Understanding Quantitative

Research Design

Dr. Reynaldo C. Cruz

Graduate School Professor
La Consolacion University Philippines
• QUANTITATIVE RESEARCH is a
research approach where data are
assigned some numerical values,
and data analysis is done using
precise measurement. Quantitative
analysis can range from the
examination of simple
frequencies/frequency distributions
to the description of events or
phenomena using descriptive
statistics, and to the investigation
of correlation and causal
hypotheses using various
statistical tests.
• QUALITATIVE RESEARCH is a
broadfield of inquiry that uses
unstructured data collection methods,
such as observations, interviews, or
documents to find themes and meanings
to inform our understanding of the world.
Qualitative research tends to uncover
the reasons for behaviors, attitudes, and
motivations, instead of just the details of
what, where, and when. Normally
qualitative research does not make use
of precise measurement.
o MIXED METHOD OF RESEARCH is
the type of research in which a
researcher or team of researchers
combine elements of qualitative and
quantitative research approaches for
the purposes of breadth and depth of
understanding and corroboration
(Cresswell & Clark, 2011)
 Dimensions of Contrast Among the Three
Methodological Approaches
Qualitative Mixed Method Quantitative
Dimensions Position
Position Position
Methods Qualitative methods Mixed methods Quantitative methods

Researchers QUALS Mixed QUANS

methodologists
Paradigms Constructivism (and Pragmatism; Postpositivism
variants) transformative Positivism
perspective
Research QUAL research MM research QUAN research
questions questions questions (QUAN questions; research
+ QUAL) hypotheses
Form of data Typically narrative Narrative + Typically numeric
numeric
Purpose of (Often) exploratory (Often) confirmatory
research plus confirmatory Confirmatory plus plus exploratory
exploratory
 Dimensions of Contrast (Cont.)
Qualitative Quantitative
Dimensions Mixed Method
Position Position
Position

Both inductive and Rooted in conceptual

Role of theory; Grounded theory; deductive logic; framework or theory;
logic inductive logic inductive-deductive hypothetico-deductive
research cycle model
Ethnographic
MM designs, such as correlational; survey;
Typical studies research designs
parallel and experimental; quasi-
or designs and others (case
sequential experimental
study)

Probability, purposive
Sampling Mostly purposive Mostly probability
and mixed

Integration of
Thematic strategies: Statistical analyses:
thematic and
Data analysis categorical and descriptive and
statistical; data
contextualizing inferential
conversion
Validity/ Trustworthiness; Inference quality;
Internal validity;
trustworthiness credibility; inference
External validity
issues transferability transferability
Contrast Between QUANS and QUALS Methodology
QUANTITATIVE QUALITATIVE
Research process is deductive. Research process is inductive.

Social reality, meaning is

Measure objective facts.
constructed.

Focus on variables. Focus on in-depth meaning.

Values are present & explicit

Value-free research.
(empathy).

Independent of context. Contextual importance.

Contrast Between QUANS and QUALS Methodology
QUANTITATIVE QUALITATIVE
Many cases, subjects. Few cases, participants.
Statistical analysis Thematic analysis
Researcher as the central tool
Objective instruments of data
for data collection.
collection.

Highly structured research Loosely structured research

process. process.

Researcher is detached Researcher is immersed

(outsider). (insider).
(Adapted from Neuman, 1997: 14)
Contrast Between QUANS and QUALS Methodology
QUANTITATIVE QUALITATIVE
result oriented
process oriented

holistic perspective
particularistic and analytical

objective “outsider view” subjective “insider view” and

distant from data closeness to data

generalization by comparison
generalized by population
of properties and contexts of
membership
individual organism
 Elements of the Research Process
• Deductive thinking (Quantitative)

THEORY

HYPOTHESIS

OBSERVATION

CONFIRMATION
 Elements of the Research Process
• Inductive thinking (Qualitative)

OBSERVATION

PATTERNS

HYPOTHESIS

THEORY
 Procedures in Conducting
Quantitative Research
• Identify and define a problem or question
• Formulate the hypothesis(es) [a possible answer
to the question]
• Collect, analyze, and interpret the data/
information related to the question and hypothesis
• Form conclusions derived from the information
analyzed
• Use the conclusions to verify or reject the
hypothesis.
 Step1: Identifying the Research
Topic/Problem
Bases:
When there are… From:
 Knowledge gaps or  Books & Journals
 Unknown information  Conferences or Seminars
 Group Discussions
 Conflicting data &
research results
 Meetings &
Consultations
 New approaches which
 Field Experiences
need to be tested out &
monitored.
 From General to Specific
The research topic must be translated into more specific research
problems & objectives.

