Business Research Methods

UNIT ONE
THE CONCEPT OF RESEARCH
AIMS AND OBJECTIVES
At the end of this unit, you should be able to

 define the term research
 list and describe the objectives of research
 list and describe the attributes of research
 distinguish between research and common sense
1.1 INTRODUCTION
This unit attempts to explain the concept, objectives, and characteristics of research.
1.2 SCIENTIFIC METHODS OF RESEARCH
All of us today realize the tremendous importance of the findings of ‘Science’ or ‘Scientific
Research’. Scientific developments have brought great changes in our way of life. Scientists
of different fields are constantly engaged in the advancement of knowledge of their own
areas. In spite of the tremendous influence of science upon modern civilization, there exists as
yet no standard definition of science. Therefore, it is essential to define the term ‘science’
before we go to the procedures of scientific research.
1. Definition of science
‘Science’ is a systematized knowledge derived through observation, experimentation or any
other systematic method or procedure in order to determine the nature of the phenomena
being studied or the causes behind specific events.
The above definition of science considers it both as
I. A body of knowledge and
II. A system of procedures and methods.
These two aspects of science should be dealt with in further detail.
1
I. Science as a body of knowledge
Science has often been taken to refer to all sorts of rigorously obtained and ordered bodies of
knowledge. The scientific knowledge can be divided into several branches on the basis of
areas covered by them and on the basis of their structural differences.
(a) Rationalistic science: Philosophy & Theology.
(b) Formal Science: That deals with logic and pure mathematics.
(c) Applied Science: Medicine and Engineering.
(d) Natural Science: Physics, Chemistry and Biology.
(e) Social Science: Sociology, Political Science, Psychology, Economics and so on.
Even those persons, who take ‘Science’ as a body of knowledge, define it in different ways.
This will be clear from the following definitions:
i) Henry Poincare, a famous mathematician, defined ‘Science’ as knowledge of not of
things but of their relations.
ii) Havelock Ellis, a noted Psychologist, regards science as the search for the reasons of
things.
These two definitions are, however, not contradictory to each other. They explain the same
thing in two different ways. Knowing the relation among the different things helps to know
the reasons of things.
The wide definition of science given above includes both pure and applied subjects. Some
important methodologists in sociology take it in more specific sense, i.e., in empirical terms.
As such science to term is one sub-part of ordered knowledge which is concerned with the
empirical phenomena only, i.e., the world or phenomena which can be observed or
experimented. The statements which constitute the scientific knowledge explain the
phenomena. Such statements or explanation must satisfy a set of criterion that determines
their accuracy. Moreover, bodies of scientific knowledge consist of statements about classes
of empirical sciences, like logic and mathematics. Such statements are of different levels,
taking different forms, depending upon the number of units or the size of the sample. A body
of scientific knowledge may thus consist of such different statements, known as hypothesis,
theory and law.
2
II. Science as a system of procedures and methods
The present state of knowledge is important of course. But it is important mainly because it is
a base for further scientific operation, for further scientific theory and research. It is this
scientific operation that generates a system of procedures and methods to discover new theory
and inter-connected conceptual schemata that are fruitful for further research. Science has,
therefore, been defined also in terms of the major processes that take place within it, like
definition, identification, organization, verification, prediction and application. This sort of
definition of science indicates the actual procedures of research followed in the development
of scientific knowledge.
Here we are talking of science as an activity. This second view is called a dynamic view,
which means serving to discover or reveal. Its stress is on the problem solving activities of the
scientists. George Lundberg in his book ‘Social Research’ holds the second view of science,
i.e., considering it as a method. To him “Scientific method consists of systematic
‘Observation’, ‘Classification’ and ‘Interpretation’ of data …. Scientific generalizations are
characterized by their degree of formality, rigorousness, verifiability and general validity.”
Emphasizing the importance of the scientific methods in collection of the information about
facts, Mr. Karl Pearson in his book ‘The Grammar of Science’ says, “There is no short-cut to
truth, no work to gain knowledge of the universe except through this gateway of scientific
method”.
Many socialists, social scientists and scientists hold this second definition of science. As
between the two approaches, scientists would perhaps lay more stress on the method and the
process rather than on knowledge and the results.
As said in the beginning, science should be taken in terms of both as a body of knowledge and
also as a system of procedures and methods. ‘Knowledge’ and ‘Activities’ are the two
different aspects of science. Both taken together alone give a complete definition of science.
The two aspects of science are closely inter-related with each other. Activity is a necessary
condition for the existence and development of any kind of knowledge. In other words,
progress is possible only through conservation, transmission, and the accumulation of
knowledge which is possible only through a process of analysis and verification. The reverse
relation also exists in some respects. All scientific activities and their success depend upon
knowing the environment, which is provided by the existing scientific knowledge.
3
The difficulty encountered in attempting to define the term ‘Science’ arises mainly from the
tendency to confuse the content to science with its method. Much of the content of science is
constantly changing, so what may be scientific, i.e., accepted as true today, may become
unscientific tomorrow.
At the end it can be concluded that the different definitions given about science by many
persons, did not find any clear and comprehensive meaning. But there is a consensus among
authoritative writers with regard to the essential attributes of scientific method and process of
scientific research. On the basis of such a consensus, science may be defined quite accurately
and functionally as ‘an objective, logical and systematic method of analysis of phenomena,
devised to permit the accumulation of reliable knowledge’. The key terms of this definition
indicate what we call ‘essential criterion’ of scientific methods and will be dealt in detail
subsequently, but before that the difference between scientific knowledge and common sense
should be clearly understood.
1.3 SCIENTIFIC KNOWLEDGE AND COMMON SENSE KNOWLEDGE
Scientific knowledge is definitely different from the common sense knowledge because of the
specific systematic procedures involved in it. Before we come to the specific differences
between the two, let us take a few examples of each. Examples of common sense knowledge
or statements are given below:
1. Illness is the punishment for committing crime.
2. Economic depression is due to the Jews.
3. The earth is flat and the sun goes round the earth causing day and night.
4. Night follows the day. So day is the cause of night.
5. Bad omens, like a cat crossing the street while going to the examination hall, lead to
the failure in examination.
Examples of scientific statements are also given here to indicate the difference in their
structure and nature:
Christians are more conservative towards adoption of birth control methods than the Hindus.
1. Muslims tend to vote for congress.
2. Higher rates of crime occur among residents of slum than among residents from
developed housing areas.
4
3. Intermediate students are politically less socialized than the graduates.
Difference between scientific knowledge and common sense knowledge
Examining the nature and structure of the above groups of statements we can differentiate
scientific knowledge from the common sense one:
1. The uses of concepts, conceptual schemes and theoretical structures are strikingly different
in each of them. In common sense we find loose use of theories and concepts. As such
fanciful explanations of the phenomena, natural or human, are blindly accepted. For example
take the statements -
(a) Illness is the punishment for crime.
(b) An economic depression is due to Jews.
The scientists on the other hand systematically build their theoretical structures. They test
them for internal consistency. They put of concepts to empirical tests. The scientists realize
that the concepts are man made terms to represent reality but there are always chances of lack
of close relation between the concepts and the reality. Hence they must always be empirically
tested.
2. Besides the concepts the scientists also test the hypothesis systematically. A common man
may test his hypothesis or theory, but in a selective fashion. That means he selects those
factors or evidences, which are consistent with his hypothesis. Take statements like ‘Bengalis
are musical.’ This includes only those Bengalis who are musical but neglects, those, who are
not. These types of statements are results of preconceptions and peoples’ tendencies of
selective support. This quality has been found among armchair philosophers or social
evolutions who first formed some ideas, concepts, hypothesis and then selected those
examples which support their concepts or hypothesis.
