See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/351080413

Variables in Social Science Research

Article · April 2021

CITATIONS READS

2 9,598

2 authors:

Loveleen Kaur Ritu Mittal

Punjab Agricultural University Punjab Agricultural University
9 PUBLICATIONS 8 CITATIONS 52 PUBLICATIONS 76 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Ritu Mittal on 24 April 2021.

The user has requested enhancement of the downloaded file.

64 Indian Res. J. Ext. Edu. 21 (2&3), April & July, 2021

Variables in Social Science Research

Loveleen Kaur1 and Ritu Mittal2

1.Ph.D. Scholar, 2.Asstt. Prof., Department of Extension Education and Communication Management,
Punjab Agricultural University, Ludhiana, Punjab
Corresponding author e-mail : [email protected]
Paper Received on February 08, 2021, Accepted on March 20, 2021 and Published Online on April 01, 2021

ABSTRACT
In social science research, a study is framed using various variables. Variable is an entity that can take different
values across individuals and time. It is a symbol to which numerals or values are assigned. They are the
characteristics or conditions that can be observed, manipulated or controlled by the researcher. The variable that
really varies provides information to the research situation in detail. It is important to have the understanding of
the variables in order to use them properly and to discover relevant and meaningful results from it. It is also helpful
to understand and evaluate their application in other studies. Thus the present paper has been prepared for detail
ed explanation the concept of variable and their use in social sciences research.
Keywords: Variables; Independent variable; Dependent variable; Social science;

I n the process of formulating a research problem, behaviors of children. These behaviors are associated
there are two important considerations: the use of with the mastery of school tasks­­­­, reading words,
concepts and the construction of hypothesis. doing arithmetic problems, drawing pictures and so on.
Researchers once through with the formulation of The various observed behaviours are put together and
hypothesis, have to proceed further to test them so as expressed in a word ‘achievement’ (Kerlinger, 1995).
to accept or reject them. It is always necessary to use Concepts are not inherited in the nature itself but
concepts or properties that permit measurement or are man-made. They are the mental constructs reflecting
observation. These properties which are to be studied a certain point of view and focusing upon certain aspects
are called variables. A variable as the name implies is of phenomenon while ignoring others. For example,
something which varies. It may be weight, height, anxiety modernization, development, growth etc. are all concepts
levels, income, body temperature etc. having different meaning (Ahuja, 2015).
In social science research variable play an important Construct : According to Kerlinger (1995), a
role in increasing the clarity of research problem. Variable construct is a concept with the added meaning of having
helps to avoid subjectivity and to bring about true picture been deliberately and consciously invented or adopted
of events or phenomena or behaviour which the social for a specific scientific purpose.
science researchers are dealing with. There are two more A construct are theoretical creations that are based
terms, concepts and constructs, which are related to on observations but which cannot be observed directly.
variables. These are discussed as follow: It is an abstraction from reality, selecting and focusing
Concept : A concept is an idea expressed in words. It on certain aspects of reality and ignoring others (Mangal
expresses an abstraction formed by generalization from and Mangal, 2013 & Ahuja, 2015).
particulars. Concept expresses an abstract idea, mental In simple words, when a concept is operationalized
image or perception. ‘Weight’ is a concept; it expresses for measurement, it becomes a construct. A construct
numerous observations of things that are more or less is usually designed for particular research purpose so
‘heavy’ or ‘light’. For example, achievement is an its exact meaning relates only to the context in which it
abstraction formed from the observation of certain is found. For example, achievement is a construct which
Indian Res. J. Ext. Edu. 21 (2&3), April & July, 2021 65

