latent constructs
(‘Product Selection Management’ and ‘Product Design Management’), and the number of the
latent constructs of the original conceptual TQM model in Figure 10.1 (Model 1.0) increased
from eight to nine, hence a nine-factor model was realized. The nine-factor model was therefore
deemed fit for the current study. The final nine factors for the model are Leadership and Top
Management, Company Supplier Quality Management, Client Focus and Involvement, Company
Quality System Evaluation, Company Vision and Plan Statement, Product Design Management,
Product Selection Management, Construction Process Management and Improvement, and
Construction Employees’ Involvement and Motivation. It is therefore recommended that the
developed model and theory of TQM with particular emphasis on operationalization should form

437
the basis for further refinement of the concept and thereby make it useful to the construction
firms’ TQM in Ghana and other developing countries.

13.3.3 Practical and Policy Implication

As a result of the identified contributions that the current study makes as revealed by the
findings, the following policy implications and practical recommendations have been identified:
i. The policy implication suggests that TQM can be improved through the construction firms’
adherence to the core principles of the TQM.
ii. Stakeholders responsible for project quality in Ghana and other developing countries should
adopt effective management strategies to encourage all construction firms to implement TQM.
iii. Future TQM in the Ghanaian construction industry should contain the nine factors.
iv. Planning, organising, monitoring, measuring and controlling of TQM would be feasible if the
stakeholders are aware of indicator variables that define TQM.

13.4 LIMITATIONS
Interesting and valuable findings have emerged from this study. However, the following
limitations regarding the current study should be considered. Firstly, the research was only
conducted among top management in the Ghanaian construction industry. Given enough
resources, it would be preferable to conduct a similar research study among both top
management and junior staff (all employees) in the Ghanaian construction industry. Also, the
consideration of other developing countries could be included. Secondly, the use of additional
items or constructs might improve the inherent reliability and validity of the measures used.
Thirdly, several nested models, especially the measurement models, could have been evaluated
to check the suitability of other alterative models. The current study was purely exploratory and
confirmatory in nature. Fourthly, although the internal reliability tests indicated high internal
consistency and therefore a well-constructed research tool, some constructs revealed high
correlational values. This may be due to the fact that only one questionnaire was used to collect
information among the firms under study. A review of the research tool would have benefited
findings in this study. A final limitation is related to the sample: in addition to the
aforementioned limitations, the study has shown that some of the SEM measures may have been
438
influenced by the sample size of the study. All empirical studies are limited by the nature of the
sample studied. The exploration of the dependent variable (TQM) has shown that it has a very
complex organization (multi-faceted) and claims for further interpretations.

13.5 RECOMMENDATIONS FOR FURTHER RESEARCH

The following suggestions for further studies have been identified:
i. Further studies should be conducted to examine factors related to the limitations of the current
study. More rigorous and detailed testing of measurement scales among both top management
and junior staff (all employees) in the Ghanaian construction industry and other developing
countries would increase the knowledge of TQM in the construction industry.
ii. Some features of TQM variables may be improved, while some might remain in common.
Equally important are those groups of individual characteristics which were not considered in the
conceptualization of the current model. Therefore, a validation of the Ansah Integrated Holistic
TQM Model for the Ghanaian construction industry presented in Figure 11.13 (Model 2.0) is
recommended.
iii. The SEM software with EQS version 6.2 methodology used in data analysis should be
encouraged.

13.6 CONCLUSION
An integrated TQM model for the Ghanaian construction industry was developed based on the
existing quality management issues, practices and grounded theories. It was postulated that
overall TQM is directly related to the influence of the exogenous (latent) variables in predicting
or determining overall TQM for the Ghanaian construction industry. The postulated model was
analysed with the use of SEM software EQS Version 6.2. The fit statistics for the measurement
and structural models had an adequate fit to the sample data. The finalized empirical model
revealed that the exogenous variables (leadership and top management, company supplier quality
management, client focus and involvement, company quality system evaluation, company vision
and plan statement, product design management, product selection management, construction
process management and improvement, and construction employees’ involvement and
motivation) had a statistically significant influence in determining TQM for the Ghanaian
439
construction industry. Specifically, the exogenous variables such as construction employees’
involvement and motivation had a weak (indirect) influence on determining TQM for the
Ghanaian construction industry. From the findings, it is therefore concluded that the nine-factor
model schematically portrayed in Figures 11.12 and 11.13 (Model 2.0) represents an adequate
description of TQM for the Ghanaian construction industry.

The results of this study have theoretical, methodological and policy (practical) value because
respondents for the Delphi study were drawn from academics and construction professionals.
The respondents for the questionnaire survey were among the top management of construction
firms. Furthermore, the respondents had a good working knowledge of the studied environment.
In addition, the questionnaire survey results which modelled using SEM was a validating study
of a conceptual model developed from synthesized theories established from literature and more
importantly from the Delphi study. Hence, it is considered that the presented model for TQM
interpretation maintains its validity. The result of the study provided information that can inform
governmental, the Association of Building and Civil Engineering Contractors of Ghana and
individual firms as they plan for and implement TQM. Secondly, the study provides indicators
that will be a baseline for implementing TQM. Also, the conceptual model of TQM for the
Ghanaian construction industry which has been formulated in this study will provide a reference
to researchers who will study TQM in the near future. The current study lends support to other
studies that have utilised alternative methods to establish the factors which influence TQM for
the Ghanaian construction industry. These studies have concluded that TQM for the Ghanaian
construction industry is multifaceted as is also claimed in the current study.

The current study utilizes the more robust modelling method of SEM. By adopting this
methodology, the current study was able to model the influence of the selected multi-faceted
variables and the constructs which were statistically significant. The practical implication is that
the TQM for the Ghanaian construction industry model can be enhanced by improving on the
factors of leadership and top management, company supplier quality management, client focus
and involvement, company quality system evaluation, company vision and plan statement,
product design management, product selection management, construction process management
440
and improvement, and construction employees’ involvement and motivation to enable their
application by the construction firms. Moreover, the Association of Building and Civil
Engineering Contractors of Ghana and other stakeholders responsible for construction projects in
Ghana can adopt effective management strategies to improve on the TQM in the Ghanaian
construction industry. In future, all firms managing construction projects should adopt the nine-
factor model to enhance the quality of the implementation of TQM in the construction industry.

441
 REFERENCES
Aaker, A., Kumar, V. and George, S. (2009). Marketing research. (10th edn.). New York:
John Wiley and Sons.
Abdul-Rahman, H. (1995). The cost of non-conformance during a highway project: A case
study. Construction Management and Economics, 13:23-32.
Abiodun, I.O. and Afangadem, A.U. (2007). An assessment of the effectiveness of regulatory
measures for ensuring quality and standards in the construction industry of some selected
cities in Nigeria. Journal of the Nigerian Institute of Building, pp. 63 – 70.
Addison, T. (2003). E-commerce project development risks: Evidence from a Delphi survey.
International Journal of Information Management, 23(1):25-40.
Adeogun, B.K. and Okafor, C.C. (2013). Occupational Health, Safety and Environment (HSE)
Trends in Nigeria. International Journal of Environmental Sciences, Management and
Engineering Research, 2(1): 24-29, Jan-Feb. Available from http:// www.ijesmer.com
[Accessed 23 October 2015].
Adetunji, I. O. (2005). Sustainable construction: A web-based performance assessment tool.
Ph.D. Dissertation. Department of Civil and Building Engineering, Loughborough
University, UK.
Adler, M. and Ziglio, E. (1996). Gazing into the oracle: The Delphi method and its
application to social policy and public health. London: Kingsley Publishers.
Adusa-Poku, N. Y. (2014). Assessing total quality management (TQM) in the Ghanaian
construction industry: An exploratory study in Kumasi. Master of Science Dissertation.
Department of Building Technology, Faculty of Architecture and Building Technology,
Kwame Nkrumah University of Science and Technology, Ghana.
African Development Fund. (ADF) (2008). Ghana country gender profile. Human Development
Department (OSHD). Available from: http:// www.afdb.org [Accessed 30 August 2016].

African National Congress (1994). Reconstruction and Development Programme White Paper.
Available from http://www.anc.org.za/ancdocs/policy/white.html. [Accessed 18 November
2015].

442
Agha, A. S. (2011) TQM in construction industry. Available from
http://faculty.kfupm.edu.sa/CEM/bushait/cem515/term-
papers/TQMinconstructionindustry.pdf. [Accessed 12 March 2015].
Ahire, S.L., Golhar, D.Y. and Waller, M.A. (1996). Development and validation of TQM
implementation constructs. Decision Sciences, 27(1):23-56.
Ahmed, S. (1993). An integrated total quality management (TQM) model for the construction
process. Ph.D. Dissertation, School of Civil & Environmental Engineering, Georgia
Institute of Technology Atlanta, GA, USA.
Ahmed, SM., Azhar, S. and Castillo, M. (2002). Measurement of construction processes for
continuous improvement. Final Report submitted to the Building Construction Industry
Advisory Council, Department of Education (DOE), State of Florida
Aigbavboa, C. O. (2013). An integrated beneficiary-centred housing satisfactory model for
publicly funded housing schemes in South Africa. Unpublished DPhil in Engineering
Management, University of Johannesburg. South Africa. Available from
http:www.ujdigispace.uj.ac.za [Accessed 15 June 2016].
Allens, A.R. (1994). Quality management in the construction phase of the traditional
procurement system in South Africa: The case of the Western Cape. University of Cape
Town in Cape Town, Western Cape, South Africa.
Al-Momani, A. (2007). TQM in construction: A case study of Jordan. In: Hughes, W. (Ed).
Proceedingss of the Inaugural Construction Management and Economics “Past, Present
and Future” Conference, 16th -18th July 2007, University of Reading, UK, pp. 741-752
Al-Musleh, A. M. (2010). Development of A Framework for Total Quality Management
Principles in the Construction Companies with Special Reference to the Construction
Companies in the State of Qatar. PhD Thesis. School of Engineering and Material
Sciences at the Queen Mary, University of London, London, United Kingdom.

Amara, R. (1975). Some methods of futures research. Menlo Park: Institute for the Future.
Amaratunga, D., Baldry, D., Sarshar, M. and Newton, R. (2002). Quantitative qualitative
research in the environment. Work Study, 51(1):17-31.

443
Ambrose, G.A.L. and Claasen, S.J. (2004). A total quality management approach to civil
engineering consulting. SA Journal of Industrial Engineering, 15(1):67-81.

American Quality Foundation and Ernst & Young. (1991), International quality study: The
definitive study of the best international quality management practices. Cleveland, OH.:
Ernst & Young.
Anderson, J.C., Rungtusanatham, M. and Schroeder, R.G. (1994). A theory of quality
management underlying the Deming management method. Academy of Management
Review, 19(3):472-509.
Ankrah, N. I. (2007). An investigation into the impact of culture on construction project
performance. PhD Thesis. SEBE, University of Wolverhampton, Wolverhampton.
Ansah, S.K. (2011). Partnering: An alternative contractual arrangement for construction project
delivery in Ghana. Journal of Polytechnics in Ghana (JOPOG), 5(1):32 – 48.
Ansah, S.K., Aigbavboa, C. and Thwala, W.D. (2015). Total quality management
implementation in the Ghanaian construction industry: A tool for improving construction
productivity. In: Mojekwu, J.N., Nani G., Atepor, L., Thwala, W. D., Ogunsumi, L., Awere
E., Ocran, S.P. and Bamfo-Agyei, E. (Eds). Proceedings of the 4th Applied Research
Conference in Africa. (ARCA) Conference, 27-29 August 2015, Ibadan, Nigeria, pp. 336-
347.
Anvuur, A., Kumaraswamy, M. M. and Male, S. (2006). Taking forward public procurement
reforms in Ghana. CIB W107 Construction in Developing Countries. In: Proceedings of
the International Symposium, Construction in Developing Economies: New Issues and
Challenges, 18th - 20th January 2006, Santiago: Chile.
Arditi, D. and Gunaydin, H. M. (1997). TQM in the construction process. International Journal
of Project Management, 15(14):235-243.
Aubrey, C. (2005). Quality management in financial services. Glencoe: McGraw Hill.
Babbie, E. R. (1990). Survey research methods. Belmont, California: Wadsworth.
Bagozzi, R.P. and Yi, Y. (2012). Specification, evaluation, and interpretation of structural
equation models. Journal of the Academy of Marketing Science, 40(1):8-34.
Bamisile, A. (2004). Production management. Lagos, Nigeria: Foresight Press.

444
Barrett, P. (2007). Structural equation modeling: Adjudging model fit. Personality and
Individual Differences, 42(5):815-824.
Beales, R. (2005). Delphi creates fresh CDS challenge: A new cash-settlement method for
simplifying credit derivatives faces a stern test with the latest bankruptcy. The fFnancial
Times, 43.
Bell, S. (1996). Learning with information systems: Learning cycles in information systems
development. New York: Routledge.
Black, S.A. and Porter, L.J. (1996), Identification of the critical factors of TQM. Decision
Sciences, 27(1):1-21.
Boaden, R. and Dale, B. (1992). TQM in the construction industry: A preliminary analysis.
International Journal of Technology Management, 7(4/5): 244-253.
Boote, D.D. and Beile, P. (2005). Scholars before researchers: On the centrality of the
dissertation literature review in research preparation. Educational Researcher, 34(5):3-15.
Boshoff, S. (2010). The Cape Town 2030 initiative: A contested vision for the future. In:
Counter-currents: Experiments in sustainability in the Cape Town region. Cape Town:
Jacana Media and African Centre for Cities, pp.70–79.
Bowen, P., Pearl, R. and Akintoye, A. (2007). Professional ethics in the South African
construction industry. Building Research & Information, 35(2):189-205.
Brace, I. (2008). Questionnaire design: How to plan, structure and write survey material for
effective market research. Philadelphia, United States: Kogan Page Publishers.
Brieschke, P.A. (1992). Reparative praxis: Rethinking the catastrophe that is social science.
Theory into Practice, 31(2): 173-180.
Brill, J., Bishop, M. & Walker, A. (2006). The competencies and characteristics required of an
effective project manager: A web-based Delphi study. Educational Technology Research
and Development, 54(2):115-40.
British Standards Institution (BSI). (1992). BS 7850 Total quality management. London: British
Standards Publishing.
British Standards Institution (BSI). (2009). History of BSI group. Available from
http://www.bsigroup.com/en/About-BSI/About-BSI-Group/BSI-History/ [Accessed 17
September 2016].
445
Bryman, A. (2001). Social research methods. Oxford, UK.: Oxford University Press.

