1 1

2 0

BM2101
Research Methodology

Research Theory
Objectives
 Recognize your research philosophy
 Classify different types of research methodologies
 Relate research designs with relevant research methodologies
 Interpret the style of the research document
 Judge the appropriateness of the chosen research design
 Distinguish between inductive and deductive approaches
 Define variables, measurements and their related terms
 Compare structure of quantitative, qualitative and mixed methods research study
 Consider the implication of time constraints on the choice of research design
 Explain the difference between quantitative, qualitative and mixed methods designs
 Able to critique others’ research designs
 Develop and justify one’s own research design

2
Some Remarks
 The response rate for the survey may be too low, and this may necessitate a change in research strategy,
such as by interviewing a new group of respondents
 Cross-sectional studies: involving the study of a particular phenomenon (or phenomena) at a particular
time
 Longitudinal studies: its main strength is its capacity to study change and development. This type of study
may also provide you with a measure of control over some of the variables being studied.
 Even with time constraints it is possible to introduce a longitudinal element to your research.
 Archival and documentary research
 Documentary research can be conducted alongside a grounded theory strategy based on qualitative
interviews and use a similar produce to analyse both sets of data.
 Can involve documentary research within a case study strategy

3
Survey
 Usually associated with deductive research approach
 May be descriptive, interpretive, or causal studies
 Popular and common strategy in business and management research and is most frequently used to
answer What, Who, Where, How much and How many questions.
 Survey is not a method of data collection. To collect survey data, we use measuring instruments such as
questionnaires and observation devices
 Allows you to collect quantitative data which you can analyze quantitatively using descriptive and
inferential statistics.
 In descriptive surveys, the intent is to infer population characteristics from a subset of the population,
called a sample
 In interpretive or causal studies based surveys, researchers ask respondents for their viewpoints on a
particular topic, preferences on a range of options, or reasons for their actions
 Case study is about depth; but survey is about breadth
4
Survey
 Surveys are popular because they provide a quick and efficient way to obtain broad answers based on a
sample before generalizing it to the population. The weaknesses of surveys include possible researcher,
sampling, and response biases. They are also less appropriate if in-depth answers are required
 Surveys may be ad hoc or carried out at regular intervals
 Cross-sectional surveys: gather information about a population using a sample at a point in time
 Not suited for understanding changes over time

 Longitudinal study: collect data over time to monitor changes. Three types:
 Trend studies using different samples.

 Cohort studies using different samples from the same cohort. Cohort is a group of individuals who

share a common characteristic such as year/period of birth or class.

 Panel studies using the same sample. Panel study helps to increase the sample size to improve the

precision of our sample estimates.

 With same samples, there is a risk of participants dropping out of the survey.
 Tend to be more expensive than cross-sectional surveys because of the need to monitor samples over

time. 5
Survey
 Pilot survey: done on a small but similar sample to determine if the survey design and method of data
collection (e.g., a questionnaire) may be improved before the actual larger scale survey.
 Sampling: Consists of identifying the:
 Sampling unit or elements: smallest unit of observations that is of interest to the researcher in a

sample.
 Household in consumer-surveys, organizations or employees in business surveys, …
 Target population: aggregate of all sampling units

 Called as target audience in consumer surveys

 Also called as theoretical population; theoretical here means a concept or idea and not a real
thing
 Sampling frame (if any): list of elements from which sampling takes place

 Unlike the theoretical population, this is real

 Not all survey have a sampling frame. For example, it is not possible to obtain a list of every fish in
a lake.
6
Survey
 Sampling: Consists of identifying the:
 Sampling method

 Probability samples: use random sampling to draw the samples from a sampling frame
 As long as the sampling frame is close to the target population, probability samples are more
accurate than non-probability samples.
 Main types of probability samples:
 Simple random samples: draw the sampling units randomly. Despite its simplicity and wide
discussion in elementary statistical textbooks, we rarely use simple random samples in actual
surveys because of the need for a large sampling frame.
 Systematic samples: first divide the target population into smaller lists and then draw the
elements following a particular pattern after a random start.