include good OCPA, getting scholarship and getting a job. and a variable. Concepts are mental images or
Variable : As already discussed, variable is something perceptions and therefore their meanings vary markedly
which varies. It is a symbol to which numerals or values from individual to individual, whereas variables are
are assigned. Variable is a characteristic that is common measureable. Concepts are subjective impressions; their
to a number of individuals, groups, events, objects etc. understanding may differ from person to person. It is
Thus, age (young, middle-aged, old), income class (lower, therefore important for the concepts to be converted
middle, upper), caste (low, intermediate, high), education into variables as they can be subjected to measurement
(illiterate, less educated, highly educated) etc., are all (Kumar, 2015).
variables. The individual cases differ in the extent to Sometimes variable and construct are used
which they possess the characteristics (Ahuja, 2015). interchangeably. But both the terms have differences.
Variable is an entity that can take different values Variable is a measurable representation of an abstract
across individuals and time. For example, age can be construct and therefore variable is measurable and
considered a variable because age can take different construct cannot be measured directly. Construct is
values for different people or for the same person at theoretically defined concept and is scientific and theory
different time. When a characteristic has only one value, specific whereas variable is a quantity that can vary.
it is a constant, not a variable. Constructs are conceptualized at theoretical plane while
It is not unusual to see some confusion between variable are measured at empirical or observational level
variables and the attributes or categories of which they (Raiphea, 2015).
consist. ‘Gender’ is a variable consisting of two categories Importance of variables : The role of variables in
of male and female. ‘Income’ is a variable consisting of extension research is as follow:
different categories of poor, middle class and rich people. • The variable that really varies provides information
The researcher has to be clear of this distinction between to the research situation in detail.
variable and category (Ahuja, 2015). • Variables and their classification provide mass of data
Variables are the characteristics or conditions that for establishing relation up to the prediction point.
can be observed, manipulated or controlled by the • By clear understanding and identification of
researcher. The measurement of variables may be research variables, researcher can strengthen the
different from everyday notions of measurement such quality of the research.
as weight and temperature. Measurements of variables • In addition a thorough understanding of these
can involve merely categorization (e.g. sex, country etc.). variables and their interactions help us to make
Examples of important variables in sociology, psychology research more useful.
and education are: sex, income, education, social class, Classification of variable : Following classification of
occupational mobility, anxiety, religion, political variables are generally used in research :
preference, achievement, age etc. (Kerlinger, 1995).
Independent and dependent variables : The variables
Definitions: which are well known to the researchers and the most
• A variable is a property that can take on different frequently used in any empirical research are the
values (Kerlinger, 1995). independent and dependent variables.
• Variables are attributes or qualities which exhibit An independent variable is the presumed cause of
differences in magnitude and which vary along some the dependent variable, the presumed effect. When we
dimension (Dooley, 2008). say A causes B, it means A is independent variable and
• A variable can be regarded as some kind of B is dependent variable. Independent variable is one which
yardsticks that gives us a basis for the evaluation explains or accounts for variation in dependent variable.
of the single unit of analysis (Galtung, 1967). The independent variable is antecedent and the
• Variable is an aspect of a testing condition that can dependent variable is the consequent. The independent
change or take on different characteristics with variable is the variable manipulated by the experimenter
different conditions (McBurney, 2007). to see its effect on dependent variable. For example, if
Difference between concept, construct and variable investigator wants to study the effect of teaching
: Measurability is the main difference between a concept methods upon the classroom achievements of students
66 Indian Res. J. Ext. Edu. 21 (2&3), April & July, 2021