Topic Street Children

Problem What is the situation

of street children in
urban areas?

Objective (s) To conduct a situation

analysis on the needs
of street children in
three urban areas.
 What is a research problem?
• It is an issue that is borne out of a
perplexing situation where difficulties are
observed to exist.
• It is considered as a situation that needs to
be changed or addressed.
• These problems consists of:
a. areas of concern
b. conditions to be improved
c. Difficulties to be eliminated
d. Questions seeking answers
Example Showing the Main Problem Statement
and the Sub-Problem Questions
The primary purpose of this study was to determine if there is
any significant difference between the effects of written exercises
and frequency of tests on the retention and academic achievement
of students in mathematics.

More specifically it attempted to answer the following

questions:

1. Does giving frequent tests in mathematics promote better

retention of the lessons taught in the class than just giving
written exercises to the students?

2. Does giving frequent tests in mathematics promote better

academic achievement of the lessons taught to the students
than just giving written exercises in the class?
Example Showing the Main Problem Statement
and the Sub-Problem Questions:

3. Do the effects of written exercises and test

frequency on the academic performance of students
with different abilities in mathematics vary according to
the term of the examination?

4. etc.
 Step 2: Formulating Hypothesis
What is hypothesis?
• A tentative, reasonable, testable assertion regarding the
occurrence of certain behaviors, phenomena, or events; a
prediction of study outcomes. (Fraenkel and Wallen, 1993)

• a testable statement of a predicted relationship or difference

among two or more variables. (Mertler and Charles, 2008)

• a tentative statement about the expected relationship or difference

between two or more variables
What makes a “good” hypothesis?
• Posit a relationship or difference between
variables
• Reflect a theory or a body of literature
upon which they are based.
• Be brief and to the point
• Be testable
• It should be stated clearly and
unambiguously in the form of a declarative
sentence.
 Statement of the Hypothesis
• The hypothesis are stated in the null form of “no
difference” or of ‘no relationship” between and among
the involved variables.
• Should include a very clear and objective direction as to
how the hypothesized variables are to be investigated.
• Should be very well-formulated and relevant to the main
problem of the study and its sub-problems.
• Each stated hypothesis jibes with the pertinent sub-
problem statement.
• On sub-problems which are “hypothesis-free,” it should
be stated at the outset of the section.
• Or only the pertinent hypothesized sub-problem numbers
are stated before presenting the corresponding null
hypothesis in said section.
 The Null Hypothesis (Ho)

• A statement of equality between groups in an

investigation.

• states the absence of difference or relation

 
• is a statement of “no significant relationship” or “no
significant difference” between two or more
variables or groups that are to be measured and
tested through inferential statistics.
 Examples of Null Hypothesis

1) The mathematical ability of males is equal to the

mathematical ability of females.
2) Boys and girls do not differ significantly in
mathematical ability
3) There will be no significant difference in the average
score of ninth graders and the average score of twelfth
graders on the ABC memory test.
4) There is no significant relationship between personality
type and job success.
5) There is no significant difference in voting patterns as a
function of political party.
6) The brand of ice cream preferred is independent of the
buyer’s age, gender, and income.
 The Alternative Hypothesis (H1)

• usually called the research hypothesis, it states the presence

of a difference or relation. It contradicts the null hypothesis.
• is the reverse statement of a null hypothesis. It states that
“there is a significant difference” or “there is a significant
relation” (for a non-directional / two-tailed test) or that the
difference or relation is in a certain direction (for a
directional / one-tailed test).
• A statement of inequality between groups in an investigation.
• Is a definite statement of the relationship or difference
between two variables.