7
Survey
 Sampling: Consists of identifying the:
 Sampling method

 Main types of probability samples:

 Stratified samples: commonly used in actual surveys. First stratify the target population
according to some criteria such as gender, income, workplace, residence, or cohort. Then
draw the elements randomly from each stratum.
 Proportional stratified sampling: samples are drawn in equal proportion

 Disproportional stratified sampling: sample proportions are unequal

 Cluster samples: select random clusters rather than elements. Often used in city surveys to
identify potential respondents.

8
Survey
 Sampling: Consists of identifying the:
 Sampling method

 Non-probability samples: if a sampling frame is not available or too difficult to construct, then
non-probability samples may be used.
 Convenience samples: select based on convenience. Used when researcher is more
interested in some quick responses than in representatives.
 Purposive samples or judgemental samples: our judgement or the choice of experts is
preferred to random sampling
 Sampling errors depend on the quality of judgement. Different experts may not agree

on what is representative.

9
Survey
 Sampling: Consists of identifying the:
 Sampling method

 Non-probability samples: if a sampling frame is not available or too difficult to construct, then
non-probability samples may be used.
 Quota samples: similar to stratified sample except that chance selection is not used in each
stratum. Chosen out of convenience rather than random.
 Quota samples are popular because stratification reduces the need to select large

samples. The greater homogeneity of the sampling elements with each stratum reduces
the sampling error. A sampling frame is not required, which makes it cheaper to draw a
sample
 Snowball samples: begins with a few respondents who provide referrals for the researcher to
contact additional respondents.


10
Survey
 Sampling: Consists of identifying the:
 Sample size: No guide on sample size. Not to consider 30, as t distribution will be close to normal

curve
 Qualitative surveys: aim should be to gather as many different views as possible.
 Quantitative surveys:
 Consider cost or difficulties to access potential respondents
 Higher power (): this means reduce the Type II error. Type II error is reduced when:
 The spread of each curve is smaller (called spread size)

 The curves are further apart (called effect size)

11
Survey
H0 is actually Person is actually
Decision True False Decision Innocent Guilty
Reject H0 Type I error Correct Guilty Type I error Correct
Accept H0 Correct Type II error Innocent Correct Type II error

 Which error is serious? Depends on the situation. For court, we don’t wish to have a justice system that
sends innocent to jail, so Type I error is serious. For manufacturing, rejecting a good batch (Type I error) by
mistake is less serious than not rejecting a bad batch (Type II error)

 If we have only few firms and we wish to do some analysis:

 Cross-sectional data is not sufficient for regression analysis due to lack of data.
 Panel study can help in getting more data, as it is obtained over time from the same sample.

12
Errors in Surveys
 Non-sampling errors:
 Administrative errors: mistakes in data collection or processing

 Respondent errors: giving incorrect answers because they cannot recall or want to hide information or

not in mood to talk

 Conceptual errors: part of the researcher, such as the use of different and confusing concepts of culture

 Measurement errors: the measure of a concept is inappropriate

 Sampling errors: chance variations such as the variation of sample means from the population mean.
 From central limit theorem: the size of the variation is , where is the population standard deviation and

n is the number of samples.

13
Case Study
 In-depth inquiry into a topic or phenomenon within its real-life setting
 “case” in case study may refer to a person or group, an organization, an association, a change process, an
event as well as many other types of case subject
 Case study research sets out to understand the dynamics of the topic being studied within its setting or
context. “Understanding the dynamics of the topic” refers to the interactions between the subject of the
case and its context
 Case study research is often used when the boundaries between the phenomenon being studied and the
context within which it is being studied are not always apparent. Understanding context is fundamental to
case study research.
 The interaction between a phenomenon and its context is best understood through in depth case studies.
 Some positivist researchers have advocated using case studies inductively to build theory and develop
theoretical hypotheses, which can be tested subsequently. In this way, the use of the case study is
advocated in the early, exploratory stage of research as a complement to deductive research.