Bryman, A. (2007). Barriers to integrating quantitative and qualitative research. Journal of

Mixed Methods Research, 1(1):8-22.
Buckley, C. (1995). Delphi: A methodology for preferences more than predictions. Library
Management, 16(7):16-19.
Buckley, C.C. (1994). Delphi technique supplies the classic result? Australian Library Journal,
43(3):158-64.
Buertey, J. I. T., Kaku, M. A. and Adjei-Kumi, T. (2013). Delays to large construction projects
in Ghana: A risk overview. In: Laryea, S. and Agyepong, S. (Eds.). Procedings of 5th West
Africa Built Environment Research (WABER) Conference, 12–14th August 2013, Accra,
Ghana, pp.367 – 380.
Buertey, J., Amoa, P. and Adjei-Kumi, T. (2011). Construction cash flow prediction model in
Ghana: A case study of the District Assembly Common Funded Projects. Pentvars
Business Journal, 5(2).
Burati, J. L. and Kalidindi, S.N. (1991). Quality management in the construction industry. ASCE
Journal of Construction, Engineering and Management, 117(2):341-359.
Burke, R. (2007). Project management techniques. Auckland, New Zealand: Burke Publishing.
Burns, A.C. and Bush, R.F. (2002). Marketing research: Online research applications. New
Jersey: Prentice Hall.
Burrell, G. and Morgan, G. (1994). Sociological paradigms and organizational analysis:
Elements of the sociology of corporate life (6th edn.). London: Heinemann, pp.1-37.
Byrne, B.M. (2010). Structural equation modelling with AMOS. Basic concepts, applications
and programming. New York/London: Routledge.
Cain, C.T. (2004). Performance measurement for construction profitability. Oxford: Blackwell.
Cavalli-Sforza, V. and Ortolano, L. (1984). Delphi forecasts of land use – transportation
interactions. Journal of Transportation Engineering, 110(3):324-39.
Cherrington, D.J. (1995). The management of human resources (4th edn.). Englewood Cliff, NJ.:
Prentice- Hall.
Chini, A.R. and Valdez, H.E. (2003). ISO 9000 and the U.S construction industry. Journal of
Management in Engineering, 16(2): 69-76.
446
Choi, T.Y. and Eboch, K. (1998). The TQM paradox: Relations among TQM practices, plant
performance, and customer satisfaction. Journal of Operations Management, 17(1): 59-75.
Chou, C.P. and Bentler, P.M. (1995). Estimates and tests in structural equation modelling. In:
Hoyle, R.H. (Ed.). Structural equation modeling: Concepts, issues, and applications.,
Thousand Oaks, CA.: Sage Publications.
Choudhry, R.M., Fang, D. and Mohamed, S. (2007). The nature of safety culture: A survey of
the state-of-the-art. Safety Science, 45(10):993-1012.
Churchill, G.A. (2001). Basic marketing research. Fort Worth: The Dryden Press.
Coates, J.F. (1975). Review of Sackman Report - Technological forecasting and social change,
7(2). New York: American Elsevier Publishing.
Comrey, A.L. and Lee, H.B. (1992). A first course in factor analysis. New Jersey: Lawrence
Erlbaum Associates.

Conca, F.J., Llopis, J. and Tari, J. J. (2003). Development of a measure to assess quality
management in certified firms. Journal of Operational Reasearch, 156(2004): 683-697.
Construction Industry Development Board (CIDB) (2004). SA Construction Industry Status
Report - Synthesis review on the South African construction industry and its development.
Pretoria: CIDB.
Construction Industry Development Board (CIDB) (2010) The CIDB construction industry
indicators summary results: 2009. Pretoria, South Africa: CIDB.
Construction Industry Development Board (CIDB) (2012). Construction industry indicators,
March 2012. Pretoria, South Africa: CIDB.
Construction Industry Development Board (CIDB). (2007). The National Infrastructure
Maintenance Strategy in support of ASGISA and Government growth objectives. Pretoria,
South Africa: CIDB.
Cooper, D. and Schindler, P. (2006). Marketing research. New York: McGraw-Hill.
Cooper, D.R. and Emory, W. (1995). Business research methods. Chicago: US: Irwin.
Cordon, C. P. (2013). System theories: An overview of various system theories and its
application in healthcare. American Journal of Systems Science, 2(1):13-22 DOI:
10.5923/j.ajss.20130201.03.

447
Creative Research Systems (2003). The survey system. Available from
http://www.surveysystem.com/sscalc.htm [Accessed 21October 2015].
Creswell, J.W. (1994). Research design: Qualitative, quantitative, and mixed methods
approaches. Thousand Oaks: Sage.
Creswell, J.W. (2003). Research design: Qualitative, quantitative, and mixed method
approaches. Thousand Oaks, California: Sage Publications.
Creswell, J.W. and Clark, V.L. (2007). Designing and conducting mixed methods research.
Thousand Oaks: CA: Sage Publications.
Creswell, J.W., Clark, V.L., Gutmann, M.L. and Hanson, W.E. (2003). Advanced mixed
methods research designs. In: Tashakkori, A. and Teddlie, C. (Eds.). Handbook of mixed
methods in social and behavioral research, Thousand Oaks, CA.: Sage Publications. pp.
209-240.
Crisp, J., Pelletier, D., Duffield, C., Adams, A. and Nagy, S. (1997). The Delphi method?
Nursing Research, 46:116-8.
Critcher, C. and Gladstone, B. (1998). Utilizing the Delphi technique in policy discussion: A
case study of a privatized utility in Britain. Public Administration, 76(3):431-49.
Crosby, P.B. (1979), Quality is free. New York: McGraw-Hill.
Crotty, M. (1998). The foundations of social research: Meaning and perspective in the research
process. London: Sage Publications.
Cuhls, K. (2003). Delphi method. Available from:
http://www.unido.org/fileadmin/import/16959_DelphiMethod.pdf. [Accessed 26 October
2016].
Culp, G., Smith, A. and Abbott, J. (1993). Implementation TQM in consulting engineering firm.
Journal of Management in Engineering, 9(4):340-366.
Czaja, R. and Blair, J. (1996). Designing surveys: A guide to decisions and procedures.
Thousand Oaks, CA., London: Pine Forge Press.
Czinkota, M.R. and Ronkainen, I.A. (1992). Global marketing 2000: A marketing survival
guide. Marketing Management, January/February, pp.37-44.

Dale, B.G. and Plunkett, J.J. (1990). Managing quality. New York: Philip Allan.

448
Dalkey, N.C. & Helmer, O. (1963). An experimental application of the Delphi method to the use
of experts. Journal of the Institute of Management Sciences, 9:458-467.
Dean, J.W. Jr. and Bowen, D.E. (1994). Management theory and total quality: Improving
research and practice through theory development. Academy of Management Review,
19(3):392-418.
Delbecq, A.L., Van de Ven, A.H. and Gustafson, D.H. (1975). Group techniques for
program planning: A guide to nominal group and Delphi processes. Glenview, IL:
Scott, Foresman & Company.
Deming, W. E. (1986). Out of the crisis: Quality, productivity and competitive position.
Cambridge: Cambridge University Press.
Denzin, N. K. (1970). The research act in sociology. London: Butterworth.
Denzin, N. K., and Lincoln, Y. S. (2003). Strategies of qualitative inquiry. Thousand Oaks, CA:
Sage Publications.
Dikmen, I. and Birgonul, M. T. (2003). Strategic perspective of Turkish construction companies.
Journal of Management in Engineering, 19(1):33-40.
Dingsdag, D.P., Biggs, H.C., Sheahan, V.L. and Cipolla, C.J. (2006). A construction safety
competency framework: Improving OH&S performance by creating and maintaining a
safety culture. Brisbane: Cooperative Research Centre for Construction
Innovation.Disaster Management Institute (DMI) nd. Accident Causation Theories
Dion, P.A. (2008). Interpreting structural equation modelling results: A reply to Martin and
Cullen. Journal of Business Ethics, 83:365-368.

Dow, D., Samson, D. and Ford, S. (1999). Exploding the myth: Do all quality management
practices contribute to superior quality performance. Production and Operations
Management,8:1–27.
Du Plessis, C., Irurah, D.K. and Scholes, R.J. (2003). The built environment and climate change
in South Africa. Building Research and Information, 31(3-4):240-256.
Dublin, R. (1978). Theory of building. A practical guide to the construction and testing of
theoretical models (Rev. edn.) New York: The Free Press.

449
DuBrin, A.J. (1995). Leadership: Research findings, practice, and skills. Boston, MA:
Houghton Mifflin Company.
Duncan, J.M., Thorpe, B. and Sunmer, P. (1990). Quality assurance in construction. Aldershot,
Great Britain: Gower Publishing Company.
Dunn, W.N. (1994). Public policy analysis: An introduction. Englewood Cliffs, NJ: Prentice
Hall.
Easton, G. (1993). The 1993 state of U.S. TQM: A Baldrige examiner’s perspective. California
Management Review, 35(3):32-54.
Easton, G.S. and Jarrell, S.L. (1998). The effects of TQM on corporate Economist. Construction
and the Internet. The Economist. 15 January 2000.
Egan, S. J. (1998). Rethinking construction. Report of the Construction Task Force, HMSO.
London: Department of the Environment.
Ellram, L. (1996). The use of the case study method in logistics research. Journal of Business
Logistics, 17(8):93-138.
Els, D.A. and Delarey, R.P. (2006). Developing a holistic wellness model. South African
Journal of Human Resource Management, 4(2):46-56.
Eto, H. (2003). The suitability of technology forecasting/foresight methods for decision systems
and strategy. A Japanese view. Technological Forecasting and Social Change, 70:231-
249.
Feigenbaum, A.V. (1991). Total quality control (3rdedn.). New York: McGraw-Hill. Fellows, R.
and Liu, A. (1997). Research methods for construction. Oxford: Blackwell Science.

Fernandez-Muniz, B., Montes-Peon M.J. and Vazquez-Ordas, J.C. (2007). Safety culture:
Analysis of the causal relationships between its key dimensions. Journal of Safety
Research, 38(6):627-641.
Field, A. (2005). Discovering statistics using SPSS (and sex, drugs and rock ‘n ‘roll), (2nd edn.).
London: Sage Publications.
Fitzgerald, M.K. (2005). Safety performance improvement through culture change. Process
Safety Environmental Protection, 83(4): 324-330.

450
Flick, U. (2009). An introduction to qualitative research. Thousand Oaks, CA: Sage
Publications.
Flynn, B.B., Schroeder, R.G. and Sakakibara, S. (1994). A framework for quality management
research and an associated measurement instrument. Journal of Operations Management,
11:339-366.
Fock, H.K.Y. and Woo, K.S. (1998). The China market: Strategic implications of Guanxi.
Business Strategy Review, 9(3):33-43.
Fornell, C. (1983). Issues in the application of covariance structure analysis: A comment.
Journal of Consumer Research, 9(4): 443-448.
Forzano, G. (2008). Research methods for the behavioural sciences. United States: Cengage
Learning EMEA.
Foster, V. and Pushak, N. (2011). Ghana’s infrastructure: A continental perspective. World Bank
Policy Research Working Paper No. 5600. Available at SSRN:
https://ssrn.com/abstract=1792258.
Garland, R. (1991). The mid-point on rating scale: Is it desirable? Marketing Bulletin,
2(May):66-70.
Garrod, B. (2008). The Delphi technique. Institute of Rural Sciences, University of Wales:
Aberystwyth, UK.
Garvin, D.A. (1984). What does ‘product quality’ really mean? Slogan Management Review,
26(1):25-43.
Ghana Statistical Service. (GSS). (2012). 2010 Population & Housing Census: National
Analytical Report. Accra, Ghana: GSS.
Ghobadian, A. and Woo, H.S. (1996), Characteristics, benefits and shortcomings of four major
quality awards. International Journal of Quality & Reliability Management, 13(2):10-44.
Goldratt, E. M (1995). The haystack syndrome. Croton-on-Hudson, NY: North River Press.

Goodman, C.M. (1987). The Delphi technique: A critique. Journal of Advanced Nursing,
12:729-734.

451
Green, B., Jones, M., Hughes, D. and Williams, A. (1999). Applying the Delphi technique in a
study of GP’s information requirements. Health and Social Care in the Community,
17(3):198-205.
Grisham, T. (2008). Cross cultural leadership. Doctor of Project Management, School of
Property, Construction and Project Management, RMIT, Melbourne.
Grobbelaar, S. 2001. Die kwaliteitsvereistes van die ISO 9000: 2000 dokumentasie en die invloed
daarvan op die konstruksiebedryf. Unpublished BSc (QS) dissertation, University of
Pretoria, RSA.
Gunning, J.G. and McCallion, E.M. (2007). TQM in large Northern Ireland contracting
organisations. In: Boyd, D. (Ed.). Proceedings of the 23rd Annual ARCOM Conference, 3-
5 September 2007, Belfast, UK, Association of Researchers in Construction Management,
pp. 577-586
Guy, R.F., Edgley, C.E, Arafat, I. and Allen, D.E. (1987). Social research. Puzzles and solutions.
Boston: Allyn and Bacon.
Hackman, J.R. and Wageman, R. (1995). TQM: Empirical, conceptual, and practical issues.
Administrative Science Quarterly, 40(June):309-342.
Häder, M. and Häder, S. (1995). Delphi und kognitionspsychologie:
Einzugangzurtheoretischenfundierung der Delphi-Methode. ZUMA-nachrichten,
37(19):12.
Hair, J. F., Anderson, R.E., Tatham, R.L., Black, W.C. (1998). Multivariate data analysis.
Upper Saddle River, NJ: Prentice Hall.
Hair, J.F., Black, W.C., Babin, J.B., Anderson, R.E. and Tatham, R.L. (2006). Multivariate data
analysis (6th edn.). Upper Saddle River, NJ.: Pearson/Prentice Hall.

Hallowell, M. and Gambatese, J. (2010). Qualitative research: Application of the Delphi method
to CEM research. Journal of Construction Engineering and Management, 136(Special
Issue: Research Methodologies in Construction Engineering and Management):99-107.
Hamilton, J. (1997). The macro-macro interface in the construction of individual fashion forums
and meanings. Clothing and Textiles Research Journal, 15(3):164-171.

452
Handfield, R.B. (1993). A resource dependence view of just-in-time purchasing. Journal of
Operations Management, 11:289-311.
Hardy, J.D., O’Brien, A.P. and Gaskin, C.J. (2004). Practical application of the Delphi technique
in a bicultural mental health nursing study in New Zealand. Journal of Advanced Nursing,
46(1):95-109.
Harrington, J.H. and Voehl, F. (2012). Applying TQM to the construction industry. The Total
Quality Management Journal, 24(4):352-362.
Harris, F. and McCaffer, R (2005). Modern construction management. New Jersey: Blackwell
Science Publishers.
Harris, F. and McCaffer, R. (1995). Modern construction management (4th edn.). Cambridge:
Cambridge University Press.
Harris, F. and McCaffer, R. (2002). Modern construction management. E.P.P Books Series,
Accra, Ghana.
Harris, F., McCaffer, R. and Edum-Fotwe, F. (2006). Modern construction management. New
Jersey: Blackwell Publishing.
Harris, M.M. and Schaubroeck, J. (1990). Confirmatory modeling in OB/HRM: Technical issues
and applications. Journal of Management Information Systems, 16:337-360.
Hassin, E., Tookey, J.E. and Vidalakis, C. (2007) Sustainable development and TQM
implementation in Libya: A study of the electrical generation industry. In: Egbu, C.O. and
Tong, M.K.L. (Eds). Proceedings of the 3rd Scottish Conference for Postgraduate
Researchers of the Built and Natural Environment (PRoBE), 20-22 November 2007,
Glasgow Caledonian University, pp.203-215.
Hasson, F., Keeney, S. and McKenna, H. (2000). Research guidelines for the Delphi survey
technique. Journal of Advanced Nursing, 32(4):1008-15.

Hatzichristos, T. and Giaoutzi, M. (2005). Landfill siting using GIS, fuzzy logic and the Delphi
method (author abstract). International Journal of Environmental Technology
Advancement, 6(1/2):218.
Haupt, T.C. and Whiteman, D.E. (2003). Deploying TQM on construction sites: inhibiting
factors. The International Journal of Construction Management, 3(2): 51-68.

453
Hayes, B.E. (1998). Measuring customer satisfaction: Survey design, use and statistical analysis
methods. Milwaukee: ASQ Quality Press.
Helmer, O. (1977). Problems in futures research: Delphi and causal cross-impact analysis.
Future, 9(1):25-52.
Henderson, J.R.and Ruiker, K. (2010). Technology implementation strategies for construction
organisations. Engineering, Construction and Architectural Management, 17(3): 309 - 327.
Hendricks, K.B. and Singhal, V.R. (1996). Quality awards and the market value of the firm: An
empirical investigation. Management Science, 42(3): 415-436.
Hendricks, K.B. and Singhal, V.R. (1997). Delays in new product introductions and the market
value of the firm: The consequences of being late to the market. Management Science,
43(4):422-436.
Henry, A.L. (1994). The factors associated with insolvency among contractors in the South
African construction industry: A case study of the Western Cape Region. University of
Cape Town in Cape Town, Western Cape, South Africa.
Heppner, P.P. and Heppner, M.J. (2004). Writing and publishing your thesis, dissertation and
research - A guide for students in the helping professions. Belmont: Brooks & Cole
Thomson Learning.
Hill, S. (1995). The social organization of boards of directors. The British Journal of Sociology,
46(2):245–279.
Hillebrandt, P.M. (2000). Economic theory and the construction industry (2nd edn.).
Basingstoke: Macmillan.
Hinze, J. (2001). Construction contracts (2nd edn.). McGraw- Hill Companies.
Hishamuddin, I. (2007). Empirical analysis on factors influencing customer loyalty in Malaysian
telecommunication industry. Multimedia University, Malaysia.
Hitchcock, G. and Hughes, D. (1989). Research and the teacher. London: Routledge.