Contrary to this the scientists take into consideration not only those variables that have been
hypothesized as the possible causes, but also those variables which are not effective or are not
related. Take the statement “Higher rates of crime occur among residents of slum than among
the residents from developed housing projects.”S
A scientist would not make above types of statements in the beginning without making
observation. He will first simply assume some sort of relationship between living conditions
and delinquency rates. He would then take into consideration delinquency rates of different
areas or of different neighborhoods, for example, of slums and developed areas. He would
5
compare them, and then only after finding empirically slum and delinquency to be related, he
will make such type of statements. His statements are thus based upon systematic
observations and recording of facts and not upon purely personal judgments. Unscientific
statements are found in the areas of both the social and the natural sciences. As for example,
unless and until through observation and logical demonstration it was established that the
earth is round and moves on its own axis around the sun, the people considered that the earth
is flat and the sun moves around the earth causing day and night.
3. A scientist not only observes the relationship between two or more variables under different
natural conditions but also tests the relationship under controlled conditions. He does not
assume any causal relationship between the antecedent event and the consequent event simply
on the basis of their occurrence after one another. He tests it empirically as well as under
controlled condition whenever possible.
A layman takes even fortuitous or accidental simultaneous occurrence of two events as causal
viz., ‘night follows the day. Therefore, day is the cause of night.’
4. Final difference between the common sense knowledge and the scientific knowledge is that
the scientist, while attempting to explain the relationship among the observed phenomena,
carefully rules out metaphysical explanation for they are not testable. It is different from
statements based upon belief.
Thus, the two types of knowledge differ in: (a) their manner of assumptions, (b) objectives,
and (c) procedures. Second and third are the most important criterion.
1.4 SCIENTIFIC RESEARCH
By the term ‘research’ is meant any enquiry or investigation regarding any phenomena or
event in order to discover facts. ‘Science’ as explained before, includes a body of knowledge
and a system of procedures. A scientific research, therefore, means ‘an investigation carried
on through systematic procedures.’ Thus, investigation carried on in the field of any science
comes under scientific research. In this sense research in social sciences is also scientific. In
the next chapter why and how of social science will be examined.
The best definition of scientific research, in the opinion of the present writer has been given
by Fred, N. Kerlinger in his book, and ‘Foundations of Behavioral Research.’ To him,
‘Scientific research is systematic, controlled, empirical, and critical investigation of
hypothetical propositions about the presumed relations among natural phenomena.”
6
1. The above definition of scientific research includes its two major characteristics. That
the scientific research is systematic and controlled, which means that scientific
investigation is ordered and the research situation is tightly disciplined. Among the
many alternative explanations of a phenomenon, all but one are systematically ruled
out. One can thus have greater confidence in the tested and established relationship
among the variables.
2. That scientific investigation is empirical, which means that it has to be tested or
checked against objective reality. It is easy to exaggerate or to over generalize. Hence
each study must be subjected to the court of empirical enquiry and test. Thus, criterion
of testing one’s hypothesis or theories systematically and empirically makes the results
of scientific research different from the no-scientific ones.
To fulfill the above two characteristics scientific research makes use of what we call the
‘scientific method’.
1. Scientific method
A. Definition
It has been emphasized before that science implies two things, a body of knowledge as well as
a system of methods.
When we talk about scientific method we are actually talking about the methodological aspect
of science, i.e., its system of procedures, followed in any investigation. Science or scientific
method has been defined differently by different persons. Some definitions spell out the steps
or procedures to be followed to make it scientific, while the others point out its characteristics.
A few of these are given below to make evident the steps or procedures of it as well as its
characteristics. George A. Lundberg in his book ‘Social Research’ has specified its
procedures and some of its essential characteristics. To him a scientific method consists of
systematic observation, classification and interpretation of data. And the main difference
between our day to day generalizations in one hand and the conclusions derived through the
application of the scientific methods on the other, lie in the degree of formality, rigorousness
verifiability and general validity of the latter.”
Another social scientist defines science in terms of its six major processes that take place
within it. “These are testing, verification, definition, classification, organization and
7
orientation which include prediction and application.” The above definitions indicate that
scientific method is a system of techniques and processes. Techniques may differ from one
science to other. But all the sciences are similar in the sense that they all aim to fulfill the
essential characteristics of scientific methods, like objectivity, reliability and validity. They all
aim to discover general laws.
From the above definition and discussion we can make the following derivations about
scientific method.
1. It is a system of procedures which are similar to some extent and different in others in
case of different sciences. However, it has some basic standardized steps or
procedures, which are observation, classification, interpretation and application.
Following these procedures, any study or investigation can be made scientific.
2. It has some specific goals, which are development of theories or laws, which provide
explanations and predictions.
3. There are different techniques for different category of sciences, viz., natural, formal,
social and so on. The techniques differ in relation to the phenomena to which they are
suited and in terms of their exactness or other scientific characteristics.
4. That there are certain essential criterions which must be fulfilled in order that any
study can be called scientific. These are objectivity, reliability and validity. Besides
these the study must also fulfill the notion of prediction and control.
B. Goals of science and scientific method
Any science or scientific study aims to explain the occurrence of events in terms of their
causes. Such explanations are general in nature in the sense that they instead of trying to
explain each and every separate behavior or characteristics of the individuals or the units, deal
with them in common.
Such explanations, when derived through scientific procedures and stated in the form of
scientific language are known as theory. Therefore, the ultimate goal of science is the
development of theories. A scientific theory provides explanation of the events by specifying
the exact relations of one set of empirical events with the others. Therefore, on the basis of
scientific theories one can predict the occurrence of events in future. Natural or physical
sciences are able to provide explanations or show relationship among the events more
specifically than the social sciences and, therefore, are able to provide more exact predictions.
8
But due to varying and changeable nature of the society or the social events such explanations
or predictions in social sciences are not that exact.
C. Essential criterions of scientific methods
The essential criterions of scientific methods are (i) objectivity, (ii) reliability, (iii) validity,
(iv) verifiability, (v) definiteness, (viS) generality, (vii) control and predictability.
(i) Objectivity
The study of any phenomena may be either subjective or objective. A study is called
‘subjective’ when it is influenced by the scientist’s individual values, feelings and beliefs. On
the contrary, any study is considered as ‘objective’ when it is based upon the objective
observation of the factual events and not upon personal judgments. Any scientific study has
also to be free from the remarks like, good or bad.
Objectivity is an essential criterion to consider any study as scientific or non-scientific.
Therefore, the scientific methods provide a rigorous, impersonal mode of procedure dictated
by the demands of logic. The criterion of objectivity is easily maintained in natural or
physical sciences, whereas it is very difficult to maintain objectivity in social sciences,
because here our units of study are human beings themselves with whom we live and share
our feelings and opinions. Besides this, people often have the tendency to consider their
culture as superior. As a result often sociological and anthropological studies have suffered
from ‘Ethnocentrisms’. By ethnocentrism is meant judging other’s culture according to ones
own culture and values.
In social sciences achieving absolute objectivity has not yet been possible. Therefore, the
social scientists are constantly engaged in improving the techniques or methods of data
collection. Besides improving the methods it is also essential to train the investigators in
techniques of observation and interview.
Given below are examples of subjective and objective statements:
Subjective statement
(a) ‘Western families are highly individualistic and disorganized.’
(b) ‘Jews are cruel and selfish.’
(c) Objective statement
(d) Muslims are more conservative than Hindus in adaptation of family planning.
(e) Birth rate in developing nations is higher than that in developed nations.
9
(ii) Reliability
This is the second essential characteristic of a scientific method. In simple language reliability
means the degree of accuracy of measuring instrument. It refers to the degree to which scores
of a test remain constant for the same unit of measurement over times. According to F. N.