14
Case Study
 Case study research may beneficially use some combination of archival records and documentation,
different forms of observation, ethnography, interviews and focus groups, questionnaires, reflection and
the use of research diaries and other research aids.
 Case study research is likely to prove to be challenging because of its intensive and in-depth nature and
your need to be able to identify, define and gain access to a case study setting.
 Case study strategies can be distinguished based on:
 Single case versus multiple cases
 Holistic case versus embedded case

 Literal replication: cases will be carefully chosen on the basis that similar results are predicted to be
produced from each one.
 Theoretical replication: another set of cases may be chosen where a contextual factor is deliberately
different. The impact of this difference on the anticipated findings is predicted by the researcher.

15
Case Study Vs Case Method
 Interpretive study that seeks in-depth or intensive investigation of a particular case to discover or explore something
new rather than test a hypothesis
 Tells a big story through the lens of a smell case
 Case studies are different from case methods that are widely used in business schools to teach students
 Case is a unit of study: person, team, project, organization, province, country, process, activity, or situation
 Unit should be a clearly defined and bounded system
 Actor’s perspective guides the research
 Historical case study: there is a little control over extraneous variables in a historical case study
 Case method: not a research design
 Students study a written business case in advance of each class and debate the issues in class
 Case method is a form of teaching and learning as an alternative to lectures, role playing, and other methods of
instruction
 Goal is to engage students in their learning on a real or imaginary business case

16
Case Study
 Holistic: has thick description; the researcher puts the various parts together in a coherent manner
 Thin description is one where the explanation is superficial
 Both qualitative and quantitative data can be used
 Qualitative data is more common in interpretive case studies
 Causal case studies:
 Has some relationship
 such as in tracing the development of an organization or city, qualitative data also predominate
 Interpretive case studies: Researcher aims to interpret the data by developing conceptual categories,
supporting or challenging the assumptions made regarding them.
 Evaluate case studies: researcher goes further by adding their judgement to the phenomena found in the
data.

17
Case Study
 Sampling:
 Typical case
 Unique case
 Test case
 Multiple cases: used when
 Unsure if the selected case it typed, in which case he may use another case;
 Wants to compare and contrast different cases, such as the different ways in which village schools
are run
 Wishes to sample theoretically to build theory, to generate new ideas from a small number of cases
until the point of diminishing returns.

18
Ethnography
o Used to study the culture or social world of a group
o Literally means a written account of a people or ethnic group
o Ethnographers study people in groups, who interact with one another and share the same space, whether
this is at street level, within a work group, in an organization or within a society
o Realist ethnography
o Interpretive ethnography
o Critical ethnography
o While ethnography is a demanding strategy to use because you would need to develop some grounding in
this approach and because of the time scale and intensity involved, it may be relevant to you. If you are
currently working in an organization, there may be scope to undertake participant observation of your
workgroup or another group in the organization.

19
Action Research
 Emergent and iterative process of inquiry that is designed to develop solutions to real organizational
problems through a participative and collaborative approach, which uses different forms of knowledge, and
which will have implications for participants and organization beyond the research project
 Purpose: to promote organizational learning to produce practical outcomes through identifying issues,
planning actions, taking action and evaluating action.
 Process: the process of action research is both emergent and iterative. An action research strategy
commences within a specific context and with a research question but because it works through several
stages or iterations the focus of the questions may change as the research develops.
 Cycle 1: Teasing out the issues
 Cycle 2: Understanding the customer project
 Cycle 3: Acting on knowledge
 Each cycle involves diagnosing, planning action, taking action and evaluating action