Ho, S.K.M. and Fung, C.K.H. (1994). Developing a TQM excellence model. Total Quality
Managment Magazine, 6(6):24-30.
Hoinville, G. and Jowell, R. (1978). Survey research practice. Aldershot, Great Britain: Gower
Publishing Company.

454
Holey, E.A., Feeley, J.L., Dixon, J. and Whittaker, V.J. (2007). An exploration of the use of
simple statistics to measure consensus and stability in Delphi studies. BMC Medical
Research Methodology, 7(52):1-10.
Hooper, D., Coughlan, J. and Mullen, M. (2008). Structural equation modelling: Guidelines for
determining model fit. Electronic Journal of Business Research Methods, 6(1):53-60.
Hopkins, A. (2006). Studying organisational cultures and their effects on safety. Safety Science,
44: 875-889.
Hossain, M. M., Prybutok, V.R., Abdullah, A.B.M., and Talukder, M (2010). The development
and research tradition of statistical quality control. International Journal of Productivity
and Quality Management 5(1): 21-37.
Hoyle, R. (1995). SEM concepts, issues and applications. London: Sage Publications.
Hsu, C.C. and Sandford, B.A. (2007). The Delphi technique: Making sense of consensus.
Practical Assessment, Research and Evaluation, 12(10):1-8.

Hsu, I-Y., Su, T-S., Kao, C-S, Shu, Y-L, Lin, P-R. and Tseng, J-M. (2012). Analysis of business
safety performance by structural equation models. Safety Sciences, 50(1):1-11.
Hu, L. and Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation M odeling, 6(1):1-55.

Hudson, C. R., Sears, G.A. and Keoki. S.S. (2000). Construction project management (4th edn.).
Canada: John Wiley & Sons.
Imbeah, K. A. (2012). Framework for the implementation of total quality management (TQM) in
real estate firms in Ghana. MSc Thesis, Facullty of Architecture and Building Technology,
KNUST, Ghana.
Institute of Management Accountants (IMA). (1999). Theory of constraints (TOC) Management
System Fundamentals. Institute of Management Accountants and Arthur Andersen LLP,
USA: , IMA Publication Number 99342
Institution of Occupational Safety and Health (IOSH) (2004). Promoting a positive culture - a
guide to health and safety culture. 4.2. Leicestershire: IOSH.

455
International Labour Organization (2001). The construction industry in the twenty first century:
Its image, employment prospects and skill requirements. Geneva: International Labour
Office.
International Organization for Standardization (ISO) 8402 (1994). Quality management and
quality assurance – vocabulary. , Geneve, Switzerland: ISO.
Iqbal, S. and Pipon-Young, L. (2009). The Delphi method. The Psychologist, 22(7):598-600.

Ishikawa, K. (1985). What is total quality control? The Japanese Way. Englewood Cliffs, NJ.:
Prentice-Hall.
Jack, E.P. and Raturi, A.S. (2006). Lessons learned from methodological triangulation in
management research. Management Research News, 29(6):345-357.
Jean, K. (1992). Livelihood strategies among farm youth in Rwanda. Michigan: Michigan State
University.
Jha, K. and Iyer, K.C. (2006). Critical factors affecting quality performance in construction
projects. TQM & Business Excellence, 17(9): 1155-1170.
Johnson, R. B., Onwuegbuzie, A. J. and Turner, L. A. (2007). Toward a definition of mixed
methods research. Journal of Mixed Methods Research, 1(2):112–133.
Joubert, W., Cruywagen, J. H. and Basson, G. A. J. (2005). Will the implementation of a TQM
system benefit South African construction companies? South Africa Journal of Industrial
Engineering, 16 (1):29-40.
Juran, J. M. (1988). Quality control handbook. New York: McGraw-Hill.

Juran, J.M. and Gryna, F.M. (1993). Quality planning and analysis (3rd edn.). New York:
McGraw-Hill.
Kanji, G.K. and Asher, M. (1993). Total quality management process: A systematic approach.
Advances in TQM Series. Abingdon: Carfax.
Kanji, G.K. and Asher, M. (1996), 100 methods for total quality management. London: Sage
Publications.
Keeney, S., Hasson, F. and McKenna, H.P. (2001). A critical review of the Delphi technique as a
research methodology for nursing. International Journal of Nursing Studies, 38(2):195200.
Kelloway, E. (1998). Using LISEL for SEM: Researchers guide. London: Sage Publications.

456
Kenya Institute of Management (2009). Total quality management: Theory, concepts and
practice. Macmillan Publishers.
Kerlinger, F. and Lee, H. (2000). Foundations of behavioural research. Fort Worth, TX:
Harcourt.
Kerzner, H. (2003). Project management: A System approach to planning, schedualing, and
controlling. New Jersey: John Wiley & Sons.
Khadour, L. and Darkwa, J. (2008). TQEM in the UK construction industry: Some key findings
from a survey. In: Dainty, A. (Ed.). Proceedings of the 24th Annual Association of
Researchers in Construction Management (ARCOM) Conference, 1-3 September 2008,
Cardiff, UK, pp.239-246.
Kinloch, G.C. (1977). Sociology theory: Its development and major paradigms. New York: Mc
Graw-Hill.
Kiwus, C.H. and Williams, T.P. (2001). Application of TQM to environmental construction.
Journal of Management in Engineering, 17(3):176-184.
Kline, R.B. (2005). Principles and practice of structural equation modeling. (2nd edn.). New
York: Guilford Press.
Kline, R.B. (2010). Principles and practice of structural equation modeling. (3rd edn.). New
York: Guilford Press.
Koh, T.Y. and Low, S.P. (2010). Empiricist framework for TQM implementation in construction
companies. Journal of Management in Engineering, 26(3):133-143.
Kubal, M.T. (1994). Engineered quality in construction. In: Assessing contractor quality
performance. Construction Management and Economics, 20: 211.
Kumaraswamy, M. M. and Dissanayaka, S. M. (2000). ISO 9000 and beyond: From a Hong
Kong construction perspective. Journal of Construction Management and Economics,
18(7): 783-796.
Kwakye, A. A. (1998). Construction project administration in practice. Ascot, Berkshire: The
Chartered Institute of Buildings.
Kwaw, P.K., Yalley, P.P., Cobbinah, J.F. and Opintan-Baah, E. (2011). Ghana’s public
procurement reform and continuous use of traditional procurement system: The case of the
construction industry. Takoradi Polytechnic Journal of Technology, 1(1): 51-64.
457
Lad, J. and Beck, B. (2009). Construction quality: The key to successful capital projects delivery.
Pharmceutical Engineering, 29(6).
Landeta, J. (2006). Current validity of the Delphi method in social sciences. Technological
Forecasting and Social Change, 73(5):467-482.
Lang, T. (1995). An overview of four futures methodologies (Delphi, environmental scanning,
issues management and emerging issue analysis). The Manoa Journal of Fried and
Halffried Ideas (about the future), 7 (Occasional Paper Number Seven) - Hawaii Research
Center for Futures Studies, Honolulu.
Larsson, S., Pousette, A. and Törner, M. (2008). Psychological climate and safety in the
construction industry-mediated influence on safety behavior. Safety Science, 46(3):405-
412.
Lee, A. S. (1991). Integrating positive and interpretative approach to organisational research.
Organisational Science, 2(4):342 – 365.
Leedy, D.P. and Ormrod, E.J. (2010). Practical research: Planning and designing (9th edn.).
Upper Saddle River, NJ.: Pearson Education.
Leedy, P.D. and Ormrod, J.E. (2005). Practical research: Planning and design. (8th edn.).
Upper Saddle River, NJ.: Prentice Hall.
Lennon, S.L. and Burns, L.D. (2000). The diversity of research in textiles, clothing, and human
behavior: The relationship between what we know and how we know. Clothing Textiles
Research Journal, 18(4):213-226.
Lewis, T.M. (2007). Impact of globalization on the construction sector in developing countries.
Construction Management and Economics, 25(1): 7–23.
Lincola, Y.S. and Guba, E.G. (1985). Naturalistic inquiry. London: Sage Publications.

Lingard, H., Wakefield, R. and Cashin, P. (2011). The development and testing of a hierarchical
measure of project OHS performance. Engineering Construction and Architectural
Management, 18(1):30-49.
Linstone, A. and Turoff, M. (1975). The Delphi method: Techniques and applications. Reading,
MA: Addison-Wesley.

458
Linstone, H.A. (1978). The Delphi technique. In: Fowlers, J. Handbook of Futures research,
(Ed.). J. Fowlers, Westport, CT: Greenwood Press, pp. 273-300.
Linstone, H.A. and Turoff, M. (2002). The Delphi method: Techniques and applications.
Available from: http://www.is.njit.edu/pubs.php [Accessed 23 November 2016].
Lobelo, L. (1996). An investigation into factors associated with insolvencies amongst civil
engineering contracting firms in South Africa. University of Cape Town in Cape Town,
Western Cape, South Africa.
Loo, R. (2002). The Delphi method: A powerful tool for strategic management, policing. An
International Journal of Police Strategies and Management, 25(4):762-769.
Love, P. E. D., Li, H., Irani, Z., and Holt, G.D. (2000). Re-thinking TQM: Toward a framework
for facilitating learning and change in construction organisations. The Total Quality
Management Magazine, 12(2):107-117.
Love, P.E.D. and Li, H. (2000). Quantifying the causes and costs of rework in construction,
Construction Management and Economics, 18:479-490.
Love, P.E.D., Edwards, D.J. and Sohal, A. (2004). TQM in Australian contracting organisations:
Pre-conditions for successful implementation. Engineering, Construction and Architectural
Management, 11(3):189-198.
Low, S.P. and Jasmine, A.T. (2004). Implementing total quality management in construction
firms. Management in Engineering, 20(1):8-15.
Lucko, G. and Rojas, E.M. (2010). Research validation: Challenges and opportunities in the
construction domain. Journal of Construction Engineering and Management,
136(1):127135.

Luus, C. (2003). The ABSA residential property market database for South Africa: Key data
trends and implication. In: Proceedings of the IMF/BIS Conference on Real Estate
Indicators and Financial Stability, Washington DC, 27–28 October.
Maart, L. and Soa, S. (1996). From political to development practice: Facing the challenges
of field. Durban: Olive Subscription Service.
Maguire, M. (1987). Doing participatory research: A feminist approach. Amherst, MA: The
Center of International Education, University of Massachussetts.

459
Mahmood, W., Yusoff, W. and Mohammed, A. H. (2008). A conceptual framework for the
development of quality culture in the construction industry. In: Dainty, A. (Ed.).
Proceedings of the 24th Annual Association of Researchers in Construction Management
(ARCOM) Conference, 1-3 September 2008, Cardiff, UK, pp.247-256.
Malhotra, N.K. (1999). Marketing research: An applied orientation (3rd edn.). New Jersey:
Prentice Hall.
Mangan, J., Lalwani, C. and Gardner, B. (2004). Combining quantitative and qualitative
methodologies in logistics research. International Journal of Physical Distribution &
Logistics Management, 34(7):565-578.
Mann, R. and Kehoe, D. (1994). An evaluation of the effects of quality improvement activities
on business performance. International Journal of Quality & Reliability Management,
11(4):29-44.
Mann, R.S. (1992). The development of a framework to assist in the implementation of Total
Quality Management. PhD thesis, Department of Industrial Studies, University of
Liverpool, UK.
Manu, P., Ankrah, N., Proverbs, D. and Suresh, S. (2010). Exploring the causal influence of
construction project features in accident causation. In: Barrett, P., Amaratunga, D., Haigh,
R., Keraminiyage, K. and Pathirage, C., eds. Proceedings of the CIB 2010 World
Congress, Salford, UK, 10th – 13th May, 2010., pp. 241-252.
Marcoulides, G. and Hershberger, S. (1999). Multivariate statistical methods: A first course.
Structural Equation Modeling, 6(1):127-133.

Martinez-Lorente, A. R., Dewhurst, F. and Dale, B.G (1998). Total quality management: Origin
and evolution of the term. The Total Quality Management Magazine, 10(5):378-386.
Martins, E.C. and Terblanche, F. (2003). Building organisational culture that stimulates
creativity and innovation. European Journal of Innovation Management, 6(1): 64-74.
Maruyama, G. (1998). Basics of structural equation modelling. London: Sage Publications.
Masini, E. (1993). Why futures Studies? London: Grey Seal.
Masser, I. and Foley, P. (1987). Delphi revisited: Expert opinion in urban analysis. Urban
Studies, 24(3):217-224.

460
May, T. (2001). Social research - Issues, methods and process (3rd ed.). Buckingham &
Philadelphia: Open University Press.
Mazzochi, F. (2008). Complexity in biology. Exceeding the limits of reductionism and
determinism using complexity theory. European Molecular Biology Organization Reports,
9(1):10-14.
Mbande, C. (2010). Overcoming construction constraints through infrastructure deliver. In:
Proceedings of the Association of Schools of Construction of Southern Africa (ASOCSA),
Fifth Built Environment Conference, Durban, South Africa, 18–20 July.
McClave, J.T., Benson, P.G. and Sincich, T. (2008). Statistics for business and economics. (10th
edn.). United States: Pearson Prentice Hall.
McCutcheon, R.T. and Taylor, F.L.M. (2003). Employment and high-standard infrastructure.
Johannesburg: Work Research Center for Employment Creation in Construction.
McFarlin, D.B., Coster, E.A. and Mogale-Pretorius, C. (1999). South African management
development in the 21st century. Journal of Management Development, 18(1):63-78.
McIntyre, C. and Kirschenman, M. (2000). Survey of TQM in construction industry in upper
Midwest. Journal of Management in Engineering, 16(5):67-70.
McIntyre, L.J. (1999). The practical skeptic: Core concepts in sociology. Mountain View, CA:
Mayfield Publishing.
McKenna, H. (1994). The Delphi technique: A worthwhile research approach for nursing?
Journal of Advanced Nursing, 19(6):1221-1225.

Meadows, D. H. (2009). Thinking in systems. A primer. London, UK: Earthscan.

Mensah, S. (2012). Effects of project management practices on project performance: The case of
three organisations. Published monograph. Lambert Academic Publishing.
Meredith, J.R. and Shafer, S.M. (1999). Operations management for MBAs. New York: John
Wiley & Sons.
Metri, B.A. (2005). TQM critical success factors for construction firms. Management
Economics, 10(2):61-72.
Milford, R. (2010). Public capacity payment and procurement issues should be a challenge to the
operations of contractors in South Africa. Paper presented at the Fifth Built Environment

461
 Conference of the Association of Schools of Construction of Southern Africa, Durban,
South Africa, 18–20 July.
Miller, M.M. (1993). Enhancing regional analysis with the Delphi method. Review of Regional
Studies, 23(2):191-212.
Ministry of Water Resources, Works and Housing (MWRWH) (2012). A Draft Building Code
for Ghana. Accra: Ministry of Water Resources, Works and Housing.
Mitchell, J.C. (1983). Case and situation analysis. Sociological Review, 31(2):187-211.