Kerlinger, “Reliability can be defined as the relative absence of errors of measurement in a
measuring instrument.” It would be wrong to expect that a measuring instrument will be
perfectly free from errors. Achieving absolute reliability like the achieving of absolute
objectivity of the measuring instruments is less possible in social or behavioral subjects, like,
sociology, psychology or education. Here, therefore, the success of measurement depends
upon the degree or extent to which errors can be eliminated. Important instruments of
investigation in the areas of social or behavioral sciences are questionnaire, schedule,
interview, content analysis and case history. Their reliability depends upon the way they are
constructed and used. The extent to which they are reliable, they are dependable also.
Therefore, synonymous words, used for the term reliability are dependability, stability,
consistency, predictability and accuracy.
When the events or the nature of the phenomena or the behavior of the units vary from one
occasion to another we can say that they are of instable nature and therefore, are variable or
unpredictable. Predictability, which is a rather essential quality of a scientific study, depends
upon stability and consistency of events. The extent to which a measuring instrument has
achieved accuracy, precision or exactness is the extent to which it is dependable also.
Science strives constantly for exactness. It is not satisfied with half-truths and is intolerant of
careless procedures. In science we aspire to promote more specific description, reliable
prediction and control. Kerlinger in his book, ‘Foundations of Behavioral Research’ and
Jahoda & Selltiz in ‘Research Methods in Social Relations’ mention three different ways of
testing reliability of a measuring instrument.
(1) If repeated study of the same thing with the same or comparable measuring instrument
under the same conditions gives the same or similar result then the instrument can be
reliable.
(2) In case the instrument is reliable the difference in results will occur either when the
conditions determining or affecting the results have changed, or when the sample on
which the study is carried on is completely a different one. But such difference should
10
not be due to the measuring instrument, as for example, due to different observers or
due to difference in the mental position of the observer from time to time. This
definition indicates the quality of stability of the reliable instruments.
(3) The second way of knowing the reliability of a measuring instrument is by knowing
whether the measures obtained from a measuring instrument are true measures of the
property being measured. This is an accuracy definition of reliability. Here we are
really asking whether the measures are accurate or not.
(4) The third way of testing reliability is by measuring the exact amount of error in a
measuring instrument. Reliability can thus be defined as the relative absence of errors
of measurement in a measuring instrument.
(iii) Validity
Any measuring instrument is valid when it measures most accurately the objects or
individuals and their characteristics. The questions that emerge in relation to the validity of a
measuring instrument are: -
(1) What does it measure?
(2) Are the data it provides relevant to the characteristics in which we are interested?
(3) Do the differences in scores represent true differences in the characteristics we are
trying to measure or are the differences due to the influence of other factors?
The “best” definition of validity has been given in Selltiz and others in their book, ‘Research
Methods in Social Relations’.
“Validity of a measuring instrument may be defined as the extent to which differences in
scores on it reflect true differences among individuals, groups, or situations in the
characteristics which it seeks to measure, or true differences in the same individual, group or
situation from one occasion to another, rather than constant or random errors.
There is often a direct and close congruence between the nature of the object measured and
the measuring instrument. If we are measuring certain physical properties, like color of the
body or hair or of eyes, or are measuring some attributes that are direct and concrete in nature,
like, age, income, and level of education then maintaining validity is not a great problem. The
length of an object can be easily measured by laying of the sticks or measures that contain a
standard number system in feet or meters. Weight of any movable object can also be
measured easily. On the other hand, there are many social characteristics which are highly
11
abstract in nature and can be measured only indirectly. Examples of such characteristics are
achievement, aspiration, attitude, aggressiveness, socialization and authoritarianism. It is
difficult to measure these characteristics validly, i.e., in exact amount. Such abstract
characteristics need first to be defined in measurable terms. For example, to measure
aggressiveness we will first have to define which behaviors can be taken as aggressive and
which not. Unfortunately, there are no established scales to measure such type of neither
abstract characteristics nor definite standardized rules to develop a scale. As a result, we have
a large number of scales being used in Sociology or Psychology, but none of them have been
repeatedly tested for reliability or validity.
A valid and reliable measuring instrument can be said to be one that is able to measure the
characteristics both accurately and distinctly. That means the differences in scores on the
measuring instruments should reveal true differences among the individuals, objects or units
of measurement. The measuring instruments vary among themselves in the specificity or
exactness of measurement. Among the different techniques of data collection, observation
becomes comparatively less reliable and valid than the questionnaire and interview methods.
Types of Validity
The most important classification of the types of validity has been provided by a joint
committee of the American Psychological Association, the American Educational Research
Association, and the National Council on Measurements used in education. The four types of
validity that have been discussed by them and by many other methodologists are: (a)
Predictive Validity (b) Concurrent Validity, (c) Content Validity, and (d) Construct Validity.
Each of these will be examined briefly in the following pages. The greatest emphasis has
always been laid on construct validity, for it is considered as the most important form of
validity from the point of view of scientific research.
(a) Predictive Validity: By predictive validity is meant the extent to which the study
implies predictive power. We predict a dependent variable from an independent one.
We predict the existence or non-existence of a relation between any two variables, like
‘relation between method of teaching and students performance in the classes or
‘relation between family background and voting behavior’. Predictions do not
essentially mean predicting the occurrence of some event in future only. We may
predict some present or future state of affairs. As for example if we are studying
12
students’ class achievement on the basis of some tests, we can predict what would be
their achievement in future. We can also assess and predict as to students of which
year are the best or are likely to be the best.
(b) Concurrent Validity: Concurrent validity and predictive validity is often considered as
the same by many persons, like F. N. Kerlinger in his book “Foundations of
Behavioral Research”. But there is slight difference between the two. Tests that help to
distinguish the individuals who differ in their present state is said to have concurrent
validity, and a test which distinguishes individuals who would differ in the future is
said to have the predictive validity. Accuracy of the second one depends upon the first.
Thus, they are similar and related to each other. Both are equally important. Predictive
validity is ordinarily associated with practical problems and outcomes. Here interest is
mainly in the extent to which such studies would help to solve problems. When we use
one criterion to measure other characteristic we are faced with the problem of
adequacy of the criteria we are using as indicator. Teacher’s effectiveness in the class
may be determined on the basis of his or her knowledge of the subject, expression,
fluency of the language, capability to answer students’ questions logically in the class,
extent to which the students grasp his teachings, his capacity to maintain discipline in
the class and so on. Predictive validity depends upon choosing the test criterion or
indicators to measure each of the relevant concepts.
a) Content Validity: Content validity is the representativeness of the content. It involves
judging the representativeness of the items, i.e., how far they represent the content or
the universe of the content of the property being measured. Here one selects from the
universe of many items related to a property only those, which are considered as
representative in the context of a particular research work. As previously discussed a
teachers’ effectiveness can be measured through several different criteria. The set of
criteria can be called content of the universe ‘effectiveness.’ One has to select some
items from these according to the nature of the subject of study and the population.
Take another example. In a study of social status one would take the criteria of
income, education, occupation and sex as indicators.
b) Difficulty in determining content validity lies in the fact that it is easier to select items
from universe of content which are more obvious than others, which are not. Difficulty
13
is also faced when items are too few or when the items vary from one place to another,
as in case of the study of social status. It also becomes difficult when items are too
many. This difficulty is often faced in case of the construction of scales. Content
validity of a collection of items can always be challenged in respect of how large or
how good the contents are.
c) Construct Validity: We have seen before that the validity of the tests or measurement
depends upon the extent to which differences in scores on it reflect true differences
among units of observation. ‘Construct validity’ means determining which factors or
properties can explain the variation in the scores or say variance of the test. In other
words, which factors are responsible and, therefore, should be used to explain the
differences among individuals or groups in their test scores on a measuring instrument.