then the teaching methods constitute the example of population i.e. same gender, similar qualification, similar
independent variable and classroom achievements as range of land holding etc. By this method the effect of
dependent variable. Similarly, in the study entitled extraneous variable can be nullified. For example, in a
‘Accessing the level of effectiveness of extension study ‘Information seeking behaviour of women
training programmes for enhancing core competencies regarding personal health and hygiene practices’ entitled
of extension personnel’, independent variable is was conducted by Kaur et al, 2019. As gender can
extension training programme (Aishwaryaet al, 2019). impact the information seeking behaviour of respondents,
Independent variable is also called experimental, thus the effect of gender as an extraneous variable was
manipulated, treatment, cause and input variable. Other removed by including only women as respondents in
names of dependent variable are measured, response, the study.
effect and outcome variable. Through randomization : The sample should be selected
A variable which is independent in one study can randomly and whenever possible one should randomly
be dependent in another. Take the case of relation assign subjects to either experimental group or control
between farmer’s income and availability of water. If group. As the subjects are assigned randomly to different
we take income as dependent variable and water groups, the greater is the possibility of being equal.
availability (for irrigation) as independent variable, the Building the extraneous variable right into design :
relation between two can be shown as: higher the To build the extraneous variable right into design as an
availability of water, higher would be the income and independent variable is also one way to control
vice versa. But, if we want to show the relationship extraneous variable. For example, if gender is the
between income (independent variable) and quality of extraneous variable, it can be controlled by taking the
life (dependent variable), we may say: higher the income, subject of both genders and then isolating the effect of
higher the quality of life (or living standard). In the first both, separately. It will also provide additional research
study income is the result and in the second, study it is information. As it was done in the study entitled
the cause (Ahuja, 2015). ‘Information and communication technologies (ICT) use
Extraneous variable : Besides independent variable by the students of CCSHAU, Hissar’ conducted by
several other factors operating in real life situation may Malik and Godara (2020) where they took both male
affect changes in dependent variable. These factors, and female students and studied the variation in their
not measured in the study, may increase or decrease level of ICT use.
the magnitude or strength of relationship between Matching the groups : If we match the experimental
dependent and independent variables. Extraneous and control groups including the extraneous variable and
variables are those variables that are not related to the then study its effect, effect of the extraneous variable
purpose of study but may affect the dependent variable. is controlled (Malaviya, 2004)
It is therefore essential that extraneous variables are Intervening variable : A variable which is hypothesized
controlled. to exist but cannot be observed and is presumed to occur
Suppose, an investigator is interested in studying to explain the relationship between the independent and
the efficacy of method of instruction on the achievement dependent variables is called intervening or hidden
scores (dependent variable) of some trainees. The variable. It is also called latent variable.
methods to be evaluated are lecture, seminar and According to Kerlinger (1995), the constructs
discussion (independent variable). The investigator which are non-observable, have been called intervening
discover that the achievement scores i.e. the dependent variables. It cannot be seen, heard or felt. It is inferred
variable is positively correlated with intelligence (an from the behaviour. Motivation, creativity, achievement,
extraneous variable) of the subject (trainees). Thus the tiredness, boredom etc. are intervening or latent
variable intelligence (not of direct interest to the variables.For example; if we study the effect of teaching
investigator) needs to be controlled (Kumar, 2015). methods on classroom achievement of students of
The researcher can minimize or nullify the effect B.Sc.In this association between teaching methods and
of extraneous variable by following ways: classroom achievement needs to be explained. Other
Selecting the subjects from homogeneous population variables intervene such as anxiety, fatigue, motivation
: The subjects may be selected from homogeneous etc. are intervening variable.
Indian Res. J. Ext. Edu. 21 (2&3), April & July, 2021 67

In research terminology, Actually, we cannot assume that the measured

• Change variables are called independent variables. (attribute) and manipulated (active) anxieties are the
• Outcome/effect variables are called dependent same. We may assume that both are anxiety in a broad
variables. sense, but they are certainly not same (Kerlinger, 1995).
• Unmeasured variables affecting cause and effect Moderator and control variable : Moderate variable
relationship are called extraneous variables. is secondary independent variable which is selected for
study to determine if it affects the relationship between
• Variables those link cause and effect relationships
the primary independent variable and the dependent
are called intervening variables.
variable. Control variables are those which may affect
Active and attribute variable : Those variables that
the relationship between the independent variable and
can be manipulated, changed or controlled are called
the dependent variable, and which are controlled (effect
active variables. Award of prizes, giving punishment etc.
cancelled out) by eliminating the variable or holding the
are examples of active variables. Those variables that
variable constant. For example, if we take a hypothesis:
cannot be manipulated changed or controlled and that
“widow’s attitudes towards remarriage are related to
reflect the characteristics of the study population are their socio-economic background.” Here widow’s
called attribute or assigned variables. Age, sex, education attitudes are dependent variable and their socio-
and income are attribute variables (Kumar, 2015). economic background is independent (primary) variable.
For example, if a study is designed to measure the It is possible that widows having children or no children
relative effectiveness of three teaching modules (module may also affect attitude towards remarriage. Thus, the
A, module B and module C). The structure and content third factor ‘widows with children’ (secondary
of these could vary and any module might be tested on independent variable) can also affect their attitude
any population group. The content, structure and towards remarriage.
testability of a module on a population group may also The difference between a control and a moderator
vary from researcher to researcher. As a researcher variable is that the effect of control variable are
does have the ability to control or change the teaching minimised, eliminated or held constant while the effects
modules, he can decide what constitutes a teaching of the moderator variable are studied. Since both control
model and on which group of the student population it and moderator variable are independent variables, it is
should be tested. Hence it is an active variable. On the up to researcher to determine the independent,
other hand, a researcher does not have any control over moderator and control variable.
characteristics of the student population such as their Qualitative and quantitative variables : The qualitative
age, sex or motivation to study. These characteristics variables refer to those which cannot be manipulated
of the study population are called attribute variables. after the research is started and which consists of
This active-attribute distinction is general, flexible categories that cannot be ordered in magnitude.
and useful. Some variables are by their very nature Variables such as colour, sex, religion etc. are of
always attributes, but other variables that are attributes qualitative type. They are ready-made and descriptive
can also be active. This characteristic makes it possible in nature. Qualitative variables are also known as
to investigate the same relations in different way. For organism variables i.e. the variable of the organism.
example, we can measure the anxiety of subjects. Quantitative variables refer to those variables
Anxiety is in this case an attribute variable. But we can which are composed of categories that can be ordered
manipulate anxiety, too. in magnitude i.e. it may exist in greater or smaller
We can induce different degree of anxiety, for amounts. Examples of quantitative variables are age,
example, by telling the subjects of one experimental income, size of land holding, size of group, intelligence
group that the task they are about to do is difficult, that etc. With quantitative variables, precise measurements
their intelligence is being measured, and that their future are possible because they can easily be ordered in terms
depends on the scores they get. The subjects of another of increasing or decreasing magnitudes. For example
experimental group are told to do their best but not to age of the members of young farmers club can be
strain themselves too much; the outcome is not too arranged on a continuum or scale in order of magnitude
important and will have no influence on their futures. from young to old (Mulay, 1980).
68 Indian Res. J. Ext. Edu. 21 (2&3), April & July, 2021