Mitchell, M. and Jolley, J. (1992). Research design explained (2nd ed.). Forth Worth, TX:
Harcourt Brace Jovanovich College Publishers.
Mohammed, S.M. and Rotimi, J.O. (2002). A review of quality control and assurance in building
construction. A paper presented at the Millennium Conference, held at A.B.U Zaria.
Mohammed, U. S. (2011). Green buildings: A sustainable construction approach to sustainable
built environment. A paper presented at the workshop held at Multi-Purpose Hall, Gombe
State University, 3-7 July 2011, pp. 5.
Molenaar, R.K., Park, J-II. and Washington, S. (2009). Framework for measuring corporate
safety culture and its impact on construction safety performance. Journal of Construction
Engineering and Management, 135(6):488-496.
Morgan, D. L. (2006). Connected contributions as a motivation for combining qualitative and
quantitative methods. Washington, DC: American Public Health Association.
Mosadeghrad, AM. 2014. Why TQM programmes fail? A pathology approach. The Total
Quality Management Journal, 26(2):160-187.
Mulaik, S.A., James, L.R., Van Alstine, J., Bennet, N., Lind, S. and Stilwell, C.D. (1989).
Evaluation of goodness-of-fit indices for structural equation models. Psychological
Bulletin, 105(3):430-45.
Munn, P. and Drever, E. (1990). Using questionnaires in small-scale research: A teachers' guide.
Edinburgh: SCRE.
Musonda, I. (2012). Construction health and safety (H&S) performance improvement. A client-
centred model. Unpublished DPhil in Engineering Management, University of

462
 Johannesburg. South Africa. Available from http://www.ujdigispace.uj.ac.za [Accessed 02
June 2016].
Muthén, L.K. and Muthén, B.O. (2007). Mplus, user’s guide. (5th edn.). Los Angeles, CA:
Muthén & Muthén.
Nazhad M. B., Mosavi J. S. and Kordabadi S. S. (2012). Organizational culture and quality
management (ISO/9001) case study: Tehran Universities employees. Journal of Basic and
Applied Scientific Research, 2(12):12590-12599.
Nee, P.A. (1996). ISO 9000 in construction. New York, USA.: Wiley.
Nesan, L. J. and Holt, G.D. (1999). Empowerment in construction organisations: The way
forward for performance improvement. England: Research Studies Press.
Neuman, W.L. (2006). Social research methods. Qualitative and quantitative approaches. (6th
edn.). Boston: Pearson.
Newcombe, R., Fellows R., Langford, D. and Urry, S. (2001). Construction management in
Practice. (Vol. 1). London: Blackwell Science.
Ngowi, A. B. (2000). Impact of culture on the application of TQM in the construction industry in
Botswana. International Journal of Quality & Reliability Management, 17(4/5):442-452.
Ngowi, AB. 2001. Impact of culture on the application of TQM in the construction industry in
Botswana. International Journal of Quality and Reliability Management, 17(4/5): 442-452.
Nicco–Annan, J. (2006). Partnering in construction. The Quantity Surveyor, 1:14-19.
Norberg, J. and Cumming, G. (2008). Complexity theory for a sustainable future. New York,
NY: Columbia University Press.
Nuamah, F. A., Manu, P., and Manu, E. (2013). A preliminary inquiry into the applicability of
client-contractor partnering in the Ghanaian Construction industry. In: Laryea, S. and
Agyepong, S. (Eds). Proceedings of 5th West Africa Built Environment Research (WABER)
Conference, 12 – 14th August 2013, Accra, Ghana, pp.77 – 85.
Nukulchai, W. K. (2003). Towards total quality management in higher education: The
experience of Asian Institute of Technology. Asian Institute of Technology.
Oakland, J. (2000). TQM – Text with cases (2nd edn.). Butterworth Heinemann.
Oakland, J. and Marosszeky, M (2006). Total quality in the construction supply chain. Oxford,
UK: Elsevier.
463
Ofori, G. (2012). Developing the construction industry in Ghana: The case for a central agency.
Available from http://www.buildingcontractorsgh.com [Accessed 24 November 2015].
Okoli, C. and Pawlowski, S. (2004). The Delphi method as a research tool: An example, design
considerations and applications. Information Management, 42:15-29.

Oliver, G.B.M. (1992). Quality management in construction - Implementation in design services

organisations. London: CIRIA.
Ozaki, R. (2003). Customer–focused approaches to innovation in house building. Construction
Management and Economics, 21(6): 557-564.
Pallant, J. (2007). SPSS, Survival manual: A step-by step guide to data analysis using SPPS
Version 1.5. (3rd edn.). McGraw Hill.
Parente, R.J., Hiob, T.N., Silver, R.A., Jenkins, C., Poe, M.P. and Mullins, R.J. (2005). The
Delphi method, impeachment and terrorism: Accuracies of short-range forecasts for
volatile world events. Technological Forecasting and Social Change, 72(4):401-11.
Pheng, L.S. and Teo, J.A. (2004). Implementing TQM in construction firms. Journal of
Management in Engineering, 20(1): 8-15.
Philips Quality (1995). Philips quality - Let’s make things better. Eindhoven, The Netherlands:
Corporate Quality Bureau, Philips Electronics N.V.
Phillips, R. (2000). New applications for the Delphi technique. San Diego: Pfeiffer and
Company.

Phua, F. T. T. (2006). Predicting construction firm performance: An empirical assessment of the

differential impact between industry- and firm-specific factors. Construction Management
and Economics, 24(3):309-320.
Pinsonneault, A. and Kraemer, K.L. (1993). Survey research methodology in management
information systems: An assessment. Journal of Management Information Systems,
10(2):75-105.
Powell, T.C. (1995). TQM as competitive advantage: A review and empirical study. Strategic
Management Journal, 16: 15-37.
Project Management Institute. (2000). A guide to the project management body of knowledge
(PMBOK Guide). Newtown Square: PSI.
464
Pyzdek, T. (2003). Quality engineering handbook. CRC Press.

Quattrone, P. (2000). Constructivism and accounting research: Towards a trans-disciplinary

perspective. Accounting, Auditing & Accountability Journal, 13(2):130- 155.
Ramachandran, V. (2010). TQM in construction. Available at
http://www.brighthub.com/office/project-management/articles/86518 [Accessed 24 June
2015].
Raskin, M.S. (1994). The Delphi study in field instruction revisited: Expert consensus on issues
and research priorities. Journal of Social work education, 30:75-89.
Rategan, C. (1992). TQM. Journal of Property Management, 57: 32-34.
Rayens, M.K. and Hahn, E.J. (2000). Building consensus using the policy Delphi method.
Policy, Politics and Nursing Practices, 1(2):308-31.
Raykov, T. and Marcoulides, G. (2000). A first course in structural equation modeling. New
Jersey: Lawrence Erlbaum Associates.

Reid, N.G. (1988). The Delphi technique, its contributions to the evaluation of professional
practice. In: Ellis. R. (Ed.). Professional competence and quality assurance in the caring
professions. Chapman & Hall, pp. 230-62.

Republic of Ghana. (2011). Ghana at a glance. Know more about Ghana. Available from
http://www.ghana.gov.gh/index.php?option=com_content&view=category&layout=blog&
id=92&Itemid=238. [Accessed 10 December 2016].

Republic of South Africa. Department of Public Works. (1999). Creating an Enabling

Environment for Reconstruction, Growth and Development in the Construction Industry.
The White Paper, Pretoria, South Africa: Government Printing Works.
Robson, C. (2002). Real word research: A research for social scientists and
practitionerresearch. Oxford: Blackwell Publishing.

Rogers, M.R. and Lopez, E.C. (2002). Identifying critical cross-cultural school psychology
competencies. Journal of School Psychology, 40(2):115-141.

Ross, J.E. (1993). Total quality management. Delray Beach, FL.: St. Lucie Press.

465
Rowe, G., Wright, G. and Bolger, F. (1991). Delphi - A re-evaluation of research and theory.
Technological Forecasting and Social Change, 39:238-251.

Ruyter, J.C. and Scholl, N.B. (1998). Positioning qualitative market research: Reflections from
theory and practices. Qualitative Market Research Journal, 1(1):7-14.

Saad, M. U. and Dahiru, D. (2015) Project quality management practices in multinational and
indigenous construction firms of Nigeria. In: Laryea, S. and Leiringer R. (Eds.).
Proceedings of the 6th West Africa Built Environment Research (WABER) Conference, 10-
12 August 2015, Accra, Ghana, pp.951-960.
Sackman, H. (1974). Delphi assessment: Expert opinion, forecasting, and group process. Santa
Monica, CA: Rand Corporation.
Saizarbitoria, I.H. (2006). How quality management models influence company results –
conclusions of an empirical study based on the Delphi method (ISO 9000 and EFQM
quality management standards of Europe). Total Quality Management and Business
Excellence, 17(6):775-94.
Sanni, A.A. and Windapo, A.O. (2008). Evaluation of contractor‘s quality control practices on
construction sites in Nigeria. In: Dainty, A. (Ed.). Proceedings of the 24th Annual
Association of Researchers in Construction Management (ARCOM) Conference, 1-3
September 2008, Cardiff, UK, pp.257-265.
Sarantakos, S. (2005). Social research (3rd edn.). Basingstoke: Palgrave Macmillan.

Saraph, J.V., Benson, G.P. and Schroeder, R.G. (1989). An instrument for measuring the critical
factors of quality management. Decision Sciences, 20:810-829.
Scheele, D.S. (2002). Reality construction as a product of Delphi interaction. The Delphi
method: Techniques and applications, 37.
Schoemaker, P.J.H. (1993). Multiple scenario development: Its conceptual and behavioral
foundation. Strategic Management Journal, 14(3):193-213.

Schreiber, B.J., Stage, K.F., King, J., Nora, A. and Barlow, A.E. (2006). Reporting structural
equation modelling and confirmatory factor analysis results: A review. The Journal of
Educational Research, 99(6):323-338.

466
Schumacker, R. and Lomax, R. (2004). A beginner’s guide to structural equation modelling.
New York: Lawrence Erlbaum.
Seaver, M. (Ed.). 2003. Gower handbook of quality management. Aldershot, Great Britain:
Gower Publishing Company.
Sekaran, U. (2000). Research method for business: A skill building approach. New York: John
Wiley and Sons.
Seo, D.C., Torabi, M.R., Blair, E.H. and Ellis, N.T. (2004). A cross validation of safety climate
scale using confirmatory factor analytic approach. Journal of Safety Research, 35(4):427-
445.
Sharma, P.C. (2006). A textbook of production engineering. New Delhi: Rajendra Ravindra
(PVT).
Shenkar, O. and Von Glinow, M.A. (1994). Paradoxes of organizational theory and research:
Using the case of China to illustrate national contingency. Management Science, 40(1):56-
71.
Shook, C.L., Ketchen, D.J., Hult, T. and Kacmar, J.M. (2004). An evaluation of structural
equation modeling in strategic management research. Strategic Management Journal,
25:397-404.
Sila, I. and Ebrahimpour, M. (2002). An investigation of the total quality management survey-
based research published between 1989 and 2000. International Journal of Quality and
Reliability Management, 19(7):902-970.
Silva, M.C. (1977). Philosophy, science, theory. Interrelationships and implications for nursing
research. Image, 9(3), 59-63.
Simmonds, C. (1977). The nature of futures problems in futures research: New directions.
London: Addison-Wesley Publishing Company.
Singh, P. J. and Alan Jr., S. (2004). Relationship between TQM and innovation: An empirical
study. Journal of Manufacturing Technology Management, 15(5):394 - 401.

Skulmoski, G.J., Hartman, F.T. and Krahn, J. (2007). The Delphi method for graduate research.
Journal of Information Technology Education, 6:1-21.

467
Slack, N., Chambers, S., Harland, C., Harrison, A. and Johnston, R. (1995), Operations
management. London: Pitman Publishing.
Smallwood, J.J. (2000). A study of the relationship between occupational health and safety,
labour productivity and quality in the South African construction industry. Unpublished
PhD, University of Port Elizabeth in Port Elizabeth, Eastern Cape, South Africa.
Smith, M.A. (1997). Perceptions of quality in journalism and communications education: A
Delphi study. Journal of the Association for Communication Administration, 1:32-50.

Smith, M.H. (2004). A sample/population size activity: Is it the sample size of the sample as a
fraction of the population that matters? Journal of Statistics Education, 12(2):1-12.

Society of Actuaries (SOA). (1999). Final report of the 1999 Delphi study. Schaumburg, IL:
South Africa at a glance (2011). Provinces & Metropolitan Municipalities. Available from
http://www.southafricaataglance.com/south-africa-provinces.html. [Accessed 17
September 2016].
Sommerville, J. and Sulaiman, N.F. (1997). An analysis of cultural elements influencing the
implementation of TQM in the UK construction industry. Norwegian Building Research
Institute, pp.77-85.
South African Government Information. Available from
http://www.info.gov.za/aboutsa/economy.htm [Accessed 26 November 2009].
South African Reserve Bank. (2009). Monetary policy review - May 2009. Pretoria: Reserve
Bank.
Spinelli, T. (1983). The Delphi decision-making process. Journal of Psychology, 113:73-80.

Statistics South Africa (2009). Labour Force Survey. Available from

http://www.statssa.gov.za/publications/P0210/P0210September2000,2001,2002,2003,2004
,2005 2006,2007.pdf.[Accessed 02 January 2010].
Statistics South Africa (Stats SA). (2010). Statistical Release P0441 Gross Domestic Product
Fourth Quarter: 2009. Pretoria, South Africa: Stats SA. Available from
www.statssa.gov.za [Accessed 12 December 2015].
Steiger, J. (2007). Understanding the limitations of global fit assessment in structural equation
modelling. Personality and Individual Differences, 42(5): 893-898.
468
Steingard, D.S. and Fitzgibbons, D.E. (1993), A postmodern deconstruction of TQM, Journal of
Organization Change Management, 6(5)27- 42.
Stenbacka, C. (2001). Qualitative research requires quality concepts of its own. Management
Decision, 39(7):551-555.
Steyn, H., Basson, G., Carruthers, M., du Plesis, Y., Prozesky-Kuschke, B., Kruger, D., van Eck,
S. and Visser, K. (2004). Project management: A multi-disciplinary approach. Pretoria:
FPM Publishing.
Stitt-Gohdes, W.L. and Crews, T.B. (2004). The Delphi technique: A research strategy for career
and technical education. Journal of Career and Technical Education, 20(2):55-67.
Subhash, K., and Narag, A.S. (2007). Recommending a TQM model for Indian organisations.
The TQM Magazine, 19(4):328-353.
Sweeny, L. (2009). A study of current practice of corporate social responsibility (CSR) and an
examination of the relationship between CSR and financial performance using structural
equation modeling (SEM). Unpublished Doctoral thesis, Dublin Institute of Technology,
Ireland.

Tabachnick, B. G. and Fidell, L. S. (2007). Using multivariate statistics. (5th edn). New York:
Allyn and Bacon.

Tabachnick, B.G. and Fidell, L.S. (2001). Using multivariate statistics. (4th edn.). Boston: Allyn
& Bacon.

Tackaberry, P. (1998). Intellectual property risks in China: Their effect on foreign investment
and technology transfer. Journal of Asian Business, 14(4):1-38.
Takim, R., Akintoye, A. and Kelly, J. (2004). Analysis of performance measurement in the
Malaysian construction industry. In: Ogunlana, S. O., Chareonngam, C., Herabet, P. and
Hadikusumo, B. H. W. (Eds.). Globalization and Construction, AIT Conference Centre,
Bangkok, Thailand, pp.533–546.
Talha, M. (2004). Total quality management (TQM): An overview. Journal of the Bottom Line:
Managing Library Finances, 17(1):15-19.

469
Tam, L. W. H. (2000). Quality Management Theory and Practice: Some Observations of
Practices in Australian Academic Libraries. Library Management, 21(7): 349-356.

Tamblyn, A. & Shelton, D. (1996). Market research manual for providers of vocational
education and training. Melbourne: Victoria Office of Training and Further Education.