Thus, the researcher wants to locate the variables or factors which are called
constructs. All the characteristics are not measurable or identifiable directly. They are
abstract in nature and need to be defined operationally for measurement. These are
called ‘constructs’. The researcher after locating the variables or factors attempts to
trace out the nature, type and degree of relationship among different variables. For
example, in the study of political socialization and political participation by the author
it was found that scores on socialization and participation of students varied.
Attempting to trace out the factors it was found that the scores varied with level of
education of the students and their family background.
Thus in a study of any problem the scientists are interested in the measurement of the
property, in test performance of the individuals in order to categorize them in groups and to
distinguish the groups. They are also interested in the relationship between variables or
constructs. This helps in predicting the behavior of the individuals.
(iv) Verifiability
Scientific research is a continuous process of verification. The process of verification consists
of repeated study of the same thing under similar conditions in order to check the accuracy of
the conclusion or inferences drawn. If the predictions made are found to be correct or if
repeated study gives the same result then the study can be considered as accurate, reliable or
valid. In a descriptive study verification consists of corroboration of the expressed results,
generally by replication of the observations by unbiased observers. How many replicate
14
observations are needed, and how much difference among the various observations is to be
tolerated are questions that can be answered only by reference to the hypothesis.
Examples of unverifiable statements are:
(a) God is kind
(b) Democracy is good
(c) Examples of verifiable statement are:
(d) Marriage choice and religion are related,
(e) Voting behavior varies with religiously.
(v) Definiteness
Scientific method implies the quality of definiteness. Any conclusions derived through this
are always definite in nature. The scientist can test the correctness of or accuracy of
generalization drawn by him when they are definite in nature. As for example the conclusion
‘The Degree of Political Consciousness and Participation increases with age and education’ is
definite in nature and therefore can be tested. But it would be difficult to test the statements,
like, ‘change in modes of production leads to change in the society.’
(vi) Generality
Scientific conclusions are not particularistic but general in nature. They are not related to the
specific or unique characteristics or behaviors of the individual units rather are related to the
common characteristics of a class of individuals. This does not mean that the scientists are not
at all interested in individuals. They study the individuals as the elements of groups, of
classes, or of society, that supply us with information about classes or groups. As such the
generalizations drawn about them cover not only the individuals who have been studied but
also the similar individuals who were not selected in sample. The generalizations thus cover
groups or classes irrespective of place, and time.
(vii) Control and predictability
Another criterion of scientific method is control. This is perhaps, the most important criteria
in the sense that it is the degree of control which differentiates physical and natural sciences
from the social sciences. Also that it determines the fulfillment of other criteria, as, reliability,
validity, verification, predictive generality, etc. How to achieve control has been discussed in
detail in the chapter ‘Research Design.’
15
1.5 OBJECTIVES OF RESEARCH
1. Curiosity about unknown
“Curiosity is an intrinsic trait of human mind and a compelling drive in the exploration of
man’s surroundings.” It is a natural instinct in the mankind. Even a small child is curious
about the unknown objects that he notices around him and tries to understand them in his own
ways. The same curiosity drives a scientist to explore unknown factors working behind the
social phenomena. When he observes various social activities of man so complex and varied,
he simply marvels at their nature and tries to understand them in their true significance.
2. Desire to understand the cause and effect of widespread problems of human beings
According to P.V. Young – “The search for cause and effect relationship has been more
relentless than almost any other scientific effort upon which human energies have been
spent.” People want not only an account of events but want to know how they happened.
3. Appearance of Novel and Unanticipated Situations
Man is often faced with many acute and difficult problems. An ordinary person reacts
emotionally to these, but a scientist sets down dispassionately to find out their cause and thus
evolves a lasting solution to such intricate problems. In quite a large number of cases such
problems have inspired the scientist to go into their detail and study the basic factors causing
these problems.
4. Desire to discover new and test old scientific procedure as an efficient way to gain useful
and fundamental knowledge
Such research is not in fact a research in phenomena, but a research in techniques or methods
used in research. A number of such researches have been made to evolve better and most
refined techniques for dealing with human problems. Of late there has been growing emphasis
upon the use of quantitative or statistical methods in research in order to make it more definite
and mathematically precise.
1.6 SUMMARY
We have said that research is used for understanding, explaining, predicting and to some
degree controlling human behavior. It has the attributes of scientific method. It is carefully
directed, formal, systematic and intensive process, which is closed, tied to theory and to
16
theory development ultimately, however, it has a unique and specific purpose, and that is to
provide ‘information’ or knowledge.
Exercise 1
1. In your own words, define ‘research’
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
Exercise 2
1. Why do we engage in the research process?
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
Exercise 3
1. How do you characterize research?
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
1.7 ANSWERS TO CHECK YOUR PROGRESS EXERCISES
Exercise 1
In simple terms, it is an attempt to find solutions to felt, real human problems.
Exercise 2
Primarily we want to solve immediate problems.
We can also carry it out to broaden our understanding of the issue at stake.
We may like to prove or disprove an existing state of affairs, etc.
Exercise 3
There are a number of characteristics of effective research undertakings.
17
It is an objective treatment of problems.
It is a step by step approach to solving problems.
It is systematic or scientific (for example, there are well established tools of data collection,
sampling techniques, etc.) In other words, we make careful considerations and justifiable
moves in the research process.
It is controlled activity i.e. factors or variables outside the objective of the research process
are kept constant or are controlled so that they will not affect the process itself.
Exercise 4
Common sense knowledge is accepted blindly while this is not the case in research;
hypothesis in common sense is tested selectively and so on.
18
UNIT TWO
TYPES OF RESEARCH
AIMS AND OBJECTIVES
At the end of this unit you are expected to
 List and describe the major types of research
 Describe the steps followed in carrying out some of these types of research
2.1 INTRODUCTION
Although research may not always lend itself to rigid system of classification, it is usually
possible and desirable to put it into one of the many general categories
2.2 TYPES OF RESEARCH
2.2.1 Experimental Research
Experimental Research - An attempt by the researcher to maintain control over all factors that
may affect the result of an experiment. In doing this, the researcher attempts to determine or
predict what may occur.
Steps involved in conducting an experimental study
Identify and define the problem.
Formulate hypotheses and deduce their consequences.
Construct an experimental design that represents all the elements, conditions, and relations
of the consequences.
1. Select sample of subjects.
2. Group or pair subjects.
3. Identify and control non-experimental factors.
4. Select or construct, and validate instruments to measure outcomes.
5. Conduct pilot study.
6. Determine place, time, and duration of the experiment.
Conduct the experiment.
Compile raw data and reduce to usable form.
Apply an appropriate test of significance.
Essentials of Experimental Research
Manipulation of an independent variable.
An attempt is made to hold all other variables except the dependent variable constant -
control.
Effect is observed of the manipulation of the independent variable on the dependent variable -
19
observation.
Experimental control attempts to predict events that will occur in the experimental setting by
neutralizing the effects of other factors.
Methods of Experimental Control
Physical Control
. Gives all subjects equal exposure to the independent variable.
. Controls non-experimental variables that affect the dependent variable.
.Selective Control - Manipulate indirectly by selecting in or out variables that cannot be
controlled.
.Statistical Control - Variables not conducive to physical or selective manipulation may be
controlled by statistical techniques (example: covariance).
Check your Progress Exercise1
1. Define experimental research.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2. List six steps involved in conducting an experimental study.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2.2.2 Descriptive Research
Descriptive research is used to obtain information concerning the current status of the
phenomena to describe, "what exists" with respect to variables or conditions in a situation.
The methods involved range from the survey that describes the status quo, the correlation
study that investigates the relationship between variables, to developmental studies that seek
to determine changes over time.