 
 Directional Alternative Hypothesis

• reflects a difference or relationship between groups, and

the direction of the difference or relationship is specified.

• if there is a difference, then one group has the higher /

lower value
 
• if there is a relationship, then the relationship is either
positive or negative

 
 Examples of Directional Hypothesis
1) Boys perform significantly better than girls in
mathematics
2) There is a significant positive relationship between
attitude toward math and performance in math.
3) Second graders like school less than first graders
but more than third graders
4) Students with academic disabilities will significantly
have more negative attitude about themselves if
they are placed in special classes than if they are
placed in regular classes.
5) A is greater than B (or A > B)
6) B is greater than A ( or B > A)
 Non-directional Alternative Hypothesis
• reflects a difference or relationship between
groups, but the direction of the difference or
relationship is not specified.

Examples :
1) Boys and girls differ significantly in mathematics.
2) There is a significant relationship between sex and
mathematical ability.
3) There will be significant difference between the
scores on an attitude measure of students with
academic disabilities placed in special classes and
such students placed in regular classes.
Step 3: Collecting,& Analyzing,
Interpreting Data

Data Collection:
 Choosing appropriate instrument
 Identifying a population group or a
representative sample

Data Analysis & Interpretation:

 Critical Factors to Consider in
Doing Quantitative Analysis

1.Choice of appropriate statistical

tests,
2.Accuracy in statistical calculations,
3.Analysis & interpretation of
statistical results.
 Guidelines in Choosing
Appropriate Statistical Tests
Do you like to Use Central Tendencies /
describe status of Yes Variabilities or other
phenomenon? Descriptive Measures

No

Are you comparing Use Chi Square or

one distribution of Yes Critical Ratio
scores to another?

No

Are you comparing How many Use t- test

the means of two means are 2 or Analysis
Yes of Variance
or more groups? you
comparing?
Are you comparing How many Use t-test
the means of two Yes means are 2 or Analysis
or more groups? you of Variance
comparing?

No
More than 2
Are you trying to
determine how Use Analysis of Variance
closely
performance on Think again! You either (1) need a
what variable No statistic not covered in these two
predicts chapters or (2) have made a mistake
performance on in your thinking.
another?

Yes
Are you using a
combination of 2 or Yes Use multiple Correlation
more variables to
predict an outcome? You want to
measure the extent Use
of influence of Regression
No predictor variables Analysis
to the criterion
Are you trying to variable
determine a curvilinear
relationship Yes
Use eta coefficient
No

Are you holding a third Yes Use partial correlation

variable constant?

No

Use Pearson correlation Figure 1. Vockel’s Algorithm

for Selecting Appropriate
Statistical Tests
Statistical Test Purpose Assumption
A. Measures • Typical Value or a
of representative value
Central of a mass of
Tendency quantitative data.
• Synonymous to average
1. Mean score. Appropriate for
interval and ratio
variable
2. Median
• Divides the distribution
into 2 equal parts. Suited
for ordinal data.
3. Mode • Most frequent score.
Suited for nominal and
interval data.
Statistical Test Purpose Assumption
B. Measures of • Degree of spread
Dispersion or variability of a
given set of data.

1. Range • Describing the extreme

2. Quartile scores
Deviation • Analyzing the semi-
interquartile range of
scores or item values.
1. Standard
• Deviation of values from
Deviation
the mean.
• For higher statistical
inferences like those
involving coefficients of
correlation and standard
scores.
Statistical Test Purpose Assumption
C. Measures of

Location • Percent of cases below

1. Percentiles a particular rank position
• Analyzing the value of
2. Quartiles the variable based on
quarter position.
• Analyzing the value of
3.Deciles
the variable, the
percentage being taken
in units of 10.
D. Other
Descriptive
Measures

1. Percent • Standardizing for size,

looking for a subgroup in
relation to the whole.
2. Proportion
• Comparing the relative
size of a subgroup as
against another
subgroup.
Statistical Test Purpose Assumption
The categories must be
mutually exclusive and
exhaustive.