Tang, S.L., Ahmed, S.M., Aoieoung, R.T. and Poon, S.W. (2005). Contruction quality
management. Hong Kong: Hong Kong University Press.
Tashakkori, A. and Teddlie, C. (1998). Mixed methodology: Combining qualitative and
quantitative approaches: Applied social research methods, 46. Thousand Oaks, CA: Sage
Publications.
Tashakkori, A. and Teddlie, C. (2003). Handbook of mixed methods in social & behavioral
research. Thousand Oaks: Sage Publications.

Teddlie, C. and Tashakkori, A. (2009). Foundations of mixed methods research. Thousand

Oaks, CA: Sage Publications.

Terkper, S. E. (2013). Budget statement and economic policy. Accra, Ghana: Ministry of
Finance.
Thach, E.C. and Murphy, K.L. (1995). Training via distance learning. Training and
Development, 49(12):44.
Thangaratinam, S. and Redman, C.W.E. (2005). The Delphi technique. The Obstetrician &
Gynaecologist, 7(2):120-125.
Tingey, M. (1997). Comparing ISO 9000, Malcolm Baldrige, and the SEICMM for Software: A
reference and selection guide. Englewood Cliffs, NJ.: Prentice-Hall.
Tomlinson, M.R. (2010). Ensuring mortgage repayment in South Africa. Housing Finance
International. Available from http://www.housingfinance.org.pdf [Accessed 12 December
2015].
Tong, D.Y. (2007). An empirical study of e-recruitment technology adoption in Malaysia:
Assessment of modified technology acceptance model. Multimedia University, Malaysia.
Touliatos, J. and Compton, N.H. (1988). Research methods in human ecology/home economics.
Ames, IA: Iowa State University Press.
470
Trochim, W. and Donnelly, J. (2007). The research methods knowledge base (3rd edn.).
Cincinnati, OH: Atomic Dog Publishing.
Turoff, M. (1970). The design of a policy Delphi. Technological Forecasting and Social
Change, 2(2):140-71.
Tutesigensi, A. and Pleim, V. (2008) Why small and medium construction enterprises do not
employ six sigma. In: Dainty, A. (Ed). Proceedings of the 24th Annual Association of
Researchers in Construction Management (ARCOM) Conference, 1-3 September 2008,
Cardiff, UK, pp. 267-276.
United States Department of Defense (1988), The role of concurrent engineering in weapons
system acquisition, IDA Report R-338. Alexandra: United States: Institute of Defense
Analysis.
Uysal, M. and Crompton, J. (1985). An overview of approaches used to forecast tourism
demand. Journal of Travel Research, 23(4):7-15.

Van Wyk, L. (2003). A review of the South African construction industry Part 1: Economic,
regulatory and public sector capacity influences on the construction industry. Pretoria,
South Africa: CSIR Boutek. Available from http://www.csir.co.za/akani [Accessed 15
November 2015].
Van Wyk, L. (2004). A review of the South African construction industry Part 2: Sustainable
construction activities. Pretoria, South Africa: CSIR Boutek. Available from
http://www.csir.co.za/akani [Accessed on 18 December 2015].
Verster, J. J. P. (1998). Strategies for successfully meeting the demands of quality and service
excellence. Acta Structillia. Journal for the Physical & Development Sciences, University
of the Free State, Bloemfontein, RSA.
Vibe Ghana (2014). Towards establishing the “Construction Industry Development Authority” in
Ghana. Available from http://vibeghana.com/2014/0 [Accessed 10 November 2015].
Von Bertalanffy, L. (1968). General system theory. New York, NY: George Braziller.

Vrijhoef, R. and Koskela, L. (2000). The four roles of supply chain in construction. European
Journal of Purchasing & Supply Management, 6:169-178.

471
Vuppalapati, K., Ahire, S.L. and Gupta, T. (1995). JIT and TQM: A case for joint
implementation. International Journal of Operations & Production Management,
15(5);84-94.
Wadsworth, H.M., Stephens, K.S. and Godfrey, A.B. (1986). Modern methods for quality
control and improvement. John Wiley & Sons.
Waldman, D.A. (1994). The contribution of TQM to a theory of work performance. Academy of
Management Review, 19(3):510-536.
Wessel, G. and Burcher, P. (2004). Six sigma for small and medium-sized enterprises. The TQM
Magazine, 16(4):264 - 272.
Whalen, M.J. and Rahim, M.A. 1994. Common barriers to implementation and development of
TQM Program. Industrial Management, 36(2):19.
Wheatley, M. J. (2006). Leadership and the new science: Discovering order in a chaotic world.
San Francisco, CA: Berrett-Koehler Publishers.
Whitman, N.I. (1990). The committee meeting alternative: Using the Delphi technique. The
Journal of Nursing Administration, 51(1):57-68.

Wiegmann, D. A., Zhang, H., von Thaden, T.L., Sharma, G.and Mitchell, A.A. (2002). A
synthesis of safety culture and safety climate research. University of Illinois Aviation
Research Lab Technical Report ARL-02-03/FAA-02-2.
Wilber, K. (2000). Sex, ecology, spirituality. Boston, MA: Shambhala Publications.

Womack, J.P., Jones, D.T. and Roos, D. (1990). The machine that changed the world. New
York: Massachusetts Institute of Technology, Rawson Associates.
Wong, T.C. (1999). Marketing research. Oxford, UK: Butterworth-Heinemann.

World Bank (1984). Housing and financial institutions in developing countries. Washington DC:
World Bank.
World Bank (2004). The World Bank Report 2004 (Vol. 1). Retrieved from http://www.world
bank.org/html/extpb/2004/pdf/volume_1.pdf. [Accessed 09 November 2015].
Woudenberg, F. (1991). An evaluation of Delphi. Technological Forecasting and Social
Change, 40:131-150.

472
Yin, R.K. (1994). Case study research: Design methods. San Francisco: Sage Publications.

Yousuf, M.I. (2007). The Delphi technique. Essays in Education, 20:80-9.

Yu, C.S., Cochran, D.S. and Spencer, B. (1998), Quality management practices in China. Quality
Management Journal, 5(2): 91-106
Yuan, K. (2005). Fit indices versus test statistics. Multivariate Research, 40(1):115-148.

Yuen, Y.Y. (2007). An ergonomic study of mobile phone usage in Malaysia. Multimedia
University, Malaysia.
Zairi, M. (1991). Total quality management for engineers. Woodhead Publishing.
Zairi, M. (2006). Total quality management: Deming Juran Gift to the World (2nd edn.).
Clitheroe: Spire City Publishing.

Zhang, Z. (2000a). Developing a model of quality management methods and evaluation their
effects on business performance. TQM, 11(1):129-137.
Zhang, Z. (2000b). Implementation of total quality management: An empirical study of Chinese
manufacturing firms. PhD Thesis, Faculty of Management and Organization, University of
Groningen, The Netherlands.
Zikmund, W.G. (2000). Exploring marketing research. (7th edn.). Fort Worth: Dryden Press.

473
 APPENDICES
APPENDIX A

January 20, 2017

Dear Sir
INVITATION TO PARTICPATE IN A DELPHI SURVEY
I am Samuel Kwame Ansah, a registered Ph.D. candidate in the Faculty of Engineering and the
Built Environment at the University of Johannesburg. My Ph.D. study is under the supervision of
Prof. Wellington Didibhuku Thwala (Master’s Programme Co-ordinator, Faculty of Engineering
and the Built Environment, University of Johannesburg, South Africa, and Vice-President of the
South African Council for Project and Construction Management Professions - SACPCMP).
The area of my research is developing an Integrated Total Quality Management Model for the
Ghanaian construction industry. I will be using a Delphi approach and need to compile a panel of
experts in the field to participate in this process.
I understand that your time is important. Each round of the Delphi process will take
approximately 10 - 15 minutes to complete. A more complete description of the Delphi process
is attached for your information.
It would be appreciated if you would consent to participating in the study in this capacity.

Kind regards
Samuel Kwame Ansah, Ph.D. Candidate
University of Johannesburg
Faculty of Engineering and the Built Environment
Ghanaian mobile number: +233246769673
South African mobile number: +277837420742

474
Email: [email protected]
APPENDIX B

Delphi Invitation Follow-Up Letter

10th February, 2017
Dear Sir
I would like to thank you for accepting the invitation to participate as an expert in our project to
develop an integrated Total Quality Management model for the Ghanaian construction industry.
The process of collecting input from the expert panel will probably involve no less than three
rounds.
We understand that your time is important. Each round of the Delphi process will take
approximately 20 minutes to complete. A more complete description of the Delphi process is
attached for your information.
To start with, I would like to request your curriculum vitae for our records and to confirm your
area of expertise. We would appreciate your response by the 24th of February 2017.
Kind regards

Samuel Kwame Ansah, Ph.D. Candidate

University of Johannesburg
Faculty of Engineering and the Built Environment
Ghanaian mobile number: +233246769673
South African mobile number: +277837420742
Email: [email protected]

475
 APPENDIX C
Delphi Method and Application to this study
BACKGROUND INFORMATION
The Delphi Method

The Delphi method is a process to collect data and information to solve non-analytical problems.
Used as a research tool, the process gathers knowledge from individuals (experts), analyses and
combines the information to obtain a group consensus.

The Process

The information is gathered in a series of questionnaires or surveys called rounds. The first
round is exploratory in nature and presents the participant with a standard questionnaire. The
second round will present the participant with the group response aling with his or her response
from the first round. Each expert (participants) member has the opportunity to alter his or her
answer or to voice his or her opinion about new issues collected in the previous survey. The
third round, if necessary, will finalise the statistical response of the group to form a consensus.

The Advantages of Delphi

Delphi has three features over other data collection methods: (1) anonymity, (2) controlled
feedback of results, and (3) statistical group response or consensus.
Anonymity – The members of the process are unknown to other members. This feature will help
minimise the “bandwagon effect.” In public group meetings, one participant, possibly less
knowledgeable, may be more vocal during discussion, potentially persuading more
knowledgeable panellists. Also by keeping the participants unknown, one participant may
change his or her answer to one question without publicly admitting that he or she has done so.
Controlled Feedback – The benefit to the participant of Delphi is gained by feedback of results
collected in earlier rounds. The participant will be sent the group response of colleagues and
other experts in the industry.

476
Statistical Group Response – The goal of Delphi is to move towards a group consensus.
However, the end result will undoubtedly display a range of opinions. The statistical group
response is created to assure that the opinions of all participants in the surveys are represented.
The Application of Delphi

The study aims to determine the factors that influence TQM (TQM) implementation in the
construction industry and develop a holistic integrated TQM model for the Ghanaian
construction industry. The title of the study is “An Integrated TQM Model for the Ghanaian
construction industry”. Your participation in the Delphi process for a doctoral study is very
important.

Time Commitment

The time commitment is minimal. Each round or questionnaire should take approximately 10 -
15 minutes to complete and submit.

477
 APPENDIX D

DELPHI INSTRUCTIONS FOR ROUND ONE QUESTIONNAIRE

DELPHI SURVEY – ROUND 1

Thank you for accepting to serve on the Delphi panel for this research. Your acceptance is highly
appreciated. This first round survey is intended to be completed in approximately 15-20 minutes.
Subsequent surveys will require significantly less time to complete. You will be given the
opportunity to change your response later on after all Delphi participants have completed the first
round survey and results have been analysed. The results will be communicated to you. The
results will be in simple statistics e.g. median, average, range and percentage.
When you have finished answering all of the questions, please email your response to
[email protected] by Friday, March 24, 2017.

INSTRUCTIONS

1. Please answer all of the questions to the best of your ability.

2. Please indicate your response by placing an ‘X’ in the appropriate boxes. The survey requests
that you rate the prospect of the elements influencing TQM in the construction industry in
Ghana. You are also requested to rate the impact of other factors in predicting TQM in the
Ghanaian construction industry.
3. The influence (probability) scale is presented, and only a number should be used for a
probability range. For instance, if you consider the influence (probability) range to be between
71 & 80% of the feature’s influence then you should mark ‘X’ under the box ‘8’.
If the impact is considered to be high, then ‘X’ should be marked under the ‘7’ or ‘8’ box
depending on whether your opinion is inclined more towards high or very high impact.(See the
attached questions).
Please use your experience, expertise and judgement to rate what you perceive the average
negative or positive influence of the various features are for TQM and Quality Management in
Ghanaian construction industry at large.

PROBABILITY SCALE (likelihood in percentage)

1-10% 11-20% 21-30% 31-40 41-50 51-60 61-70 71-80 81-90 91-100
X

478
IMPACT SCALE
No impact Low impact Medium impact High impact Very high impact
1 2 3 4 5 6 7 8 9 10

Q. 1.1 TQM MAIN ATTRIBUTES: To identify the main attributes that contribute to TQM
(TQM) in the Ghanaian construction industry.

TQM (TQM) What is the influence of the listed attributes on TQM (TQM) in
attributes the Ghanaian construction industry? (1=low probability,
10=high probability)
1 2 3 4 5 6 7 8 9 10 Rank
Leadership and Top
Management (TML)
Company Supplier
Quality Management
(CSQM)
Client Focus and
Involvement (CFI)
Company Quality
System Evaluation
(CQSE)
Company Vision and
Plan Statement
(CVPS)
Product Selection
and Design
Management
(PSDM)
Construction Process
Management and
Improvement (CPMI)
Construction
Employees’

479
 Involvement and
Motivation (CEIM)

Any other main attributes: Please specify: …………………………………………………….

…………………………………………………………………………………………………

Q. 1.2. TQM SUB-ATTRIBUTES: To identify the sub-attributes that contributes to Total

Quality Management (TQM) in the Ghanaian construction industry.

Q. 1.2.1 LEADERSHIP AND TOP MANAGEMENT FEATURES:

What is the impact of the listed leadership and top management
Leadership and Top features on TQM? (1=low probability, 10=high probability)
Management
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Leadership style
Leadership ability in
solving quality
problem
Leadership skills
Leadership initiatives
Leadership and top
management
knowledge in TQM
Top management
commitment in TQM
Top management
interaction with
workers
Top management
participation
Top management
learning

480
Top management
empowerment
Top management
encouragement
Top management’s
role model

Any other leadership and top management features? Please specify: …………………………..
………………………………………………………………………………………………….

Q. 1.2.2 COMPANY SUPPLIER QUALITY MANAGEMENT FEATURES:

What is the impact of the listed company supplier quality
Company Supplier management features on TQM?
Quality (1=low probability, 10=high probability)
Management
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Partnership with
Suppliers
Supplier Selection
Criteria
Participation in
Suppliers
Supplier Performance
Evaluation
Supplier Quality
Audit

Supplier
Communication
Supplier Knowledge
in/of? TQM
Location of Supplier
Suppliers’ Past
Records

481
Suppliers’
Commitment
Cooperation from
Suppliers
Suppliers’ Orientation
Any other company supplier quality management features? Please specify: ……………………
…………………………………………………………………………………………………..

Q. 1.2.3 CLIENT FOCUS AND INVOLVEMENT FEATURES:

Client Focus and What is the impact of the listed client focus and involvement
Involvement features on TQM? (1=low probability, 10=high probability)

No Low Medium High Very high

impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Client brief/ input
Client complaint
information/feedback
Market investigation
Client satisfaction
survey
Quality warranty
Client information
system
Client services
Client cooperation
Client orientation
Any other client focus and involvement features? Please specify: ………………………………
…………………………………………………………………………………………………..

Q. 1.2.4 COMPANY QUALITY SYSTEM EVALUATION FEATURES:

Company Quality What is the impact of the listed company quality system
System Evaluation evaluation features on TQM? (1=low probability, 10=high
probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Evaluation of strategy
Evaluation of overall
company performance
Evaluation of
departments’
performance
482
Evaluation of
employee
performance
Evaluation of quality
costs

Evaluation of quality
manual

Evaluation of quality
control system

Evaluation of quality
system procedures

Evaluation of end
results

Quality audit

Quality benchmarking

Quality information
system

Any other company quality system evaluation features? Please specify: ……………………….