 Includes participation, knowledge and implication

20
Grounded Theory
 Used to refer to a methodological approach, a method of inquiry and the result of a research process
 Developed as a process to analyse, interpret and explain the meanings that social actors construct to make
sense of their everyday experiences in specific situations.
 Provides a systematic approach to collect and analyse qualitative data. Its use is often well suited to the
nature of qualitative research. Whereas in quantitative research it is usual to collect a complete set of data
and then analyse these, in qualitative data it is often useful to analyse data as you collect them
 Usually referred to as taking an inductive approach. Can be appropriate to abductive as well.
 Make sure to decide how to select cases for your research.
 Select new cases to collect and analyse further data. This approach is a special form of purposive sampling,
known as theoretical sampling, which continues until theoretical saturation is reached
 Objectivist ground theory and constructivist ground theory exist.

21
Grounded Theory
 Theoretical sensitivity: using this strategy will mean you should avoid being overly sensitized to pre-existing
theoretical concepts, to allow yourself to make sense of participants’ meanings in the data to guide your
research.
 Using ground theory is time consuming, intensive and reflective

22
Grounded Theory
 Key elements of grounded theory are:
 Early commencement of data collection
 Concurrent collection and analysis of data
 Developing codes and categories from the data as these are collected and analysed
 Use of constant comparison and writing of self-memos to develop conceptualization and build a theory
 Use of theoretical sampling and theoretical saturation aimed at building theory rather than achieving

(population) representativeness
 Use of an abductive approach that seeks to gain insights to create new conceptual possibilities which

are then examined

 Initial use of literature as a complementary source to the categories and concepts emerging in the

data, rather than as the source of categorize these data. Later use to review the place of the grounded
theory in relation to existing, published theories
 Development of a theory that is grounded in the data
23
Narrative Inquiry
 A narrative is a story; a personal account which interprets an event or sequence of events. Using the term
‘narrative’ requires a distinction to be drawn between its general meaning and the specific meaning used
here
 A narrative be defined as an account of an experience that is told in a sequence way, indicating a flow of
related events that, taken together, are significant for the narrator and which convey meaning to the
researcher.
 Structural elements that are useful to facilitate analysis of narratives
 What is the story about?
 What happened, to whom, whereabouts and why?
 What consequences arose from this?
 What is the significance of these events?
 What was the final outcome?

24
Comparative Design
 Compares and contrasts several cases to draw inferences
 Overcomes a weakness of causal study design, which does not allow for the user of comparisons to draw causal
inferences
 In terms of sample size, comparative design is a small-N design. Falls somewhere between a case study and a survey
 Researchers prefer this when there are only a few cases
 Small sample size is not suitable for multivariate statistical analyses or broad large-sample surveys
 Comparative studies are intended to be causal, that is, the researcher deliberately chooses the cases to discover
causality, rather than through random sampling
 To make comparison easier, it is helpful to reduce the number of variables to a more manageable set.
 Limitations of comparative design
 The inability to prove cause; Difficulties in finding suitable, diverse cases
 Difficulties in finding comparable cases; Difficulties in sorting out rival explanations if there are many variables and
only a few cases; The outcomes are binary
 To overcome the outcomes as binary limitations, qualitative comparative analyses considering truth tables and fuzzy or
Boolean Logic is considered. 25
Experiment
 Number of independent variables is k; list of independent variables is X; dependent variable Y
 If k is small and possibility exists for manipulating some variables to ascertain their effects on the Y by keeping other Xs
constant
 Variables X1 and X2 treatments because of the frequent use of experiments in clinical and agricultural studies
 Many researchers think experiments provide stronger evidence of causation by eliminating arrival explanations through
experimental control. However, this is only possible if there are a few variables. If k is large, it becomes more difficult to
disregard the influences of other independent variables by fixing their values
 Classical Experimental Design: an experimenter uses two groups of subjects, called the control group (C) and the
experimental group (E). He then administers a treatment (T) to E, and C provides a baseline to compare the effects.
 Blind experiment: the subjects in each group do not know whether they have been given the treatment.
 Double-blind experiment: possible to blind the investigators; they do not know whether the results come from E or
C.