• Describing symmetry or
3. Skewness
asymmetry; normality or
abnormality.
4. Kurtosis • Describing peakedness
or the nature of the
distribution hump and
tails.
E. Measures of • Describe whether
Correlation or not variables are
correlated; describe
1. Pearson ( r ) the nature and
• Measure of Linear
extent of
relationship correlation of two
between and variables which are
among variables. either ratio or interval.
Statistical Test Purpose Assumption
2. Spearman • Correlation between two
(rho) ordinal variables.

3. Phi Coefficient • Correlation between two

(ro) real nominal
dichotomous variables.
4. Point-Biserial
• Correlation between a
real dichotomous
variable and an interval
5. Multiple variable.
Correlation
• Correlation between a
real dichotomous
variable and a interval
variable.
6. Partial
Correlation • Correlation of two
variables partialling out
the influence of a third
variable.
Statistical Test Purpose Assumption
2. Spearman • Correlation between two
(rho) ordinal variables.

3. Phi Coefficient • Correlation between two

(ro) real nominal
dichotomous variables.
4. Point-Biserial
• Correlation between a
real dichotomous
variable and an interval
5. Multiple variable.
Correlation
• Correlation between a
real dichotomous
variable and a interval
variable.
6. Partial
Correlation • Correlation of two
variables partialling out
the influence of a third
variable.
 Statistical Test Purpose Assumption
7. Multiple • Describing the
Regression correlation among
variables – identifying
cause and effect
relationship.
F. Inferential
Statistics

1. Student’s t • Testing a hypothesis

(One-Sample about the mean of a
Chi-Sample normally distributed
Test) population, using a one-
sample group.
2. t-test of
Correlated • Significance of
Means differences between two
(Wilcozon Test) correlated groups.
3. t-test of
Independent • Significance of
Means (Mann differences between two
Whitney U test) independent groups.
 Statistical Test Purpose Assumption

F. Inferential Finding significance

Statistics of differences or
variations
• Significance of
4. F- test One
-way Using k samples differences among three
or more independent
(Kruskal Wallis groups; data in the
Test) strength of interval
measurement.
5. F-test Two-way
• Function of 2-way F plus
(Friedman test) Using k samples the interactive effect.
plus interaction
6. Analysis of effect • An extension of ANOVA
Covariance allowing for the
correlation between
initial and final scores.
Co-variance analysis
effects adjustment in the
final score or terminal
score which allow for
differences in some
terminal variables.
 Statistical Test Purpose Assumption
F. Inferential
Statistics

7. ANCOVA
Used in finding the • An extension of ANOVA
Effect of covariate. allowing for the
correlation between
initial and final scores.
Co-variance analysis
effects adjustment in the
final score or terminal
score which allow for
differences in some
terminal variables.
 Suggested Format in Doing
Quantitative Analysis
1.Briefly introduce the topic.
2.Present the data/statistics.
3.Analyze the data.
4.Interpret the findings
 Illustrative Case
Involving Descriptive and
Non-parametric
Inferential Analysis
Research Title:

The Leadership Behavior of School

Administrators in the City Division of San Jose
del Monte and Its Influence on Teachers
Performance

Major Problem:

How do the leadership behaviors of school

administrators influence teachers’ performance?
Specific Problems:

1.What is the profile of school administrators

in terms of age, educational preparation,
administrative experiences, trainings.?
2.How may the leadership behavior of school
administrators be described in terms of task
and relationship dimensions?
3.What is the level of performance of the
teachers in terms of: teaching competency,
professional development activities, etc.?
4.Do the leadership behaviors of school
administrators exert significant influence on
the performance of the teachers?
2.Formulating the Hypotheses
• Research Hypothesis (H)
“The leadership behavior of school
administrators influences teachers’
performance.”
• Alternative Hypothesis (H1)
“The leadership behavior of school
administrators positively influences teachers’
performance”.
• Null Hypothesis (Ho)
“The leadership behavior of school
administrators does not exert significant
influence on teachers’ performance”.
Profile of the Administrators
The profile of the administrators is
presented in terms of age, educational
preparation, administrative
experiences, and trainings.
 Illustrative Case
Involving Inferential
Analysis