Q. 1.2.5 COMPANY VISION AND PLAN STATEMENT FEATURES:

Company Vision and What is the impact of the listed company vision and plan
Plan Statement statement features on TQM? (1=low probability, 10=high
probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Vision statement

Quality policy

483
Overall business
performance plan
Product quality goal
Quality improvement
plan
Formulation of vision
and plan
Concreteness of future
plan
Employees’
contribution to the
vision
Involvement of
employees in the
development of vision
statement
Any other company vision and plan statement features? Please specify: ……………………
…………………………………………………………………………………………………..

Q. 1.2.6 PRODUCT SELECTION AND DESIGN MANAGEMENT FEATURES:

Product Selection What is the impact of the listed product selection and design
and Design management features on TQM? (1=low probability, 10=high
Management probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Client brief/input
Cost implication
Environmental issues
Socio-cultural issues
Appearance/finishes
required
Strength required
Intended purpose of
the material
Where the material to
be used
Design
methods/techniques
Design detailing
Concurrent
Engineering
Design

484
experiments/testing

Quality function
deployment
Value engineering

Computer-aided
design
Any other product selection and design management features please specify: ………………..
…………………………………………………………………………………………………..

Q. 1.2.7 CONSTRUCTION PROCESS MANAGEMENT AND IMPROVEMENT FEATURES:

Construction What is the impact of the listed construction process management
Process and improvement features on TQM? (1=low probability, 10=high
Management and probability)
Improvement
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Project monitoring
and control
improvement

Equipment
maintenance and
innovation
Inventory
management
Use of quality
materials

Use of quality tools

Use of quality manual

Understanding of
work
instructions/quality
manual

485
Quality system
procedures
improvement
Obtaining ISO 9000
certification for
operation

Understanding of
quality control system
Continuity of quality
control circle
activities

Appropriate use of
system structure and
standards

Targets and priority

measures

Safe working
environment

Utilization of analysis
results

Any other construction process management and improvement features? Please specify: ……..
…………………………………………………………………………………………………..

Q. 1.2.8 CONSTRUCTION EMPLOYEES’ INVOLVEMENT AND MOTIVATION

FEATURES:
Construction What is the impact of the listed construction employees’
Employees involvement and motivation features on TQM? (1=low probability,
Involvement and 10=high probability)
Motivation
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Education and
486
Training
Salary Promotion
Position Promotion
Employee Rewards
Bonus Scheme
Conducive Working
Environment
Involvement of
Employees
withinFunctional
Team
Employee
Participation
Employee Suggestion
Improving Employee
Commitment
Job rotation
Involvement in
Quality Control Circle
Employee
Involvement in Cross-
functional Team
Employee
Recognition
Any other construction employees involvement and motivation features? Please specify: …….
…………………………………………………………………………………………………..

Q. 2. FACTORS THAT ENABLES CONSRUCTION COMPANY TO IMPLEMENT TQM

What is the influence of the listed enable factors on
construction company to implement TQM? (1=low probability,
10=high probability)

1 2 3 4 5 6 7 8 9 10 Rank
Top management support
Good quality driven
agenda
Commitment to people
Training and re-training
of staff
Institution of continuous
improvement process
within the company
Systematic TQM
487
framework
Efficient management
system
Proper understanding of
TQM among construction
professionals
Availability of resources
Good communication in
organisation
Good quality policy
Changes in company
structure
Good organizational and
quality culture
Teamwork
Availability of quality
information
Employee involvement
Good climatic condition
Good socio-economic
environment
Proper project
conceptualization
Project manager
competence
Clarity of goals and
objectives
Adequate TQM experts

Any other enable factors? Please specify: ....................................................................................

Q.3 FACTORS THAT AFFECT CONSTRUCTION COMPANY FOR NOT IMPLEMENTING

TQM
What is the impact of the listed factors on construction company
for not implementing TQM? (1=low probability, 10=high
probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Lack of efficient
TQM management
system
Lack or limited
company resources
488
Unavailable TQM
policy
Lack or limited
knowledge of TQM
Inability to employ
TQM personnel
Absent of systematic
TQM framework
Lack and Inability to
train and educate
employees on TQM
Lack of understanding
among construction
professionals in
applying TQM
Lack of coordination
of the implementation
of TQM policy within
the organisation
Limited access to
body responsible for
the implementation of
TQM policy
Lack of TQM expert
Reluctance to change
old management
technique
Complex nature of
TQM technique
Lack of commitment
from management
Perception that TQM
may not yield any
better results
TQM technique is
time consuming
TQM technique is
costly
Lack of finance in the
management of TQM
experts
Lack of enforcement
from the legislative
bodies overseeing the
489
implementation of
TQM
Lack of interest in the
application of TQM
Any other factors? Please specify: ……………………………………………………………..

Q. 4 ORGANIZATIONAL FACTORS AND ISSUES THAT AFFECT CONSTRUCTION

COMPANIES IN THE IMPLEMENTATION OF TQM
What is the impact of the listed success factors and issues that
affect the construction companies in the implementation
of TQM? (1=low probability, 10=high probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Leadership and top
management
commitment
Top management
support
Project manager
competence
Quality policy
Supplier management
Limited cash flow to
manage TQM
Change in
management
behavior/attitude
Change in status quo
Personnel to manage
and monitor TQM
application
Education and
training
Employee relations
Employee
involvement
Lack of interest in the
application of TQM
Client involvement
Teamwork
Feedback by project

490
participants
Design quality
management
Strategic quality
management
Quality data and
reporting
Analysis of quality
information
Any other factors? Please specify: ………………………………………………………………
………………………………………………………………………………………………….

Q 5 EFFECTS OF IMPLEMENTING TQM IN THE CONSTRUCTION INDUSTRY

What is the effect of the listed factors on construction industry
for the implementation of TQM? (1=low probability, 10=high
probability)
1 2 3 4 5 6 7 8 9 10 Rank
Reduction in
construction costs
Higher client
satisfaction
Improve employee job
satisfaction
Improve schedule
performance
Elimination/Reduction
of reworks
Defect-free product at
first attempt
Improve safety
Higher productivity
Lower employee
turnover
Speed up construction
work
Improve relationships
with subcontractors
Improve methods of
working
Better control over the
construction process
Increase profitability
491
Gaining competitive
advantage over other
companies
Decreasing waste
Better coordination of
activities
Any other effects? Please specify: ……………………………………………………………..

…………………………………………………………………………………………………

Q. 6. PERSONAL INFORMATION OF EXPERT PANEL MEMBERS

Title (Mr, Mrs, Ms, Rev, Dr, Prof)

Email:
Highest qualification
Field of specialisation
Professional registration (ICIOB, MCIOB, FCIOB,
MGIOC, FGIOC, MGhIS, FGhIS, etc.)
Years of experience (TQM)
Current employer

Position

Region

Thank you for taking your time to fill out this first round survey. The second round of the Delphi process will be a
follow-up email.
Please do not hesitate to contact me, my promoter, Prof. Wellington Didibhuku Thwala, or my co-promoter, Prof.
C.O. Aigbavboa,if you have any questions about this survey or the research project in general. Kindly see contact
details below.

Contact details:
Samuel Kwame Ansah, Ph.D. Candidate
Department of Construction Management and Quantity Surveying
Faculty of Engineering and the Built Environment
University of Johannesburg, South Africa
Ghanaian mobile number: +233246769673
South African mobile number: +277837420742
Email: [email protected]

Promoter
Prof. C.O. Aigbavboa (Ph.D.)
Department of Construction Management & Quantity Surveying
Faculty of Engineering and the Built Environment
University of Johannesburg, South Africa
Tel: +27 (0) 11 559 6398

492
Mobile: +27 78 795 8231
Email: [email protected]
Associate Editor - Journal of Construction Project Management and Innovation (JCPMI)

Co-Promoter
Prof. Wellington Didibhuku Thwala Ph.D. (Eng)
Professor of Construction Project Management,
Department of Construction Management and Quantity Surveying
Faculty of Engineering and the Built Environment
University of Johannesburg, South Africa
Tel: +27 (0)11 559 6048
Fax: +27 (0) 11 559 6630
Mobile: +27 83 383 5537
Email: [email protected]

493
 APPENDIX E

DELPHI INSTRUCTIONS FOR ROUND TWO QUESTIONNAIRE

DELPHI SURVEY – ROUND 2

Thank you for completing Round 1 of the Delphi survey. We recognise that the survey required a
significant time investment to complete thoughtfully. We appreciate your time and effort. The
purpose of Round 2 is to provide you with the opportunity to change your response, if desired,
given the group median response for each question.

When you have finished answering all of the questions, please kindly email your response to
[email protected] by Friday, April 21, 2017.

INSTRUCTIONS
For each element you will see two (2) values: your response from the Round 1 survey (indicated
with a yellow highlighted box), and the group median from the Round 1 survey indicated in the
column to the far right hand of each table. Please take one of the following three actions for each
category:
1. Accept the group median response by leaving the field completely unchanged.
2. Maintain your original response by placing an ‘X’ in the highlighted field*.
3. Indicate a new response by placing an ‘X’ in the appropriate field*.

* *If your response is more than ten per cent (one unit) above or below the group median, please
provide a reason for your outlying response in the field provided.

Q. 1.1 TQM MAIN ATTRIBUTES: To identify the main attributes that contributes to TQM
(TQM) in the Ghanaian construction industry.

TQM (TQM) What is the influence of the listed attributes on TQM (TQM) in
attributes the Ghanaian construction industry? (1=low probability,
10=high probability)
1 2 3 4 5 6 7 8 9 10 Rank
Leadership and Top X 10
Management (TML)

494
 Company Supplier X 9
Quality Management
(CSQM)
Client Focus and X 8
Involvement (CFI)
Company Quality X 9
System Evaluation
(CQSE)
Company Vision and X 9
Plan Statement
(CVPS)
Product Selection X 8
and Design
Management
(PSDM)
Construction Process X 9
Management and
Improvement (CPMI)
Construction X 10
Employees
Involvement and
Motivation (CEIM)
Any Comment:

495
Q. 1.2. TQM SUB-ATTRIBUTES: To identify the sub-attributes that contributes to TQM
(TQM) in the Ghanaian construction industry.

Q. 1.2.1 LEADERSHIP AND TOP MANAGEMENT FEATURES:

What is the impact of the listed leadership and top management
Leadership and Top features on TQM? (1=low probability, 10=high probability)
Management
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Leadership style X 9
Leadership ability in X 9
solving quality
problem
Leadership skills X 9
Leadership initiatives X 9
Leadership and top X 10
management
knowledge in TQM
Top management X 9
commitment in TQM
Top management X 9
interaction with
workers
Top management X 9
participation
Top management X 8
learning
Top management X 8
empowerment
Top management X 9
encouragement
Top management’s X 8
role model
496
Any Comment:

Q. 1.2.2 COMPANY SUPPLIER QUALITY MANAGEMENT FEATURES:

What is the impact of the listed company supplier quality
Company Supplier management features on TQM?
Quality (1=low probability, 10=high probability)
Management
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Partnership with X 8
Suppliers
Supplier Selection X 8
Criteria
Participation in X 8
Suppliers
Supplier Performance X 9
Evaluation
Supplier Quality X 9
Audit

Supplier X 9
Communication
Supplier Knowledge X 9
in TQM
Location of Supplier X 6
Suppliers’ Past X 9
Records
Suppliers’ X 8
Commitment
Cooperation from X 9
Suppliers
Suppliers’ Orientation X 7
Any Comment:

497
Q. 1.2.3 CLIENT FOCUS AND INVOLVEMENT FEATURES:
Client Focus and What is the impact of the listed client focus and involvement
Involvement features on TQM? (1=low probability, 10=high probability)

No Low Medium High Very high

impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Client brief/ input X 9
Client complaint X 9
information/feedback
Market investigation X 8
Client satisfaction X 9
survey
Quality warranty X 8
Client information X 7
system
Client services X 7
Client cooperation X 8
Client orientation X 7
Any Comment:

Q. 1.2.4 COMPANY QUALITY SYSTEM EVALUATION FEATURES:

Company Quality What is the impact of the listed company quality system
System Evaluation evaluation features on TQM? (1=low probability, 10=high
probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Evaluation of strategy X 8
Evaluation of overall X 9
company performance
Evaluation of X 9
departments’
performance
Evaluation of X 8
employee
performance
Evaluation of quality X 8
costs

Evaluation of quality X 8

498
manual

Evaluation of quality X 9
control system

Evaluation of quality X 9
system procedures

Evaluation of end X 9
results

Quality audit X 9

Quality benchmarking X 9

Quality information X 9
system

Any Comment:

Q. 1.2.5 COMPANY VISION AND PLAN STATEMENT FEATURES:

Company Vision and What is the impact of the listed company vision and plan
Plan Statement statement features on TQM? (1=low probability, 10=high
probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Vision statement X 8

Quality policy X 9

Overall business x 8
performance plan
Product quality goal X 8
Quality improvement X 9
plan
Formulation of vision X 8
and plan
Concreteness of future X 8
plan

499
Employees’ X 8
contribution to the
vision
Involvement of X 8
employees in the
development of vision
statement
Any Comment:x

Q. 1.2.6 PRODUCT SELECTION AND DESIGN MANAGEMENT FEATURES:

Product Selection What is the impact of the listed product selection and design
and Design management features on TQM? (1=low probability, 10=high
Management probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Client brief/input X 9
Cost implication x 8
Environmental issues X 7
Socio cultural issues X 8
Appearance/finishes X 9
required
Strength required X 9
Intended purpose of x 9
the material
Where the material to X 7
be used
Design X 8
methods/techniques
Design detailing X 9
Concurrent X 8
Engineering
Design X 8
experiments/testing

Quality function X 9
deployment
Value engineering X 9

Computer-aided X 7
design
Any Comment:
500
Q. 1.2.7 CONSTRUCTION PROCESS MANAGEMENT AND IMPROVEMENT FEATURES:
Construction What is the impact of the listed construction process management
Process and improvement features on TQM? (1=low probability, 10=high
Management and probability)
Improvement
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Project monitoring X 9
and control
improvement

Equipment X 8
maintenance and
innovation
Inventory X 7
Management
Use of quality X 9
materials

Use of quality tools X 8

Use of quality manual X 8

Understanding of x 9
work
instructions/quality
manual

Quality system X 8
Procedures
improvement
Obtaining ISO 9000 X 8
certification for
operation

Understanding of X 8
quality control system
501
Continuity of quality X 9
control circle
activities

Appropriate use of X 8
system structure and
standards

Targets and priority X 8

measures

Safe working X 9
environment

Utilization of analysis X 8
results

Any Comment:

Q. 1.2.8 CONSTRUCTION EMPLOYEES INVOLVEMENT AND MOTIVATION

FEATURES:
Construction What is the impact of the listed construction employees’
Employees involvement and motivation features on TQM? (1=low probability,
Involvement and 10=high probability)
Motivation
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Education and X
Training
Salary Promotion X
Position Promotion X
Employee Rewards X
Bonus Scheme X
Conducive Working X
Environment
Involvement of X
Employees
withinFunctional
502
Team
Employee X
Participation
Employee Suggestion X
Improving Employee X
Commitment
Job Rotation X
Involvement in X
Quality Control Circle
Employee X
Involvement in Cross-
functional Team
Employee X
Recognition
Any Comment:

Q. 2. FACTORS THAT ENABLES CONSRUCTION COMPANY TO IMPLEMENT TQM

What is the influence of the listed enable factors on
construction company to implement TQM? (1=low probability,
10=high probability)