26
Experiment
 Classical Experimental Design: an experimenter uses two groups of subjects, called the control group (C)
and the experimental group (E). He then administers a treatment (T) to E, and C provides a baseline to
compare the effects.
 Quasi-experimental designs:
 Less persuasive than the full classical experimental design
 Sometimes difficult to conduct the experiment and some adjustments have to be made.
 By not using a control group, perhaps because it is difficult to find another group large enough

for the experiment

 By not conducting a pre-test, perhaps because it is difficult or expensive to conduct, or because

the groups are large enough for the experimenter to assume that they are comparable
 By discarding the pre-test scores in the data analysis, possible because the pre-test scores are

used primarily as a check that both groups are comparable before the treatment is
administered

27
Experiment
 Classical Experimental Design: an experimenter uses two groups of subjects, called the control group (C) and the
experimental group (E). He then administers a treatment (T) to E, and C provides a baseline to compare the effects.
 Several threats to classical experimental design:
 Subjects maturing during the experiment, which may be handled by keeping the period between pre-test and
post-test short
 Measurement error; where the mathematics tests are inappropriate, such as if they are too difficult and
cheating occurs
 The testing effect, where subjects become better because they learn to do better after completing the pre-
test, or they have been “taught to the test,” that is, they have been drilled to do well
 Subjects reacting differently because they know they are being observed by taking part in an experiment
(Hawthorne effect)
 An external event, such as a change of teacher teaching the control group, which may invalidate the results
 If sample sizes are too small to provide firm conclusions, which can be increased by adding more subjects in E
and C (i.e., through replication) or by using more than two groups
 If the groups are not comparable
28
Experiment
 Classical Experimental Design: an experimenter uses two groups of subjects, called the control group (C) and the
experimental group (E). He then administers a treatment (T) to E, and C provides a baseline to compare the effects.
 Researchers use a variety of techniques to ensure that the experimental and control groups are comparable. These
strategies include:
 Randomization
 Matching: similar to the way we assign people to either side of a tug of war game
 Repeat measures design: only one group of subjects for the experiment. However, we measure each student
twice, once using the traditional method and another time using the new teaching technique.
 Sampling bias through self-selection: the participants in the experiment may not be representative of the wider
community
 Parallel Group Design: similar to classical experimental design except that the second group now receives a treatment.
Threats in classical experimental design applies here.

29
Experiment
 Repeat Measures (or paired) Design:
 Used to deal with the problem of having two groups that are not comparable
 Only one group with sample size N is used and each student is taught in two different ways (T and U), one after
another with a sufficient gap between the two periods.
 Presence of variance among participants does not matter as we are only interested in the differences within each
participant. There is less variability in the data (as the same participant is used and might be slightly difference as
they are done one after the other), which makes it easier to detect the differences in ratings.
 Randomized Block Design
 Extraneous variable: is a variable we are not interested in but its presence confounds the analysis. Also called as
noise variable. The experimenter has to deal with the unwanted noise

30
Regression
 Examines the influence of independent variables on a dependent variable
 Dependent variable is endogenous
 Independent variable is exogenous as the model does not explain changes in it. The changes in it are
determined outside the model
 Regression differs from correlation, where the relation is symmetric without any specification on whether Y
depends on X or vice versa
 To estimate the parameters from these n equations, we usually use the method of least square or ordinary
least squares (OLS) to find the line of best fit. In OLS, we minimize the sum of squares of the residuals (e i),
and hence the terms “least squares” and “best” fit.
 Residual is the difference between an observed point and the corresponding estimated point on the
regression line
 Unlike the experiment where we physically fix the values of extraneous variables to examine causality
among a few variables of interest, it is not necessary in regression. By means of statistical control, we can
determine the effects of many independent variables on the dependent variable. 31
Regression
 Sampling:
 Number of data points (n) should be much higher than number of parameters (k)
 A rule of thumb is at least 10 points for each slope coefficient ()
 If there are insufficient data points, we risk estimating an incorrect regression line
 Collect data that vary substantially in the X-direction. Otherwise, regression line will be sensitive to any
data point
 Collect representative samples


32
Theory Vs Worldview
 Distinction between a theory and a worldview
 A worldview relates to the beliefs and values of the researcher and how these inform a study.