Correlation and
Regression Procedures
Patients’ Knowledge of Tuberculosis and
Determinants of Treatment Compliance
A DOH-DOST-CLHRDC Commissioned Research

Project Leader : Dr. Reynaldo C.Cruz

La Consolacion University Philippines

Asst. Project Leaders: * Ms. Mart Juaresa Cabantog

Bulacan State University
* Ms.Jean Guillasper
Nueva Ecija Univ. of Science & Technology
* Mr. Ferdinand Rivera
Bataan Peninsula State University
* Ms. Contessa Gabriel
Ramon Magsaysay Technological University
* Ms. Evelyn Labastilla
Tarlac State University
* Ms. Grace Saniel
Population Commission
* Dr. Juanita Cruz
System Plus Foundation Colleges
* Dr. Emerlita S. Naguiat
Bulacan State University
 Objectives of the Study
 to describe the demographic characteristics of the
tuberculosis patients in Central Luzon.

 to describe the classification and types of

tuberculosis of the patient respondents.

 to determine the patients’ knowledge and attitudes

on tuberculosis.

 to identify the determinants of treatment compliance.

 Determinants of Treatment
Compliance
 Three sets of factors were hypothesized to account
on the treatment compliance of the TB patients to the
treatment regimen:

 First, knowledge of tuberculosis, specifically:

Knowledge on the acquisition, transmission,

sign and symptoms, prevention, and treatment.
 Determinants of Treatment
Compliance
 Second, health system’s variables, specifically:
 Health seeking behavior, TB stigma, health
providers, and health facilities.

 Third, the socio-economic characteristics of the

respondents, specifically:
 age, education, gender, status, income, home
ownership, religion, no. of household members, no.
of family members infected with TB, number of
rooms, history of drinking, history of smoking, no. of
stick consumption, and contact with members with
TB.
 Determinants of Treatment
Compliance
 Results of the correlation analysis revealed
that three out of five primary factors
correlated positively and significantly to the
total compliance of the patients to the
treatment regimen. These were knowledge of
disease transmission, prevention, &
treatment.
 The other factors specifically the health
systems factors and the demographic
characteristics of the patients likewise
correlated positively to the level of treatment
 Determinants of Treatment
Compliance

 Two variables however, correlated negatively

to treatment compliance: history of drinking
(-.043), and history of smoking (-.01). This
means that patients with long history of
smoking and drinking habits are generally
the ones less likely complying with the TB
regimen.
 Determinants of Treatment
Compliance
 Results of the regression analysis revealed that
knowledge of tuberculosis recorded B coefficients
with associated probability lower than the significance
set at .05. This means that knowledge of the disease
is a significant determinant of compliance to the TB
treatment regimen.

 The obtained B coefficients indicate that for every

unit increase in knowledge on how the disease could
be acquired, prevented, treated, its signs & symptoms,
how the disease could be acquired & transmitted
could generate .004, .159, .121, .025, & .108
increases in treatment compliance.
 Determinants of Treatment
Compliance
 The other factors also correlated positively with the
patients’ compliance to the treatment regimen, but not
to a significant extent. This means that health system
factors specifically, TB stigma, health seeking
behavior, health providers, and health facilities also
account on treatment compliance.

 In the same vein, the socio-economic characteristics of

the TB patients, their age, gender, family income,
educational attainment, marital status, household size,
family history, history of smoking, and history of drinking
likewise exert differential effects on their treatment
compliance but not to a significant extent.
 Determinants of Treatment
Compliance
 Results of the analysis of variance conducted revealed
an F – Ratio of 6.203 with a significance level much
lower than .05. This means that taken jointly, the
factors namely: knowledge of tuberculosis, the health
system factors, and the socio-economic
characteristics of the patients formed a very significant
set of predictors in determining the compliance of
patients to the treatment regimen.
 PRACTICUM

Direction:

Do a quantitative analysis of the

data in the Contingency Table
showing the correlation between
the leadership behavior of school
administrators and the teachers
performance.
 Thank You
and
Good Day!

Dr. Reynaldo C. Cruz