RESEARCH METHODS

DATA TYPES, VARIABLES AND

ANALYTICAL FRAMEWORK
 Session Objectives

At the end of this session, participants should be able

to:

• Distinguish the different types of variables.

• Formulate and operationalize variables.

• Come up with an Analytical Framework for their

research.

2
 Introduction
 After having formulated and analyzed the problem
we want to investigate, we must identify the data we
are going to collect in our study to meet our
objectives.

 We also want to know whether certain factors do indeed

influence the problem, and to what extent.

 To find association between problems and contributing

factors, it is essential that we carefully define the
problem itself, as well as each of the factors identified
when analyzing the problem.

 WE DO ALL THIS BY FORMULATING VARIABLES.

3
 Formulating Variables
 A variable is a characteristic of a person, object, or
phenomenon which can take on different values or can
change, such as ‘gender' and can be a subset of a
construct or a property being studied.

 Attributes are sub-values of a variable, such as 'male'

and 'female‘ i.e. the different forms/values a variable can
take.

 Variables can be :
– NUMERICAL – characteristics in the form of numbers
(e.g. age, distance, monthly revenue) or
– CATEGORICAL - non-numerical characteristics (e.g.
race, gender, colour, marital status, level of
education).
4
 Numerical Variables

 Numerical variables can either be continuous or

discrete.

 Continuous. These are variables in which numbers form

a series and are uninterrupted. With this type of data, one
can develop more and more accurate measurements
depending on the instrument used, e.g.:

 £ to $ exchange rate in dollars (1.5:1 or 1.546:1 or 1.5462:1)

 Discrete. These are variables in which numbers are

separate and distinct and can only have full values, e.g.:

 number of customer visits per week (0, 1, 2, 3, 4, etc). 5

 Categorical Variables
 Categorical variables are variables that are not quantitative
and can either be nominal or ordinal.

 Nominal variables. The attributes in these variables are

labels for classification.
e.g. Sex: male, female
Religion: Christian, Moslem, Hindu, Buddhism, etc.
Type of institution: commercial bank, discount house,
investment bank
 Ordinal variables. These are grouped variables that are
ordered or ranked in increasing or decreasing order:
e.g. High income (above $2000 per month);
Middle income ($1000-$2000 per month); and
Low income (less than $1000 per month).

Agreement with a statement:

strongly agree agree neutral disagree strongly
disagree 6
 Types of Variables

 Dependent variable – DV (criterion variable)

 Independent variable - IV (predictor variable)

 Extraneous variable
– Control variable
– Moderating/Confounding variable
– Intervening variable
 Background variable

7
 Types of Variables
1. Dependent Variable (response variable)

• This is the variable of primary interest to the researcher.

• It is used to describe or measure the problem/response
under study.
• It is the presumed effect.
• The dependent variable is what is affected
by/dependent on the independent variable.
• It changes as a result of the independent variable being
changed, and is put on the y-axis in graphs.
• The researcher’s goal is to understand and describe the
dependent variable, or explain its variability, or predict it.

8
 Types of Variables

2. Independent Variable (explanatory variable)

• Influences the dependent variable positively or negatively.

• It is the variable that is used to describe or measure the

factors that are assumed to cause or at least to influence the
problem.

• In fact the independent variable is what the

researcher manipulates. In graphs, it is put on the x-axis

• It is the presumed cause.

• It is independent of everything that occurs during the

experiment because once it is chosen it does not change.

9
 Dependent or Independent

 It is therefore important when designing an analytical

framework to clearly state which variable is the
dependent and which are the independent ones based on
the problem statement and objectives.

– if a researcher was to investigate the relationship between

salaries and educational levels in the financial sector,
salaries would be the dependent variable while educational
levels would be the independent variable.

10
 Types of Variables
3. Extraneous Variables
 Extraneous variables are undesirable additional
variables which could provide alternative explanations on
conclusions.

 They are variables that compete with the independent

variable in explaining the outcome.

 They influence the result of a study when they are not

controlled.

 Become control variables when built into the study.
However before controlling a variable one should check on
its relevance to the study.

11
 4 Extraneous Variables
3(1). Control Variable
 In experiments, control variables are those variables that are not
changed throughout the trials because the researcher is not
interested in the effect of that variable being changed for that
particular experiment.

 It is critical in experiments that these variables do not vary

and hence bias or otherwise distort the results.

 If a certain variable is suspected to influence the research

results, it should be controlled.