Steps in carrying out descriptive studies
 Stating problems
 Identifying information
 Selecting tools of data collection
 Sampling
 Designing procedures for data collection
20
 Collecting data
 Analyzing and making predictions
Check Your Progress Exercise 2
1. Define descriptive research.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2. List eight steps involved in a descriptive investigation.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
3. State the purpose of survey studies.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2.2.3 Historical Research
This is a procedure supplementary to observation in which the researcher seeks to test the
authenticity of the reports or observations made by others.
Researchers who are interested in reporting events and/or conditions that occurred in the past
employ the historical method. An attempt is made to establish facts in order to arrive at
conclusions concerning past events or predict future events.
Steps to Follow
.Isolate the problem
.Collect source materials, including primary and secondary sources
.Evaluate source material
.Formulate hypotheses
.Report and interpret findings
Primary Sources of Information - Direct outcomes of events or the records of eyewitnesses
Secondary Sources of Information - Information provided by a person who did not directly
observe the event, object, or condition
21
External Criticism - As if the evidence under consideration is authentic. The researcher
checks the genuineness or validity of the source. Is it what it appears or claims to be? Is it
admissible as evidence?
Internal Criticism - After the source is authenticated, it asks if the source is accurate, was the
writer or creator competent, honest, and unbiased? How long after the event happened until it
was reported? Does the witness agree with other witnesses?
Establishing the Genuineness of a Document of Relic
Does the language and writing style conform to the period in question and is it typical of
other work done by the author?
Is there evidence that the author exhibits ignorance of things or events that man of his training
and time should have known?
Did he report about things, events, or places that could not have been known during that
period?
Has copying altered the original manuscript either intentionally or unintentionally?
Is the document an original draft or a copy? If it is a copy, was it reproduced in the exact
words of the original?
If manuscript is undated or the author unknown, are there any clues internally as to its origin?
Checking the Content of a Source of Information
What did the author mean by each word and statement?
How much credibility can the author’s statements be given?
Advantages
The research is not physically involved in the situation under study.
No danger of experimenter-subject interaction.
The researcher locates documents, data is gathered, and conclusions are drawn outsight.
Check your Progress Exercise 3
1. Define the historical method of research.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2. List the five steps to be followed in conducting a historical investigation.
investigation.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
3. Differentiate between the following primary and secondary sources by placing an "X"
in the blank before each primary source.
22
______ Original documents ______Periodicals
______Newspapers ______Relics
______Textbooks ______Reviews of research
______Artifacts ______Remains
______Encyclopedias
2.2.4 Empirical Research
Empirical research is at the core of many fields of study. Empirical research often defines
relationships, demonstrates cause and effect and sparks our own minds to begin thinking of
other possibilities to be tested and studied. Because empirical research is so important to so
many fields, it is essential that researchers have the skills necessary to locate empirical
research articles on their topics of study.
Check your progress exercise 4
1. Define empirical research.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2.2.5 Qualitative Vs. Quantitative
The simplest way to define qualitative research is to say that it is a type of research that
involves interpreting nonnumeric data. The underlying belief of qualitative research is that
"meaning is situated in a particular perspective or context, and, since different people and
groups have different perspectives and contexts, there are many different meanings in the
world, none of which is necessarily more valid or true than another" (Gay & Airasian, 1996).
A qualitative researcher, therefore, believes that the world cannot be pinned down objective
meanings, but that all variables must be taken into account when conducting research,
including the past experiences and personalities of the researcher. According to Peck and
Secker (1999), this idea has three important implications from a research perspective:
First, it follows that the purpose of research is not to establish objective facts about the social
world because objective knowledge is impossible; rather, the aim is to explore how research
participants understand, or make sense of, the topics in which we are interested. Second, the
theories we arrive at as researchers are also inevitably our own interpretations of research
23
participants' understandings and not simply a reflection of them. Finally, for readers and
reviewers to assess qualitative research, it is necessary to present a "thick description" of the
data, illustrate this with substantial extracts, and make the processes of the analysis
transparent.
How Qualitative Research differs from Quantitative Research:
In order to gain a clearer understanding of qualitative research, it is necessary to compare its
basic purpose and focus to those of quantitative research. The differences mainly result from
the positivist (the belief that the world can be measured, understood, and generalized about)
perspective of quantitative research versus the non-positivist (the belief that the world cannot
be generalized about) perspective of qualitative research. The following chart found on page
ten of Educational Research: Competencies for Analysis and Application by Gay and Airasian
(1996) makes this point very clear:
Characteristic Quantitative Research Qualitative Research
Approach Deductive Inductive
Theory testing, prediction, Describing multiple realities,

establishing facts, hypothesis developing deep
Purpose
testing understanding, capturing
everyday life
Isolates variables, uses large Examines full context,

samples, is often anonymous interacts with participants,
Research Focus
to participants, uses tests and collects data face-to-face
formal instruments from participants
Developed before study is Begins with an initial idea

initiated, structured, formal that evolves as researcher
Research Plan proposal learns more about
participants and setting,
flexible, tentative proposal
Mainly statistical, Mainly interpretive,

Data Analysis
quantitative descriptive
24
1. Define qualitative research.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2. Define quantitative research.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
3. Which one is more preferable? And why?
-----------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------
2.2.6 Clinical Research
Clinical research is about people - patients, former patients, and people who are well. The
goal is to find out what causes human disease, and how it can be prevented and treated.
1. What is clinical research?
-----------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------
2.2.7 Action Research
Action Research is a three-step spiral process of (1) planning which involves reconnaissance;
(2) taking actions; and (3) fact-finding about the results of the action.
Kurt Lewin (1947)
Action Research is the process by which practitioners attempt to study their problems
scientifically in order to guide, correct, and evaluate their decisions and actions.
Stephen Corey (1953)
Action Research in education is study conducted by colleagues in a school setting of the
results of their activities to improve instruction.
Carl Glickman (1992)
25
Action Research is a fancy way of saying let’s study what’s happening at our school and
decide how to make it a better place.
1. Define action research.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2.2.8 Analytical Research
To analyze means to break a topic or concept down into its parts in order to inspect and
understand it, and to restructure those parts in a way that makes sense to you. In an analytical
research paper, you do research to become an expert on a topic so that you can restructure and
present the parts of the topic from your own perspective. For example, you could analyze the
role of the mother in Ethiopian family. You could break down that topic into its parts--the
mother's duties in the family, social status, and expected role in the larger society--and
research those parts in order to present your general perspective and conclusion about the
mother's role.

1. Define analytical research.
2.3 SUMMARY
This unit has familiarized you to the most common types of research available in different
fields of study. In this vein, we have looked at experimental, historical, descriptive, etc
versions of research. Each type is suited to a specific kind of problem. It is also worth nothing
that each has its own steps or designs.
1. It is a type of research in which one variable is manipulated to see its effect on the
other.
26
2. Identify problem; formulate hypothesis; construct experimental design; conduct
experiment; compile raw data; apply test of significance
1. It studies or describes people or situations, as they currently exist.
2. Stating problems; identification of information; selection of tools for collecting
information; sampling; design procedures for data collection; collect data; analyze data;
make predictions
3. The are used to assess the features of whole population or situationsS
Historical research attempts to study past events.
1. Identify problem; collect source materials; evaluate them; formulate hypothesis; report
and interpret finding.
2. Try yourself
1. It is a type of research dependent on observation and experiment, provable by them and
guided by experience rather than theory
1. It is an attempt to describe the qualitative aspect of human behavior such as feelings,
emotions, etc
2. It is an attempt to quantify human behavior and offer interpretations
3. Each one is suited to a specific type of topic or problem. Therefore one should use
them accordingly or in combination
Check Your Progress Exercises 6
1. It studies about the causes and remedies of human diseases.
1. Action research is a study conducted on specific problem(s) by practitioners in
different fields of study in an attempt to improve their practices.