 Theories, in contrast, are less abstract, drawn from the literature, and provide a guiding perspective for

designing many research process aspects. From theories, methodological approaches to design follow,
such as mixed methods research, and then data collection methods.
 Four levels for Developing a research study
 Paradigm worldview (beliefs, values of the researcher)

 Theoretical lens (e.g., feminist, racial, substantive content theories)

 Methodological approach (e.g., ethnography, experiment, mixed methods)

 Methods of data collection (e.g., interviews, checklists, instruments)

33
Theory Vs Worldview
 What theory informs: A summary of the theory use:
 The theory is visualized through a model.

 Constructs (variables) are drawn from the theory and incorporated in the study aims

 From the theory, stories from the literature about the model are drawn

 The theory constructs informed the qualitative questions addressed and the quantitative instruments

and measures
 The theory also provides “starter codes and themes” for the qualitative data analyses and for

hypothesized relationships for the quantitative data analysis

 The theory contributed to useful typologies of findings, led to working hypotheses, and illustrated the

development of caregiving over time.

34
Use of theory
 In a quantitative dissertation, an entire section of a research proposal might be devoted to presenting the
broader theory guiding the study hypotheses.
 In qualitative research, the use of theory is much more varied.
 Mixed methods research may contain a theoretical framework within which both quantitative and
qualitative data are collected.
 Theory can be a prediction of what the research expects to find or an inductive or deductive framework for
a study. It may be presented as a model or diagram of the relationship among variables or constructs.

35
Use of Theory
 The general procedures for locating and including theory are these:
 Identity the theory (or theories) from the literature search
 Place the theory (model or conceptual framework) at the beginning of the article as an a priori framework to guide
the question or hypotheses in the study
 Write about the theory by first advancing the names of the theory to be used followed by a description how it
informs the quantitative, qualitative, or mixed methods study
 Discuss the studies using the theory, especially studies closely related to the present study
 Include a diagram of the theory indicating the probable links in the theory and the major concepts or variables in
the theory
 Have the theory provide a framework for the study
 Return to the theory at the end of the study to review its role in informing findings the results. Compare the theory
use with other studies.

36
Quantitative Theory Use
 Some historical precedent exists for viewing a theory a theory as a scientific prediction or explanation for what the
researcher expects to find
 Theory in quantitative research is a “set of interrelated constructs (variables), definitions, and propositions that
presents a systematic view of phenomena by specifying relations among variables, with the purpose of explaining
natural phenomena”
 A discussion about this theory would appear in a section of a proposal in the literature review or in a separate section
called the theory base, the theoretical rationale, the theoretical perspective, or the conceptual framework.
 A theory might appear in a research study as an argument, a discussion, a figure, a rationale, or a visual model
 Investigators combine independent, mediating, and depend variables into questions based on different forms of
measures. These questions provide information about the type of relationship (positive, negative, or unknown) and the
magnitude (e.g., high or low)

37
Quantitative Theory Use
 Theory of attribution: hypothesis is tested over different settings with different population
 Theories are at three levels: (a) micro-level, (b) meso-level, and (c) macro-level.
 Micro-level theories provide explanations limited to small slices of time, space, or numbers of people, such
as Goffman’s theory of face work, which explains how people engage in rituals during face-to-face
interactions.
 Meso-level theories link the micro and macro levels. These are theories of organizations, social movement,
or communities, such as Collins’s theory of control in organizations.
 Macro-level theories explain larger aggregates, such as social institutions, cultural systems, and whole
societies. Lenski’s macro-level theory of social stratification, for example, explains how the amount of
surplus a society produces increases with the development of the society.