 A control variable increases the validity of data, eliminates

the effect of a moderator variable and,

 Leads to more convincing generalizations.

12
 Control of Extraneous Variable

 Control of extraneous variables is a fundamental

condition to causal interpretations of research (Johnson,
2001).

 Experiments control extraneous variables directly.

 The following are strategies of controlling extraneous

variables.
1. build the variable into the study - include it as an
independent variable.
2. a potentially influential variable is kept the same for all
subjects in the research – the variable is held constant.
3. remove the effects of the variable by statistical procedures
e.g., use of analysis of covariance.

13
 4 Extraneous Variables
 3(ii)Moderating/Confounding Variable
 An extraneous variable that can wholly or partially influence an
underlying true association.
 It has a strong contingent effect on the independent-dependent
variable relationship.
– associated with the problem and with a possible cause of the
problem.
– may either strengthen or weaken the apparent relationship between
the problem and a possible cause.
 May be identified, but if a variable is accounted for it is a
moderator variable,
NB: Whenever possible, avoid confounding variables because they
muddle and confuse any kind of causal assertions.

14
14
Moderating or Confounding Variable

Example: Performance

Marketing Profitability
Effort

INDEPENDENT VARIABLE DEPENDENT VARIABLE

Economic
Environment

MODERATING/CONFOUNDING VARIABLE 15
 4 Extraneous Variables
3(iii) Intervening (Mediator) Variable
 Intervening variables are variables that occur between two
other variables. Thus, an intervening variable is part of a
causal chain.
 It surfaces between the time the independent variable
operates to influence the dependent variable and the time
the impact occurs on the dependent variable.
 Surfaces as a function of the independent variable, and
helps explain the influence of the independent variable on
the dependent variable.
– may be identified, if not accounted for it is an
intervening variable.
 Causal relationship can be represented as:
Independent variable  Intervening variable Dependent
variable
EXAMPLE 1: educational level---> occupation---> income
level 16
 Types of Variables

4. Background Variables

In almost every study, background variables exist.

Such as;
– age, sex, ethnicity, educational level, schools attended,
marital status and religion, should be considered.

– these are often related to a number of independent

variables, so that they influence the problem indirectly
(hence they are called background variables).

 Only background variables important to the study should

be measured.

17
 Operationalising Variables
 We cannot directly study theoretical constructs, in most
cases.

• We need to define the variable with respect to the manner in

which we will study it. We call this an ‘operational definition’
and it allows us to indirectly study the variable.

 Operationalising variables means making them ‘measurable’

in line with objectives of our own study.

 When we define a variable using the actual measures that we

employ in our studies, this is an operational definition.

 When we define a variable in terms of its ordinary construct

or meaning this is a conceptual definition.

18
 Operationalising Variables
1. To analyse the relationship between customer satisfaction
and cost of a car.

The variables ‘customer satisfaction’ and ‘cost’ cannot be

measured directly.
• You would need to identify a series of factors or proxy variables
which form an indicator index for the level of satisfaction of
customers and others for variable cost.
For customer satisfaction we may consider attributes such as:
• Reliability
• Comfort
• Durability
• Safety
• Extras (air condition, airbags, radio etc)

For cost of car we may consider attributes such as:

• Original purchase price
• Maintenance cost
• Fuel consumption level

19
 OPERATIONALISING VARIABLES…

 To ensure that everyone (researcher, data collectors, and

readers of the report) understands exactly what has been
measured and to ensure consistency in the measurement,
it is necessary to clearly define the variables and
attributes of variables.

 For example, to define the indicator ‘maintenance cost’

of a car, it is necessary to decide what will be considered
‘costs’ e.g., is it car serving costs only or should include
insurance and security related costs of the car, how this
will be measured?

20
Example of a Framework for Defining
Variables

21
21
Examples of Variables with Different Options for
Indicators

22
22
 Analytical Framework

 It emanates directly from objectives/research problem and

consequently influences one’s methodology.

 Summarizes a systematic combination of variables that

one intends to analyze in their own study.

 This is framework which depicts the variables indicating

the causal relationship among them.

 It also indicates the direct and indirect (working through

other variables) potential relationship between
independent and dependent variable.

 It guides the researcher on proposed analysis of variables.

 Used in analytical studies. 23

 Factors Influencing Collection of
Revenue
AN ANALYTICAL
Age FRAMEWORK
Salary

Working System:
Data on Collect-
taxpayers. Ion of
Sex Experience Motivation
Enforcement revenue
of law

Specialized Training
Education education

24