1 2 3 4 5 6 7 8 9 10 Rank
Top management support X
Good quality driven X
agenda
Commitment to people X
Training and re-training X
of staff
Institution of continuous X
improvement process
within the company
Systematic TQM X
framework
Efficient management X
system
Proper understanding of X
TQM among construction
professionals
Availability of resources X
Good communication in x
organisation
Good quality policy x
Changes in company X
503
structure
Good organizational and x
quality culture
Teamwork X
Availability of quality X
information
Employee involvement X
Good climatic condition X
Good socio-economic X
environment
Proper project X
conceptualization
Project manager X
competence
Clarity of goals and X
objectives
Adequate TQM experts X

Any Comment:

Q.3 FACTORS THAT AFFECT CONSTRUCTION COMPANY FOR NOT IMPLEMENTING

TQM
What is the impact of the listed factors on construction company
for not implementing TQM? (1=low probability, 10=high
probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Lack of efficient X
TQM management
system
Lack or limited X
company resources
Unavailable TQM X
policy
Lack or limited X
knowledge of TQM
Inability to employ X
TQM personnel
Absent of systematic X
TQM framework
Lack and inability to x
train and educate
504
employees on TQM
Lack of understanding X
among construction
professionals in
applying TQM
Lack of coordination X
of the implementation
of TQM policy within
the organisation
Limited access to X
body responsible for
the implementation of
TQM policy
Lack of TQM expert X
Reluctance to change X
old management
technique
Complex nature of X
TQM technique
Lack of commitment X
from management
Perception that TQM X
may not yield any
better results
TQM technique is X
time consuming
TQM technique is X
costly
Lack of finance in the X
management of TQM
experts
Lack of enforcement X
from the legislative
bodies overseeing the
implementation of
TQM
Lack of interest in the X
application of TQM
Any Comment:

505
Q. 4 ORGANIZATIONAL FACTORS AND ISSUES THAT AFFECT CONSTRUCTION
COMPANIES IN THE IMPLEMENTATION OF TQM
What is the impact of the listed success factors and issues that
affect the construction companies in the implementation
of TQM? (1=low probability, 10=high probability)
No Low Medium High Very high
impact impact impact impact impact
1 2 3 4 5 6 7 8 9 10 Rank
Leadership and top X
management
commitment
Top management X
support
Project manager X
competence
Quality policy X
Supplier management X
Limited cash flow to X
manage TQM
Change in X
management
behavior/attitude
Change in status quo X
Personnel to manage X
and monitor TQM
application
Education and X
training
Employee relations X
Employee X
involvement
Lack of interest in the X
application of TQM
Client involvement X
Teamwork X
Feedback by project X
participants
Design quality X
management
Strategic quality X
management
Quality data and X
reporting

506
Analysis of quality X
information
Any Comment:

Q 5 EFFECTS OF IMPLEMENTING TQM IN THE CONSTRUCTION INDUSTRY

What is the effect of the listed factors on construction industry
for the implementation of TQM? (1=low probability, 10=high
probability)
1 2 3 4 5 6 7 8 9 10 Rank
Reduction in X
construction costs
Higher client X
satisfaction
Improve employee job X
satisfaction
Improve schedule X
performance
Elimination/Reduction X
of reworks
Defect-free product at X
first attempt
Improve safety X
Higher productivity X
Lower employee X
turnover
Speed up construction X
work
Improve relationships X
with subcontractors
Improve methods of X
working
Better control over the X
construction process
Increase profitability X
Gaining competitive X
advantage over other
companies
Decreasing waste X
Better coordination of X
activities
Any Comment:

507
Q. 6. PERSONAL INFORMATION OF EXPERT PANEL MEMBERS

Title (Mr, Mrs, Ms, Rev, Dr, Prof)

Email:
Highest qualification Master’s
Field of specialisation Construction Management
Professional registration (ICIOB, MCIOB, FCIOB, MGIOC,MIET
MGIOC, FGIOC, MGhIS, FGhIS, etc.)
Years of experience (TQM) 7

Current employer MAKDOSS GH

Position PROJECTS MANAGER

Region WESTERN

Please do not hesitate to contact me or my promoter and co-promoter Prof. C.O. Aigbavboa and Prof. Wellington
Didibhuku Thwala if you have any questions about this survey or the research project in general. Kindly see contact
details below.

Contact details:
Samuel Kwame Ansah, Ph.D. Candidate
Department of Construction Management and Quantity Surveying
Faculty of Engineering and the Built Environment
University of Johannesburg, South Africa
Ghana Mobile Number: +233246769673
South Africa Mobile Number: +277837420742
Email: [email protected]

Promoter
Prof. C.O. Aigbavboa (Ph.D.)
Department of Construction Management & Quantity Surveying
Faculty of Engineering and the Built Environment
University of Johannesburg, South Africa
Tel: +27 (0) 11 559 6398
Mobile: +27 78 795 8231
Email: [email protected]
Associate Editor - Journal of Construction Project Management and Innovation (JCPMI)

Co-Promoter
Prof. Wellington Didibhuku Thwala Ph.D. (Eng)
Professor of Construction Project Management,
Department of Construction Management and Quantity Surveying
Faculty of Engineering and the Built Environment
University of Johannesburg, South Africa
Tel: +27 (0)11 559 6048
Fax: +27 (0) 11 559 6630
Mobile: +27 83 383 5537
Email: [email protected]

508
 APPENDIX F
Research Introduction Letter and Questionnaire

UNIVERSITY OF JOHANNESBURG

FACULTY OF ENGINEERING AND THE BUILT ENVIRONMENT

Dear Respondent

The Postgraduate School of Engineering at the University of Johannesburg is undertaking a

research project to develop an integrated TQM (TQM) model for the Ghanaian construction
indudstry. To this end, we kindly request that you complete the following short questionnaire. It
should take no longer than 20 minutes of your time. Your response is of the utmost importance to
us.

To protect your anonymity, please do not enter your name or contact details on the questionnaire.
Summary results of this research will be available in the Department of Construction
Management and Quantity Surveying in November 2018.

Should you have any queries or comments regarding this survey, you are welcome to contact us
telephonically at +233246769673, +27837420742, or email us at [email protected]

Yours sincerely

Ansah, S.K

University of Johannesburg
Tel: + 27 11 5593284
Mobile: +233246769673, or + 27837420742

509
QUESTIONNAIRE TO EVALUATE TQM IMPLEMENTATION IN THE GHANAIAN
CONSTRUCTION INDUSTRY

Please answer the following questions by crossing (x) in the relevant box or writing in the
space provided.

SECTION A: PROFILE OF THE RESPONDENT AND THE FIRM

1. What is your gender? 1 2

Male ☐ Female ☐

2. Which age category do you fall into?

15 - 20 21 - 25 26 -30 31 - 35 36 - 40 41- 45 46 years

years years years years years years and above

      

1 2 3 4 5 6 7

3. What is your highest qualification?

Qualification Tick One

National Diploma/National Certificate  1

Bachelor’s Degree (First Degree)  2

Postgraduate – Master’s Degree (MSc., MPhil.,etc)  3

Postgraduate - Doctoral Degree (PhD)  4

510
4. For how many years have you worked in the construction industry?

1-5 years 6 – 10 11 –15 16 - 20 21 – 25 26 - 30 31years and

above

      

1 2 3 4 5 6 7

5. What is your position in the firm?

Project Site Quality Supervis Forema Other (Please specify)

Manager Enginee Control or n
r Officer
……………………………
………

     

1 2 3 4 5 6

6. How long have you held your current position?

1–5 years 6 – 10 11 –15 16 – 21 – 25 26 – 31 or more

year years 20 years 30 years
years years

      

1 2 3 4 5 6 7

511
7. How long has your firm been in existence?

1– 5 6 – 10 11 –15 16 – 21 – 25 26 – 30 31 or more
years year years 20 years years years
years

      

1 2 3 4 5 6 7

8. What type of construction project is your firm currently working on?

Building Construction Civil Engineering Other (Please specify)

……………………………………………….

  

1 2 3

9. Please, indicate your firm’s classification.

D1/K1 D2/K2 D3/K3 D4/K4

   

1 2 3 4

512
SECTION B: LEVEL OF TQM IMPLEMENTATION IN THE GHANAIAN
CONSTRUCTION INDUSTRY

10. To what extent has your company implemented TQM? Please mark your choice of response
with an [X]

A. To no extent [ ]

B. To a small extent [ ]

C. To a moderate extent [ ]

D. To a large extent [ ]

E. Completely [ ]

SECTION C: ATTRIBUTES THAT COULD LEAD TO THE IMPLEMENTATION OF TQM

IN THE GHANAIAN CONSTRUCTION INDUSTRY.

Below is a list of attributes that could lead to the implementation of TQM in the Ghanaian
Construction Industry. Based on your experience and using the scale provided, please
indicate the degree of influence each attribute has on TQM being implemented in the
Ghanaian Construction Industry.

1 = Not at all influential; 2 = Slightly influential; 3 = Moderately influential; 4 = Very

influential; 5 = Extremely influential

11. LEADERSHIP AND TOP MANAGEMENT FEATURES (LTM)

1 2 3 4 5

Code To what extent do the following

influential
Not
influential
Slightly

influential
Moderately
influential
Very
influential
Extremely

attributes influence the

at

implementation of TQM
all

513
LTM 1 Leadership style of managing
employees

LTM 2 Leadership ability in solving

quality-related problems

LTM 3 Leadership initiatives towards

TQM

LTM 4 Top management knowledge

and proper understanding of
TQM

LTM 5 Top management commitment

to TQM

LTM 6 Top management interaction

with workers

LTM 7 Top management participation

in TQM activities

LTM 8 Top management empowerment

of employees to solve quality
problems

LTM 9 Top management support of

TQM

514
12. COMPANY SUPPLIER QUALITY MANAGEMENT FEATURES (CSQM)

1 2 3 4 5

Code To what extent do the

Not at all influential

Slightly influential
influential
Moderately

Very influential
influential
Extremely
following attributes
influence the implementation
of TQM

CSQM1 Partnership with suppliers

CSQM 2 Supplier selection criteria

CSQM 3 Participation of suppliers in

TQM activities

CSQM 4 Supplier performance

evaluation

CSQM 5 Supplier quality audit

CSQM 6 Supplier communication

CSQM 7 Supplier knowledge of TQM

CSQM 8 Suppliers’ commitment to

TQM

CSQM 9 Suppliers’ orientation on

TQM

515
13. CLIENT FOCUS AND INVOLVEMENT FEATURES (CFI)

1 2 3 4 5

Code To what extent do the following

influential
Not
influential
Slightly
influential
Moderately
influential
Very
influential
Extremely
attributes influence the

at
implementation of TQM

all
CFI 1 Client brief/ input

CFI 2 Client complaint

information/feedback

CFI 3 Market investigation

CFI 4 Client satisfaction survey

CFI 5 Quality warranty to clients

CFI 6 Client information system

CFI 7 Client services

CFI 8 Client cooperation

14. COMPANY QUALITY SYSTEM EVALUATION FEATURES (CQSE)

1 2 3 4 5

Code To what extent do the

influential
Not
influential
Slightly
influential
Moderately
influential
Very
influential
Extremely

following attributes influence

at

the implementation of TQM

all

CQSE 1 Evaluation of the quality

516
 strategy framework

CQSE 2 Evaluation of overall

company performance

CQSE 3 Evaluation of employee

performance

CQSE 4 Evaluation of quality costs

CQSE 5 Evaluation of the quality

manual

CQSE 6 Evaluation of the quality

control system

CQSE 7 Evaluation of quality system

procedures

CQSE 8 Evaluation of end results

CQSE 9 Evaluation of the quality

information system

15. COMPANY VISION AND PLAN STATEMENT FEATURES (CVPS)

1 2 3 4 5

Code To what extent do the

influential
Not
influential
Slightly
influential
Moderately
influential
Very
influential
Extremely

following attributes influence

at

the implementation of TQM

all

517
 CVPS 1 Clarity of vision and plan
statement

CVPS 2 Good quality policy and driven

agenda

CVPS 3 Quality improvement plan

CVPS Formulation of vision and plan

4

CVPS Concreteness of the future plan

5

CVPS Employee contribution to the

6 vision

CVPS Involvement of employees in

7 the development of the vision
statement

16. PRODUCT SELECTION AND DESIGN MANAGEMENT (PSDM)

1 2 3 4 5

Code To what extent do the

influential
Not
influential
Slightly
influential
Moderately
influential
Very
influential
Extremely

following attributes influence

at

the implementation of TQM

all

PSDM 1 Client brief/input

PSDM 2 Cost of selection and design

of the product

518
PSDM 3

Environmental issues

PSDM 4 Socio-cultural issues

PSDM 5 Appearance/finishes required

PSDM 6 Strength required for the end

product

PSDM 7 Intended purpose of the

material

PSDM 8 Where the material to be used

PSDM 9 Design methods/techniques

PSDM 10 Design detailing

17. CONSTRUCTION PROCESS MANAGEMENT AND IMPROVEMENT FEATURES

(CPMI)

1 2 3 4 5

Code To what extent do the

influential
Not
influential
Slightly
influential
Moderately
influential
Very
influential
Extremely

following attributes influence

at

the implementation of TQM

all

CPMI 1 Project monitoring and

control improvement

CPMI 2 Equipment maintenance and

innovation

519
CPMI 3 Inventory management

CPMI 4 Use of quality materials

CPMI 5 Use of quality tools

CPMI 6 Use of a quality manual

CPMI 7 Understanding of work

instructions/quality manual

CPMI 8 Obtaining ISO 9000

certification for operation

CPMI 9 Quality control system

CPMI 10 Institution of a continuous

improvement process within
the company

CPMI11 Appropriate use of system

structure and standards

CPMI12 Efficient management system

520
18. CONSTRUCTION EMPLOYEES INVOLVEMENT AND MOTIVATION FEATURES
(CEIM)

1 2 3 4 5

Code To what extent do the

influential
Not
influential
Slightly
influential
Moderately
influential
Very
influential
Extremely
following attributes influence

at
the implementation of TQM

all
CEIM 1 Education and training/re-
training of staff
CEIM 2 Teamwork

CEIM 3 Salary/position promotion

CEIM 4 Employee rewards

CEIM 5 Bonus scheme

CEIM 6 Conducive working

environment
CEIM 7 Employee involvement in
TQM activities
CEIM 8 Availability of a suggestion
forum for employees
CEIM 9 Employee commitment

CEIM10 Recognition of employees

SECTION D: PROJECT OUTCOMES THAT COULD MOTIVATE A COMPANY TO

IMPLEMENT TQM

19. Below is a list of positive project outcomes which could motivate a Construction
company to implement TQM. Please use the scale provided to rate the degree to which each
factor motivated your company to implement TQM.
521
1=Not a motivating factor; 2=Low motivating factor; 3=Moderate motivating factor; 4=High
motivating factor; 5= Very high motivating factor

To what extent each of the

following project outcomes
Code 1 2 3 4 5
motivated your company to
implement TQM

g factor
motivatin

g factor a
Not

motivatin
gLow
motivatin
gModerate
motivatin
High
g factor
motivatin
Very high
factor

factor
19.1 Greater reduction of
construction costs

19.2 Higher product quality

19.3 Improved schedule

performance

19.4 Elimination of reworks

19.5 Defect-free product at first

attempt

19.6 Higher safety standards

19.7 Higher productivity

19.8 Good methods of working

19.9 Higher profitability

19.10 Higher reduction of

construction waste

522
SECTION E: REASONS FOR NON-IMPLEMENTATION (ADOPTION) OF TQM IN
GHANAIAN CONSTRUCTION INDUSTRY

20. Below is a list of factors that could lead to the non-implementation of TQM in the
Ghanaian Construction Industry. Using the scale provided, please indicate your level of
agreement for each factor if they lead to Non-implementation of TQM in the Ghanaian
Construction Industry.
1 = Strongly disagree (SD); 2 = Disagree (D); 3 = Neutral (N); 4 = Agree (A); 5 = Strongly
Agree (SA)

To what extent do you agree that the 1 2 3 4 5

following factors are the reason why
D N A SA
construction companies in Ghana do not
SD
adopt TQM

20.1 Lack of efficient TQM management system

20.2 Absent of TQM policy

20.3 Lack of or limited knowledge of TQM

20.4 Lack of and inability to train and educate

employees on TQM

20.5 Lack of understanding among construction

professionals in applying TQM

20.6 Lack of coordination of the implementation

of TQM policy within the organization

19.7 Lack of TQM expert

523
 20,8 Reluctance to change old management
technique

20,9 Complex nature of TQM technique

20,10 Lack of commitment from management

20.11 Perception that TQM may not yield any

better results

20.12 TQM technique is time consuming

20.13 TQM technique is costly

20.14 Lack of finance in the management of TQM

experts

20.15 Lack of enforcement from the legislative

bodies overseeing the implementation of
TQM

20.16 Lack of interest in the application of TQM

SECTION F: PERFORMANCE OF CONSTRUCTION COMPANIES IN GHANA

21. Below is a list of performance indicators for a construction company. Please use the scale
provided to rate your company’s performance of each indicator over the past year.