Continuous and discrete variable : Continuous data because it throws information away.
variables are divisible into smaller and smaller units like Stimulus and response/behaviour variable : A
age, income of farmers, height etc. The characteristic stimulus variable is the condition or manipulation created
of a continuous variable is that, within whatever limits by the researcher so as to evoke a response in an
its values may range, any value is possible. There is no organism. The general class of things that researchers
gap in the scales of a continuous variable. They can observe that relate to the environment, situation or
take on any value on the scale on which they are conditions of stimulation are referred to as stimulus
measured. For example, age can be measured in years, variable.The stimulus variables in extension research may
months and days. Similarly, income can be measured in be items like a slide show about a new crop variety, a
rupees and paisa. field day, method demonstration etc. For example in the
A discrete variable is one which involves counting study entitled ‘Role of frontline demonstration on chick
the number of events. The number of children in a family, pea production in district Ramgarh of Jharkhand’, frontline
the number of females in a particular state and the demonstration is stimulus variable (Raghavet al, 2021).
number of inhabitants in each village are some of the Any kind of behaviour of the respondent is called
examples of discrete variables, so discrete variable behavioural variable. This refers to some action or
consists of only whole numbers. In other words, a response of an individual. Behaviour is predicted from
discrete variable is one which can take only certain the response and so a behaviour variable is also known
values and none in between. For example, Number of as the response variable. At one extreme, these actions
students in a class is a discrete variable. We can count may consist of relatively simple responses such as yes
the number of students in a class. The values can be or no, true or false for a particular question. The
19, 20, 21 or 40. But it cannot be 19.5 or 36.7. There is behavioural variable in Extension may be frequency with
no value between 19 and 20. which a particular event occurs or it may be the scale
Dichotomous and polytomous variable : When a value of a particular event.
variable can have only two categories or values as in Important considerations for selection of variables:
yes/no, good/bad and rich/poor, it is known as a While selecting variables the researcher should keep
dichotomies or dichotomous variable. Some of the the following points in mind:
variables used in behavioural research are true Review of literature : The researcher should carefully
dichotomies i.e. they are characterised by the presence review the literature before deciding the variables for
or absence of a property: alive/dead, employed/ research. By reviewing the literature he becomes aware
unemployed. Other examples of two-valued variables are: of important and unimportant variables in the concerned
teacher/non-teacher, viewers/non-viewers of television, area of research. A careful review also helps the
etc. Polytomous variable have more than two categories researcher in selecting variables lying within the scope
or values. For example, religion (hindu, sikh, muslim, of his interest in defining and operationalizing variables
christian, other), of extension managerial skills director and in identifying variables which are conceptually and
(high, somewhat high, medium, low), agreement (Strongly practically important.
disagree, disagree, neutral, agree, strongly agree) etc. Use of intensive method : In this method after reviewing
Most variables are however, theoretically capable the literature, the investigator selects the relevant
of taking on continuous value. It has been common variables. These variables are then refined by asking
practice in behavioural research to convert continuous other investigators. After that those variables which
variables to dichotomies or polytomies. For example, obtained high consensus are selected for research study.
intelligence, a continuous variable, has been broken down Theoretical and logical basis
into high and low intelligence or into high, medium and There must be theoretical and logical basis to
low intelligence. determine cause and affect relationship between
It is not possible to convert a truly dichotomous independent and dependent variables.
variable such as sex to a continuous variable but it is Research design consideration : The variable must be
possible to convert a continuous variable to a dichotomy selected in view of the scope of designs of the experiment.
or a polytomy. This conversation can serve a useful The variable selected as moderator variable and the
conceptual purpose but is a poor practice in analysis of extraneous variable must be such that it fulfils the basic
 Indian Res. J. Ext. Edu. 21 (2&3), April & July, 2021 69