1. It is an attempt to break down a topic and study its parts carefully to fully understand
it
27
UNIT THREE
THE RESEARCH PROCESS
AIMS AND OBJECTIVES
 list the steps involved in carrying out research
 describe the steps briefly
3.1 INTRODUCTION
As research is a continuous process, a number of steps are involved from choosing a subject
or selecting a researchable problem to reporting the findings of the study. This unit
familiarizes you with the major steps involved in carrying out studies or research.
3.2 THE RESEARCH PROCESS
3.2.1 Selection of the problem
The first and the foremost thing to be decided by the research worker is the selection of the
research problem itself. It is probably the most important matter to be decided by the research
28
workers. Selection of an unsuitable problem would result in the inevitable failure. Following
factors may be kept in mind while selecting the problem.
(a) The problem should be such, in which researcher may be deeply interested. His own
interest will not only inspire him to work hard but also increase his power to grasp
things quickly and keep his morale high in items of difficulties.
(b) The problem should be allied with the chain of thinking or research already in
existence. Stray problems if selected become difficult to co-ordinate and do not add to
the wholesome development of the theory.
(c) The problem selected should not necessarily be new. It may be an old problem or one
on which work has already been done. There is a general temptation for the new
research worker to hit upon some new problems, as it satisfies his sense of vanity, but
from purely scientific point of view verification of an old problem may be equally
useful. Any hypothesis successfully proved in the repeated tests that hypothesis
assumes the form of a perfect theory.
(d) The topic should be within manageable limits. It should not be too comprehensive.
Here again a good deal of caution has to be taken. There is general tendency to select
vast topics, as they appear more grand. Lundberg has remarked at one place. “That
social scientists tend to be impatient of pains-taking study of limited problems and
incline rather towards the grandiose method of philosophy of history.”
(e) Vast subject of study, howsoever, grand they may appear, are unfit for real research
work. It requires a real sense of humility and a definite control over temptation on the
part of the research worker to confine himself to narrow topics. But they are very
essential for successful scientific study. Al Lundberg remarks – “If some problems of
physics today have fewer variables than some problems of sociology it is because
physicists have found it advisable to exercise modesty in the scope of the problem
which they set for themselves. They have been content to work patiently on a number
of small but related problems, but their efforts in aggregate have produced the
magnificent synthesis of the physical science as a whole.”
(f) The topic should preferably possess some direct utility. This point raises the
controversy of basic versus applied research. Basic research although so important for
29
social theory has little practical value. Applied research may process immediate utility
to some interested persons. Both of them have to be taken up but in order to make the
research work financially self-sufficient due stress has to be given to applied part.
(g) The topic should be practically feasible. Practical feasibility conveys many implied
conditions. They relate to technique, time, money and other resources at the disposal
of the research worker. The topic selected should not be such as cannot be dealt with
through existing techniques. Time, money and other resources at our disposal not only
limit the dimension of our studies but also the nature of studies. All these factors must
be thought out well in advance and a budget showing the proposed expenditure should
be prepared before the research work is actually started.
3.2.2 Preliminary Preparations
After the broad topic of study has been selected certain preliminary preparations are essential.
The researcher must acquaint himself with all the material that is available on the matter. This
may consist of written texts, research work, and unpublished information. The research
scholar should go through them. Allied matter should also be read. He should find out the
persons who have conducted research should contact them, discuss with them, the problem,
their findings, technique followed by them and the difficulties faced. This early preparation
will make him more equipped and better suited to tackle the problem under study.
3.2.3 Selection of sample
Social phenomena are very vast. It is practical impossible to contact each and every person.
Moreover, the problem under study may not concern everybody. Therefore, a sample is
generally drawn from amongst the people about whom the study has to be made. There are
different methods of drawing out a sample and they have been explained in detail in on of the
chapters. For this part only a brief mention is enough. The most essential feature of a good
sample is its representativeness. The sample should include as fully as possible the qualities
and attributes of the universe. It must be capable of yielding as accurate and reliable picture of
the total. The sampling must be simple and easy to be drawn. It should be possible to assess
the precision of the sampling and determining mathematically the limit to the errors. Above
all, it must be practicable. An unpredictable sample, however excellent, is of not real use. In
order to draw a representative sample the ‘universe’ or the total population that has to be
studied, must be known. Thus, for example, if it is proposed to conduct a survey of the
30
employment among middle class women and its effect upon the family organization in the
city of Addis, we must know the total number of middle class women employed in different
jobs in Addis. A list of all cases, of universe, known as source list is thus prepared and the
sample is drawn according to systematic sampling methods.
3.2.4 Deciding the method of study
It has been explained in the previous part of this book that different kinds of methods are used
in research. Each method has its own advantages and limitation. The research worker should,
therefore, decide in the very beginning as to what particular method he is going to use. In
selecting the method he should take into consideration not only the suitability of the method,
but his own knowledge of it also, more than one method may be applied but a prior planning
of the same is essential.
3.2.5 Nature of information to be collected
A social phenomenon is generally very complex and influenced by a number of variables. It
is, therefore, essential to decide before hand what information is to be collected.
Questionnaire and schedules are generally prepared for this purpose. They contain certain set
of questions that are to be asked from the informants. Questionnaire and are schedule bring
uniformity and thus introduce objectivity in the study. A proper drafting of these documents is
very essential. Any slip in the language may distort the meaning and many create distortion in
the replies. Information required should be complete. If some important information has been
left out it will require the same amount of labor and effort again to do it. Therefore, enough
care should be exercised in this respect.
3.2.6 Organization of survey
If it is quantitative survey a good deal of organizational work is required for successfully
conducting it. A number of field workers may be required. If they are not trained, they must
be given necessary training. Proper arrangement should be made for checking and supervising
their work so that they may not give false reports. After the report has been received, it has to
be edited. A thorough checking of the form has to be undertaken. After this the data has to be
classified, coded and presented in the form of tables. It is subjected to various kinds of
analysis and ultimately conclusions are drawn proving or disproving the hypothesis. All this
organizational work should be pre-planned. The researcher may put up his own organization
or may take the help of some already existing professional body. If the project is quite big and
31
likely to take enough time, it is generally, more useful to set up one’s own research
organization. In short surveys, help may be taken from professional organizations because
selection and training of the field workers is a time-consuming process and may require a lot
of preparatory expenditure. Even if a separate arrangement is made for this purpose,
occasional guidance may be taken and consultations made whenever some special technique
is involved.
3.2.7 Report
After the data have been analyzed deductions are drawn. These generalizations are often put
in form of a report. If the surveys have been undertaken at the instance of a third party care
should be taken to see that all the information demanded by him has been supplied. At such
times the surveyor may give his own opinion as well as the suggestions regarding the problem
under study. The form of report to be submitted should be decided before hand as it is very
helpful in deciding the nature of information to be collected and the form in which it is to be
analyzed.
Check your progress
1. The steps involved in research across different fields of study are different. Argue for
or against this.
2. --------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
---------------
3. What techniques are involved in defining research problems?
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
---------------
4. Summarize the major steps in research process.
……………………………………………………………………………………………...
3.3 SUMMARY
32
The steps are logical and sequential, that is, one step leads to the other, and provides good
stepping ground for the next one. If you follow these steps, you should have little difficulty in
writing you research paper, and you can see how well the steps fit together.
1. Even if there are slight variations across disciplines, the basic steps in the research
process start with problem identification and culminate with reporting findings.