1 = Very poor; 2 = Poor; 3 = Average; 4 = Good; 5 = Excellent

Rate how well your company

524
 Code performed for each of the following 1 2 3 4 5
performance indicators

poor
Very

Poor

Average

Good

Excellent
21.1 Construction costs

21.2 Product quality

21.3 Schedule performance

21.4 Reworks elimination

21.5 Defect-free product at first attempt

21.6 Safety standards

21.7 Productivity levels

21.8 Methods of working

21.9 Degree of control over the

construction process

21.10 Amount of waste produced

Thank you for your contribution. We value your contribution and time spent in completing this
questionnaire. If you have any queries, please do not hesitate to contact the undersigned.

Samuel Kwame Ansah

University of Johannesburg

Tel: +27 11 559 3284

Mobile: +27837420742, Email: [email protected]

525
 APPENDIX G
Model 2.0 covariance matrix (Unstandardized)

PSDM10 PSDM7 PSDM6 CEIM5 CEIM4 CEIM3 CPMI11

PSDM10 0
PSDM7 -0.011 0
PSDM6 0.01 0.003 0
CEIM5 0.036 0.001 -0.008 0
CEIM4 0.028 0.004 -0.06 0.007 0
CEIM3 0.122 0.029 0.002 -0.056 0.004 0
CPMI11 0.006 0.018 -0.027 -0.023 0.042 0.09 0
CPMI10 -0.022 0.012 0.004 -0.062 -0.043 -0.039 0.021
CPMI9 0.037 0.104 0.001 0.016 -0.033 -0.013 0.003
CPMI12 -0.029 0.002 -0.09 0.028 0.021 0.032 0.01
D19.6 0.064 0.026 -0.01 0.067 0.016 0.083 0.01
D19.5 0.047 0.024 0.036 -0.004 -0.034 0.002 -0.058
D19.4 -0.017 0.001 -0.004 0.031 0.036 0.037 0.04
D19.1 -0.006 -0.065 -0.066 0.011 -0.054 -0.008 -0.03
PSDM5 0.05 0.096 0.082 -0.029 -0.021 0.005 -0.049
PSDM2 0.027 -0.021 -0.004 -0.01 0.001 0.077 -0.005
PSDM1 0.054 -0.081 -0.036 -0.013 -0.005 0.066 0.006
CVPS7 0.085 -0.032 -0.046 0.056 0.023 0.062 0.086
CVPS6 0.066 -0.042 -0.063 0.034 -0.017 0.068 -0.014
CVPS5 0.095 0.019 -0.026 0.048 -0.058 0.071 -0.059
CFI4 -0.017 0.037 -0.047 -0.035 0.064 0.114 0.01
CFI2 0.017 -0.052 -0.003 -0.079 0.01 0.024 0.045
CFI1 -0.014 0.04 0.018 -0.099 -0.032 0.007 -0.04
CQSE3 0.022 -0.007 -0.024 0.148 0.013 -0.001 -0.019
CQSE2 0.006 -0.034 0.029 0.058 -0.077 -0.03 0.021
CQSE1 0.018 -0.012 0.024 0.087 0.015 0.045 0.024
CSQM8 -0.04 0.052 -0.027 -0.012 0.002 0.006 0.044
CSQM7 -0.037 0.047 -0.03 0.04 -0.028 -0.027 0.058
CSQM4 -0.056 0.034 -0.018 0.042 0.03 0.066 0.009
LTM9 -0.031 0.007 0.046 0.109 0.037 -0.016 0.022
LTM8 -0.004 0.037 0.027 0.042 0.016 0.03 0.031
LTM4 0.006 0.062 0.036 -0.038 -0.03 -0.017 0.026
LTM1 0.011 -0.001 0.041 -0.032 -0.022 -0.032 0.017

526
CPMI10 CPMI9 CPMI12 D19.6 D19.5 D19.4 D19.1 PSDM5 PSDM2

0
-0.009 0
0.011 0.012 0
-0.004 -0.011 0.005 0
-0.014 -0.018 -0.036 0.021 0
0.067 0.055 -0.031 -0.03 -0.006 0
0.008 0.011 0.022 -0.022 0.042 -0.001 0
0.024 0.043 0.013 0.036 -0.001 -0.018 0.027 0
-0.047 0.067 0.01 0.006 -0.045 -0.019 0.063 -0.021 0
-0.079 0.028 0.01 0.011 -0.039 -0.011 0.043 -0.007 0.022
-0.01 0.011 -0.024 0.034 -0.001 0.02 -0.037 0.07 0.002
0.031 0.039 0.02 0.04 0.024 0.041 -0.064 -0.005 -0.039
0.013 0.023 0.045 -0.003 -0.018 -0.031 -0.046 0.106 -0.063
0.044 -0.025 0.051 0.041 -0.054 -0.004 0.02 0.057 -0.008
0.058 -0.032 -0.049 0.012 -0.009 0.031 0.011 -0.008 -0.027
0.011 0.038 -0.076 0.013 -0.051 0.028 -0.035 0.051 -0.006

0.048 0.033 0.055 -0.041 -0.017 0.01 -0.005 0.078 -0.008

0.001 -0.037 -0.008 0.005 0.008 -0.067 0.052 0.003 0.017

0.011 -0.013 -0.003 0.018 0.034 -0.029 0.048 0.046 0.025
-0.046 0.035 0.011 -0.048 0.022 -0.031 -0.044 0.05 -0.02
-0.009 0.043 0.017 0.04 -0.032 -0.02 0.019 0.047 0.001
-0.013 0.03 0.063 0.05 0.011 -0.002 0.064 0.082 0.001
0.008 0.026 0.014 -0.072 -0.017 0.03 -0.002 -0.013 0.051
-0.018 0.042 0.031 0.014 0.011 0.017 0.041 0.062 0.08
-0.013 0.016 -0.081 0.034 -0.016 0.02 0.054 -0.023 -0.003
-0.027 0.008 -0.051 0.002 -0.035 -0.05 0.01 -0.003 0.002

527
PSDM1 CVPS7 CVPS6 CVPS5 CFI4 CFI2 CFI1 CQSE3 CQSE2

0
0.073 0
-0.023 0.009 0
-0.062 -0.008 0 0
0.051 0.025 0.012 0.044 0
-0.044 -0.038 0.027 -0.045 -0.01 0
-0.014 -0.002 0.006 -0.001 0.002 0.004 0
-0.021 0.055 0.005 0.109 0.088 -0.005 -0.023 0
-0.079 -0.085 -0.047 -0.019 -0.031 0.044 -0.047 -0.002 0
-0.027 0.01 -0.035 0.026 -0.025 0.016 0.004 -0.019 0.017
-0.035 0.011 0.014 0.006 0.019 0.002 -0.014 0.02 -0.028
-0.052 -0.011 -0.002 0.014 0.025 -0.014 -0.022 0.063 0.023
-0.021 0.016 -0.09 0.035 0.025 0.018 0.011 -0.006 -0.025
-0.064 0.013 0.024 0.02 -0.013 0.01 -0.055 0.094 0.019
0.081 0.031 0.077 0.042 -0.012 0.02 0.001 -0.009 0.029
-0.086 0.047 -0.01 -0.067 -0.078 0.084 -0.006 -0.109 -0.04
-0.037 0.024 -0.049 -0.058 -0.085 0.049 0.003 -0.013 -0.034

CQSE1 CSQM8 CSQM7 CSQM4 LTM9 LTM8 LTM4 LTM1

0
-0.046 0
-0.021 0.014 0
0.004 -0.01 -0.011 0
0.058 0.028 0.006 0.059 0
0.007 -0.006 -0.033 0.015 -0.01 0
-0.036 -0.117 -0.068 0.044 -0.006 0.017 0
0.031 -0.005 0.025 0.11 0.003 -0.025 0.025 0

Average absolute residual = 0.0025

Average off-diagonal absolute residual = 0.0814

% falling between -0.1 +0.1 = 99.99%

528
 APPENDIX H:
Model 2.0 covariance matrix (Standardized)

PSDM10 PSDM7 PSDM6 CEIM5 CEIM4 CEIM3 CPMI11

PSDM10 0
PSDM7 -0.339 0
PSDM6 0.343 0.079 0
CEIM5 1.116 0.02 -0.24 0
CEIM4 0.854 0.097 -1.746 0.143 0
CEIM3 4.177 0.915 0.077 -1.402 0.088 0
CPMI11 0.216 -0.567 -0.898 -0.648 1.122 2.788 0
CPMI10 -0.746 0.374 0.131 -1.651 -1.093 -1.156 -0.597
CPMI9 1.264 3.157 0.043 0.424 -0.855 -0.399 -0.085
CPMI12 -0.968 0.074 -2.917 0.746 0.544 0.979 0.284
D19.6 2.482 0.897 -0.39 -2.075 0.493 2.892 0.36
D19.5 1.558 -0.697 1.146 -0.115 -0.857 0.057 -1.698
D19.4 -0.546 -0.015 -0.131 0.794 0.887 1.053 1.136
D19.1 -0.177 -1.624 -1.744 0.237 -1.141 -0.208 0.741
PSDM5 1.717 2.994 2.694 0.819 -0.576 0.16 -1.585
PSDM2 0.902 -0.615 -0.117 -0.264 0.023 2.322 -0.14
PSDM1 -1.609 -2.181 -1.027 -0.321 -0.126 1.79 0.165
CVPS7 2.894 -0.99 -1.503 1.557 0.632 1.936 -2.714
CVPS6 1.961 -1.125 -1.806 0.834 -0.404 1.865 -0.381
CVPS5 2.905 0.52 -0.763 1.2 -1.422 1.977 -1.66
CFI4 -0.529 1.066 -1.436 -0.908 1.663 3.362 0.3
CFI2 0.578 -1.582 -0.094 -2.247 0.269 0.747 1.381
CFI1 -0.465 1.185 0.588 -2.746 -0.856 0.212 -1.188
CQSE3 0.745 -0.206 -0.752 4.084 0.345 -0.026 -0.593
CQSE2 0.207 -1.122 1.007 1.761 -2.294 -1.027 -0.705
CQSE1 0.661 -0.385 0.825 2.631 0.458 1.512 -0.804
CSQM8 -1.17 1.353 -0.763 -0.302 0.059 0.153 -1.19
CSQM7 -1.091 1.243 -0.855 1.002 -0.683 -0.755 -1.599
CSQM4 -1.646 0.889 -0.5 1.025 0.711 1.812 0.234
LTM9 -0.937 -0.188 1.358 2.736 0.902 0.452 0.627
LTM8 -0.126 -1.075 0.827 1.109 0.42 0.892 0.917
LTM4 0.18 -1.648 1.01 -0.93 -0.701 -0.466 0.696
LTM1 0.371 -0.033 1.292 -0.868 -0.562 -0.949 -0.489

529
CPMI10 CPMI9 CPMI12 D19.6 D19.5 D19.4 D19.1 PSDM5 PSDM2 PSDM1

0
-0.245 0
0.303 0.349 0
-0.149 -0.37 0.181 0
-0.392 -0.524 -1.018 0.666 0
1.825 1.501 -0.859 -0.911 -0.151 0
0.185 0.263 0.531 -0.583 0.938 -0.013 0
0.728 1.325 0.399 1.277 -0.026 -0.525 0.67 0
-1.376 1.996 0.311 0.21 -1.28 -0.537 1.529 -0.56 0
-2.097 0.76 0.269 0.339 -0.996 -0.264 0.923 -0.166 0.51 0
-0.313 0.336 -0.741 1.197 -0.016 0.566 -0.91 2.182 0.05 1.964
0.82 1.038 0.531 1.192 0.613 1.01 -1.375 -0.129 -1.005 -0.534
0.351 0.636 1.213 -0.105 -0.466 -0.773 -1.015 2.96 -1.681 -1.494
1.204 -0.698 1.429 1.318 -1.471 -0.1 0.463 1.638 -0.212 1.273
1.701 -0.936 -1.452 0.402 -0.262 0.872 0.272 -0.239 -0.805 -1.164
0.301 1.102 -2.179 0.437 -1.444 0.771 -0.821 1.538 -0.171 -0.363
1.432 1.004 1.667 1.423 -0.489 0.283 -0.129 2.437 -0.251 -0.582
0.027 -1.23 -0.266 0.191 0.266 -2.121 1.428 0.09 0.569 -2.374
0.357 -0.419 -0.11 0.678 1.094 -0.908 1.305 1.568 0.817 -0.806
-1.212 0.921 0.283 -1.483 0.562 -0.773 -0.972 1.371 -0.515 -0.826
-0.239 1.142 0.445 1.246 -0.851 -0.514 0.433 1.311 0.021 -1.254
-0.346 0.797 1.673 1.535 0.284 -0.051 1.4 2.247 0.031 -0.501
0.227 0.712 0.39 -2.264 -0.456 0.772 -0.049 -0.357 1.368 -1.539
-0.499 1.199 0.87 0.448 0.296 0.449 0.964 1.826 2.255 2.042
-0.342 0.422 -2.107 1.02 -0.397 0.485 1.166 -0.627 -0.086 -2.004
-0.755 0.225 -1.473 0.058 -0.995 -1.35 0.235 -0.095 0.047 -0.943

530
 CVPS7 CVPS6 CVPS5 CFI4 CFI2 CFI1 CQSE3 CQSE2 CQSE1 CSQM8

0
0.21 0
-0.204 0.001 0
0.708 0.287 1.129 0
-1.168 0.714 -1.226 -0.263 0
-0.05 0.165 -0.026 0.043 0.124 0
1.689 0.129 2.953 2.554 -0.14 -0.683 0
-2.873 -1.388 -0.569 -1.003 1.511 -1.574 -0.06 0
0.34 -1.02 0.776 -0.796 0.533 0.132 -0.583 0.575 0
0.29 0.326 0.141 0.47 0.053 -0.353 0.518 -0.818 -1.307 0
-0.302 -0.059 0.332 0.619 -0.371 -0.571 1.657 0.685 -0.61 0.31
0.428 -2.115 0.84 0.621 0.486 0.287 -0.167 -0.711 0.101 -0.208
0.376 0.588 0.507 -0.324 0.275 -1.46 2.564 0.568 1.725 0.667
0.933 2.007 1.134 -0.334 0.586 0.029 -0.257 0.902 0.223 -0.157
-1.276 -0.247 -1.628 -1.929 2.2 -0.154 -2.848 -1.158 -1.015 -2.673
0.726 -1.298 -1.555 -2.323 1.421 0.093 -0.374 -1.07 0.984 -0.131

CSQM7 CSQM4 LTM9 LTM8 LTM4 LTM1

0
-0.243 0
0.138 1.406 0
-0.83 0.382 -0.233 0
-1.568 1.004 -0.127 0.399 0
0.628 2.77 0.08 -0.633 0.594 0
Average absolute residual = 0.0027
Average off-diagonal absolute residual = 0.0865

% falling between -0.1 +0.1 = 99.99%

531