requirement of the research design. The independent he should also keep in view the financial resources
variable must be such that its manipulation can be easily available for the purpose. The time consideration should
done within the framework of the research design. also not be ignored. He should also see the nature of
Avoid repetition : If relationship has already been the variable because some variables are easy to study
established in previous studies, repetition must be and some are difficult.
avoided. In particular, we may choose to study the same
CONCLUSION
dependent variables while manipulating new independent
variables. Like in extension research, relationship of Variable in social science research can be defined
background variables with adoption behaviour, in different ways. But in common, it can be defined as
communication behaviour, academic performance and a measurable value that varies across individuals and
aspirations of respondents has already been established, time. It is also different from concept and construct
so it should be tested after a period of time with due and measurability is the main difference.
consideration to new independent variables. Variables play an important role in research,
Select the variables having quantitative without variables it becomes difficult to find out the
characteristics : If possible, select variables which relationships and predictions. The variables should be
have more quantitative characteristics because with the outlined in the introduction of the paper and explained
quantitative variables, precise and accurate in more detail in the methods section. Mostly emphasis
measurements are possible and these can be ordered in is led on dependent and independent variables. The effort
terms of increasing or decreasing magnitude. should be made to classify the extraneous and
Practical consideration : A researcher should also take intervening variables, so that their effect on the
into account some practical considerations in selecting dependent variables can be studied. Therefore,
variables; he should limit the number of variables to be researcher should have clear understanding of the
incorporated in the study because it is not possible to variables and should select only those variables which
study too many variables at a time. In selecting variables have direct bearing on results of the study.

REFERENCES

Ahuja R. (2015). Research Methods. Pp. 52-68. Rawat Publications, Jaipur, India.
Aishwarya S., Wason M., Padaria R.N., Rao D.U.V., Gills R., Priyadarshani P. and Gurug B. (2019). Accessing the level of
effectiveness of extension training programmes for enhancing core competencies of extension personnel: An analytical
study in Kerala. Indian Res J Ext Edu, 19 (1): 73-76
Dooley(2008)Social Research Methods. Pp. 318-319. Prentice Hall of India, New Delhi, India.
Galtung J.(1967) Theory and Methods of Social Research.Pp. 72-84. Columbia University Press, New York.
Kaur L., Kaur S. and Sharma P. (2019) Information seeking behaviour of women regarding personal health and hygiene
practices. Indian Res J Ext Edu, 19(2&3): 52-55.
Kerlinger F. N.(1995)Foundation of Behavioral Research.Pp. 26-41. Prism Book Pvt Ltd, Bangalore, India.
Kumar R.(2015)Research Methodology.Pp. 55-66. Pearson India Education Services Pvt Ltd, Noida, India.
Malaviya R. (2004)Research Methods-An Insight.Pp. 106-156. Rainbow Communication, Delhi, India.
Malik A.K. and Godara A.K. (2020) Information and communication technologies (ICT) use by the students of CCSHAU,
Hissar. .Indian Res. J. Ext. Edu., 20(4): 14-19.
Mangal S.K. and Mangal S. (2013) Research Methodology in Behavioural Sciences.Pp.96-97. PHI Learning Private Ltd, New Delhi,
McBurney D.H. and White T.L. (2007). Research Methods. Pp. 119-135. Akash Press, Delhi.
Mulay S. andWhite L. (1980) Research Methods in Extension Education. Pp. 69-79. Manasyan Publishers, New Delhi, India.
Raiphea Y.P.(2015)Variable: classification, measurement and importance in socialscience research.JBM&SSR4: 435-438.
Raghav D.K., Indrajeet, Kherwar D., Kumar A, Singh A.K. and Chauhan J.K. (2021) Role of frontline demonstration on chick pea
production in district Ramgarh of Jharkhand. Indian Res J Ext Edu, 21 (1): 30-34.



View publication stats