2. Even though these are not directly discussed, the following might be considered
brainstorming, asking questions, scanning books, researching one’s friends or families,
etc.
3. Choosing a subject, narrowing/defining the subject; reviewing relevant literature;
gathering data; analyzing data, summarizing and reporting finds.
33
UNIT FOUR
RESEARCH DESIGN
Aims and Objectives
At the end of this unit you should be able to
 define the term ‘research design’
 list the different research designs
 describe the advantages and disadvantages of the designs
4.1 INTRODUCTION
After the selection and formulation of the research problem we have to decide about the type
of research design to be followed. The design of a sociological research project can be said to
be the plan of action, the strategy and the structure of the overall procedure by which we
intend to gain more knowledge of a specific problem or a specific aspect of the society.
4.2 THE MEANING OF RESEARCH DESIGN
‘A research design is the arrangement of conditions for collection and analysis of data in a
manner that aims to combine relevance to the research purpose with economy in procedure.
This definition emphasizes fulfillment of research purpose within the limited economy.
However, this definition does not make the structure of research design very clear. In the
opinion of the author of this book the best definition of ‘Research design’ has been given by
F.N. Kerlinger which specifies the process and structure and strategy of investigation
conceived so as to obtain answers to research questions and to control “Variance”.
The terms, plan, structure and strategy need to be defined further to make the structure and
procedure of research design more clear. The ‘plan’ includes everything to be done by the
investigator in research procedure that is from writing the hypothesis, defining the hypothesis
operationally and collecting data to the final analysis of data. Thus, it means the overall
scheme or programme of research. The term structure is taken to mean more specifically the
outline scheme or the paradigm of the specific research project. By paradigm we mean a
structure of a guiding model that regulates specifying their relationship or juxtaposition to
each other. When we draw diagrams that outline the variables and their relation and
juxtaposition, we build structural schemes for accomplishing operational research purposes.
The term ‘strategy’ refers to the methods to be used to gather and analyze the data. After
34
fixing up the objectives of the research project, one has to specify the procedures and methods
of investigation. One has also to apprehend the problems that may emerge and decide the
steps to be taken to tackle the problem.
4.3 PURPOSE OF RESEARCH DESIGN
Any research design has two basic purposes, (a) to provide answers to research-questions as
validly, objectively, accurately, and economically as possible, and (b) to bring empirical
evidence to bear on the research problem by controlling variance. All the research activities
have the purpose of answering research questions. However, the manner of seeking answers
to the research questions differs from situation to situation according to the nature of the
problems. In any project the research problems are further stated in specific hypothesis so that
they can be measurable. There are various ways of testing the hypothesis depending upon
their structure. The different research designs differ among them in the manner of seeking
answers to research questions and also in the manner of testing the hypothesis. (b) The second
basic purpose of research design is the control of variance. Research designs help in
manipulation of experimental variables, i.e., in seeking and establishing the relationship
between the assumed cause or the experimental variable and the effect, i.e., the dependent
variable. In establishing such a relationship it is essential that the effect of other variables,
besides the experimental one, should be controlled. Such variables are known as extraneous
variables. By stating that the purpose of any research design is to control variance we mean
that it specifically aims to:
i. Maximize systematic variance
ii. Control or eliminate extraneous variance, and
iii. Minimize error variance
The extent to which these ends can be and are fulfilled depends upon the type, structure, and
adequacy of the research design. Each type of variance needs further elaboration.
I. Maximizing systematic variance
The systematic variance is the variation or the difference caused in the dependent variable by
the experimental variable alone and not by any other variable. By maximizing the systematic
variance means to magnify or make evident the effect of the experimental variable on the
dependent variable. The concept of variance and its control can be explained further one the
basis of an example. Suppose a professor of English, or an university administrator wants to
35
improve the students, performance in the class. For this he has to find a better method of
imparting education. He makes the hypothesis that students’ reading habit is increased or their
class performance is improved by the ‘Participation Method.’, i.e., by participating in the
class discussions or seminars than by simple lecture method. In this example the experimental
variable will be ‘Participation Method of Teaching’, the effect of which the professor or the
administrator desires to measure, as compared to the effects of Lecture Method of Teaching.
Students’ performance is the dependent variable. In another study students’ reading habit can
be taken as the dependent variable, the experimental variable being the same.
II. Control of extraneous variance
It means the control or elimination of the impact of other independent variables, which are
likely to influence the dependent variable but are considered as unwanted or extraneous in the
context of particular research work. It is essential to eliminate their impact so that variation in
the dependent variable takes place only because of the assumed cause. In case of the above
example – age, intelligence, previous school environment, family background etc., are the
additional factors which might influence the reading habit or the class performance of the
students.
The researcher’s job is to isolate, minimize, nullify or eliminate the effect of such extraneous
variables in order to demonstrate the effect of experimental variable ‘teaching methods,’ on
the dependent variable, ‘the student’s class performance or the students’ ‘reading habit’.
The best way of doing this is to make the elements (the individuals) of the two groups,
‘experimental’ and ‘controlled’ to be equal or matched on the extraneous variables. This
means that persons of same age, of similar family background and, or of same school should
be equally distributed in both the groups. Randomization and matching are the two processes
of making the elements of the groups equal or of making the two groups homogenous. These
two processes would be discussed in detail under ‘experimental research design.’
III. Minimizing Error Variance
A third group of factors, besides those mentioned above, influencing the dependent variable
are known as random variables because they are temporary factors which vary from situation
to situation. Their influence is considered as error variance. There are a number of
determinants of error variance, for instance, factors associated with individual differences
among subjects due to errors of measurement, variation of response from one trial to another,
36
guessing, momentary lose in attention, temporary fatigue, lapses of memory, and transient
emotional sate of subjects. Minimization of error variance can be achieved through:
(i) Measuring under controlled conditions, and
(ii) By increasing the reliability of measure or measuring instruments.
2. FUNCTIONS OF A RESEARCH DESIGN
Any research design tells the researcher as to what steps to follow. It tells us what things are
to be observed, how many observations should be made, i.e., what should be the size of the
sample and how should the sample units be selected. It helps to locate the variable, and tells
us how to manipulate the variables. It also tells us how to test the relationship among
variables and which statistical methods are suitable to test the relationship. Finally, it tells us
how to analyze the qualitative or the quantitative representations of the observations. It
outlines the possible conclusions to be drawn from the analysis.
4.4 REQUIREMENTS FOR A GOOD RESEARCH DESIGN
The characteristics given below are to be followed for the development of a good research
design. There is not any particular example of formulated research design which perfectly fits
the model of a good research design. However, the researchers try their best to develop one
that approximates to what can be called an adequate or a good research design.
These are
(i) Nature and scope of the problem to be studied must be stated clearly, or must be
well defined and formulated.
(ii) If any hypothesis is to be tested it must be clearly formulated.
(iii) The research design must adequately answer the research questions and test the
hypothesis.
(iv) Relevant variables must be clearly identified and operationalized. Adequate
method of collecting the information and methods of logically deriving the
conclusion must be developed. Only then control of variance is possible.
(v) The research design must be structured in a manner that it fulfills the need of
internal and external validity.
The extent to which these are fulfilled determines the scientific validity of our knowledge.
The different types of research designs differ among themselves in the extent to which the
above criteria can be fulfilled. These will be examined while discussing different types of
37
research designs. Before discussing the different types of research designs we need to define
the terminologies ‘internal validity’ and ‘external validity’.
By ‘Internal Validity’ we mean whether the experimental manipulation really makes a
significant difference. It is the extent of control that determines internal validity by
differentiating experimental variance from extraneous variance. External Validity means
generalizability, or representativeness. By generalizability we mean whether the relationship
established between the variables is applicable to the sample that has been selected. The
question that arises is in relation to the ecological and variable representativeness of studies.