38
Variables in Quantitative Theory
 A variable refers to a characteristic or attribute of an individual or an organization that can be measured or observed
and that varies among the people or organization being studied.
 Independent variables are those that influence, or affect outcomes in experimental studies. They are described as
“independent” because they are variables that are manipulated in an experiment and thus independent of all other
influences.
 Dependent variables are those that depend on the independent variables; they are the outcomes or results of the
influence of the independent variables
 Predictor variables (also called antecedent variables) are variables that are used to predict an outcome of interest in
survey method studies. Predictor variables are similar to independent variables in that they are hypothesized to affect
outcomes in a study, but dissimilar because the researcher is not able to systematically manipulate a predictor variable
 Outcome variables (also called criterion or response variables) are variables that are considered outcomes or results of
predictor variables in survey method studies. They share the same properties as dependent variables
 Intervening or mediating variables(Intervening or mediating variables) stand between the independent and dependent
variables, and they transmit the effect of an independent variable on a dependent variable

39
Variables in Quantitative Theory
 Moderating variables are predictor variables that affect the direction and/or the strength of the
relationship between independent and dependent variables, or between predictor and outcome variables
A confounding variable refers to a “third variable” that is both related to the independent (or predictor)
variable and to the dependent (or outcome) variable
 A hypothesis is a prediction about a specific event or relationship between variables
 A scale of measurement in quantitative research would be the response options available to participants on
instruments or collected by the researcher observing participants
 Endogenous variable: variables that influence other variables.
 Exogenous variable: variable not caused by other variables in the causal model

40
Variables in Quantitative Theory
 Variables are measured as either categorical or continuous. Categorical scales have two types, nominal and
ordinal. Continuous scales also have two types: interval and ratio
 Nominal scales use categories that a participant can check, such as level of education
 Ordinal scale represents categories for the participant to check but provides the categories in a rank order
 Internal or ratio scale, contains a continuous scale.
 Causality means that we would expect variable X to cause variable Y
 If you aim to test a causal claim about the relationship between two or more variables in your quantitative
study, the best choice is to conduct a true experiment, which will provide more control over potential
unmeasured variables. If you are less interested in testing a causal claim or if you cannot conduct an
experiment, then survey methods can test claims about hypothesized associations between variables.
 Temporal order means that one variable precedes another in time. Because of this time ordering, one
variable affects or predicts another variable.

41
Theories in Quantitative Research
 Researchers state their theories in research proposals in several ways, such as a series of
hypothesis, if-then logic statements, or visual models
 A second way is to state a theory as a series of if-then statements that explain why one
would expect the independent variables to influence or cause the dependent variables

42
Theories in Quantitative Research
 Visual Picture:
 Indicate the strength of the relationship

among variables by inserting valence

signs on the paths. Use positive or
negative valences that postulate or infer
relationships.
 Use two-headed arrows connected to

show unanalyzed relationships between

variables not dependent upon other
relationships in the model
 Research productivity among faculty in
pharmacy schools

43
Theories in Quantitative Research

44
Qualitative Theory Use
 It appears as a deductive explanation, as a theoretical standpoint, and as an end point for a study
 Variation in Theory Use in Qualitative Research
 First, much like in quantitative research, it is used as a broad explanation for behavior and attitudes, and

it may be complete with variables, constructs, and hypotheses.

 Second, researchers increasingly use a theoretical lens or perspective in qualitative research, which

provides an overall orienting lens for the study of questions of gender, class, and race (or other issues of
marginalized groups).
 In critical ethnography, for example, researchers begin with a theory informing their studies.

45
Qualitative Theoretical Perspectives
 Some of these qualitative theoretical perspectives available to the researcher are as follows:
 Feminist perspectives view as problematic women’s diverse situations and the institutions that frame those situations.
Research topics may include policy issues related to realizing social justice for women in specific contexts or knowledge
about oppressive situations for women.
 Racialized discourses raise important questions about the control and production of knowledge, particularly about
people and communities of color.
 Critical theory perspectives are concerned with empowering human beings to transcend the constraints placed on them
by race, class, and gender.
 Queer theory is concerned with cultural and political means, and conveys the voices and experiences of individuals who
have been suppressed.
 Disability inquiry addresses understanding this population’s sociocultural perspectives allowing them to take control
over their lives rather than a biological understanding of disability.

46
Theories in Qualitative Research
 Distinct from the deductive use and theoretical standpoint
orientation, qualitative studies apply theory as a theoretical end
point for a study. This theoretical end point means that qualitative
study ends with a theoretical model, a theory, or a conceptual
model. Qualitative research is an nductive process building from
data, to broad themes, and to a generalized model or theory.

47
Theories in Qualitative Research
 Propositional generalization—the researcher’s summary of interpretations and claims—to which is added
the researcher’s own personal experiences, called “naturalistic generalizations”
 “Pattern theories” are explanations that develop during naturalistic or qualitative research. Rather than the
deductive form found in quantitative studies, these pattern theories or generalizations represent
interconnected thoughts or parts linked to a whole.
 Our research tips on theory use in a qualitative proposal are as follows:
 Decide if theory is to be used in the qualitative proposal.

 If it is used, then identify how the theory will be used in the study, such as an up-front explanation, as an

end point, or as a transformative-advocacy lens.

 Locate the theory in the proposal early in the study or at the end.

 Locating the theory in qualitative research

 How theory is used affects its placement in a qualitative study. In those studies with a cultural theme or

a theoretical lens, the theory occurs in the opening passages of the study.

48
Mixed Methods Theory Use
 In contrast to the use of theory in qualitative research, theories are discipline-based or social justice
theories in mixed methods research.
 There are two types of theory use in mixed methods studies: discipline-based and social justice.
 Discipline-based theory drawn from the social, behavioural, or health science literature. These theories

typically identify variables or constructs, interrelate them, and form and explain expected results for
individuals and organizations.
 Social justice theory, such as feminist research, racial or ethnic research, disability research,, lifecycle

research, and the like.

 A social justice theory perspective is incorporated into a mixed methods study in these ways:
 Indicating a framework in the opening passages of a study (e.g., feminist, participatory)

 Mentioning this framework early in a study—that it relates to a marginalized or underrepresented

community and specific issues faced by that community (e.g., oppression, power).
 Lodging this framework within a theoretical body of literature, such as feminist literature or racial

literature.
49
Mixed Methods Theory Use
 A social justice theory perspective is incorporated into a mixed methods study in these ways:
 Involving the community of interest in the process of research (e.g., in the data collection).

 Taking a stand with the research question—advocating in its orientation (e.g., inequality does exist and

the research will set out to substantiate it).

 Advancing in the design the collection, analysis, and integration of both quantitative and qualitative

methods within the transformative framework.

 Talking about your experiences as a researcher and how your experiences and background shapes your

understanding of the participants and issues under study.

 Ending the study by advocating for change to help the population under study and the issue.

50
Mixed Methods Theory Use
 It is important to use in mixed methods research because theory:
 Provides a framework for organizing and reporting the quantitative and qualitative data

 Brings together separate investigations or studies and reports common concepts or ideas

 Develops a common understanding among research team members, stakeholders, and funder of

research about the underlying framework for a study

 Provides a prediction for what the researcher hopes to learn

 Links concepts and ideas, often in a temporal order; and

 Provides a visual model for understanding the overall direction of a study

51
Questions to Think
 Which one comes first: theory or data?
 How do you differentiate between universal theory and local knowledge?
 What is the difference between cross-sectional and longitudinal study?
 What is reverse causality in comparative studies?

52