This means that if the social setting in which the study or the experiment was conducted is
changed whether the relationship between the said variables will still hold. For example,
whether the relationship found between methods of teaching and class performance will be
true for the students on whom the study was conducted or would be true for the students of
different levels (post-graduate, graduate, high school and middle school) of different faculties,
of different counties and times? Similarly, does the relationship found between political
awareness and participation hold true for college students from different levels, faculties and
even for students from different types of universities, namely residential and non-residential?
‘External Validity’ depends also upon equivalence of definitions from situation to situation.
suppose we are studying the variables ‘achievement’, ‘aggression’ or ‘anxiety’.
Do the variables mean the same for people from different areas or from different cultures? For
this we have to define the terminologies. For example, when we are studying ‘anxiety’ we
have to define what kind of anxiety we mean We have to see whether all kinds of anxiety are
the same or not. For example, anxiety may be created in different ways, in one situation it
may be created by verbal statements and in another situation by electric shock. The question
is whether the two methods induce anxiety of the same type.
4.5 TYPES OF RESEARCH DESIGN
Once the problem to be investigated has been formulated perfectly, the investigator has to
design the structure of his research. Any research work has the common purpose of finding
answers to meaningful questions. Besides this, each research has its own specific purpose of
studying, analyzing and explaining the events in a different manner. It is one the basis of these
38
specific purposes that we can divide the research designs into the following four different
types. The research designs vary from each other in their specific purposes.
(I) Exploratory or formularize research design
This research design is applied when the researcher is not acquainted with the problem or the
community he wants to study. Therefore, it aims to gain familiarity with a new phenomena or
community. It aims to achieve new insights into the problem or community in order either to
formulate a more precise research problem or to formulate specific hypothesis and to locate
the possible variables.
(II) Descriptive research design
It aims to gain accurate description of the community, institution or events in a manner that
involves minimum bias and is maximum reliable.
(III) Experimental and Ex-Post facto research design
Both of these research designs aim to test causal hypothesis but in a different manner.
Besides the above-mentioned specific purpose the research designs differ from each other in
their form or structure, in the proposition of hypothesis, and in the extent to which control of
variance is needed and is possible. The four research designs would be explained in the
following pages in terms of their specific purpose, proposition of hypothesis, structure and
control of variance. The selections of the correct research design or combination of designs
determine the scientific validity of a research.
4.5.1 Exploratory or formulative research design
It has been mentioned before that this research design aims to gain familiarity with the
problem, or situation, or the community not known before. It helps in the discovery of ideas
and insights that helps in understanding and formulating a problem for the development of
hypothesis and for its more precise investigation. Without sufficient information about the
phenomenon or the setting in which the study is to be carried out, it is not possible to develop
a highly structured study design that would test specific hypothesis. Exploratory studies also
help in clarification of concepts, in locating important variables, in establishing priorities for
further research, in gathering information about practical possibilities for carrying out
research in real life setting, and in providing a census of problems regarded as urgent by
people working in a given field of study. As for example, if we want to make a study on
minority group relations we will have to explore detailed information regarding the problem,
39
the concept and the setting before we develop any structured research design. This happens
when much empirical work has not been done in the specific filed. Study of minority relations
involves similar problems as faced by Jahoda, Seltiz and others in the study of relations
between member of different social and religious groups within the United States. For them it
was essential to select such settings where the members of the different groups are in face to
face condition, possibly in neighborhood housing colony, grocery, shops, drug store, etc.
Since these investigators had little knowledge in the above mentioned areas of interest, they
needed to first make an exploratory study design for six months. For gaining familiarity they
interviewed housing experts, managers of housing projects, and race relation official of public
housing agencies. Experience and insights of these experts made a valuable contribution
towards gaining familiarity with the problem and in developing more systematic study of the
same. On the basis of the acquainted knowledge they selected public housing areas or inter-
racial housing project areas as integrated or suggested areas of race relationship. Such
exploratory studies also helped them to identify other possible factors than occupancy studies
also helped them to identify other possible factors than occupancy pattern (viz. integrated
inter-racial pattern and segregated bi-racial pattern), such as the neighborhood where the
projects were located, the racial composition of the tenants, the attitude of the management
staff, project facilities, differences in religion, education or political views, etc. Only on the
basis of all these information the investigators could develop a systematic study to determine
how the difference in occupancy pattern might produce the observed differences in attitudes
towards Negroes.
In exploratory research design one makes use of the following methods for collecting
information.
(a) Review of the related and other pertinent literature;
(b) Survey or interview of experienced people; and
(c) Analysis of insight stimulating experiences.
One may use one or more of these methods. Whatever method is chosen, it is used in a
flexible way. As one gains greater knowledge and understanding of the problem one needs to
make frequent changes in the research procedure.
40
(a) Survey of literature
One of the simplest, short-cut and economic way to gain knowledge of or insight into the
unfamiliar phenomena is to build ones research work based upon the works already done. One
can collect explicitly formulated hypothesis from similar studies and use them as a basis for
development of further research guidelines or for the development of hypothesis. Theories
and concepts not only from similar areas of study but also from other studies might be useful
in defining the problem in hand and in developing hypothesis.
(b) The experience survey
The aim of the experience survey is to obtain insight into the problem or into the relationships
between variables and to get an accurate picture of current practices by interviewing the
experienced people. Mere statistical information is not sufficient for developing insight or for
exploration into the problem. For example, the nature and amount of changes in tribal society
can be assessed better by interviewing tribal welfare officers or tribals themselves than by
going through the statistical figures along which show the change in terms of the amount of
money invested or in terms of the number of tribal persons getting education, housing, or
other facilities. The author also had to take help from experienced persons, the bureaucrats
themselves, in developing a project to study the role of bureaucrats in the political
development.
Though in our day-to-day experiences we are in a position to observe the effects of alternative
decisions and actions with respect to problems of human relations, yet very few of us ever
remember it, or put it into written form, or might have the competence to communicate it.
Therefore, the researcher has to carefully select the respondents for experience survey in a
manner that ensures a representation of different types of experience. As for example, in a
study of the relationship between work performance and job satisfaction one will need to
know above people’s job satisfaction, their work performance and also the working conditions
or facilities available to the people.
(c) Insight stimulating experiences
Specially when the area under investigation is unformulated and there are few or no
experienced persons to provide necessary information for development of insights and
hypothesis for specific research, then intensive study of a few selected examples have been
found to be highly fruitful. As for example the remarkable theoretical insights of Sigmund
41
Freud were stimulated by his intensive study of some patients. Similarly, deep intensive
anthropological studies of few primitive cultures have thrown light to the understanding of not
only similar other tribes but also in the understanding of modern man. Utility of such
examples in the formulation of the problem and hypothesis depends upon the intensity of the
study of the individuals, as in case of Freud, or up on the intensity of the study of the group,
community or culture as in case of the anthropology studies it depends upon the attitude of the
investigator. His attitude should be of seeking greater and greater amount of knowledge rather
than of testing specific hypotheses. He must have integrative power, i.e., can combine diverse
pieces of information in a manner that it leads to unified interpretation. Its utility can be
realized from the fact that often a study of a few instances may produce a wealth of new
insights new ideas, whereas, a huge collection of facts may not.
4.5.2 Descriptive research design
(a) Meaning and purpose of descriptive design
As the name itself implies, the purpose of this research design is to provide description of an
individual, a community, a society, an event, or of any other unit under investigation. The
description, provided by it, is aimed to involve minimum bias and maximum reliability. As
different from the exploratory study design, the descriptive study design tests specific but
non-causal hypothesis.
(b) Types of descriptive research design
Depending upon its structure description may be of two types:
(i) Qualitative, and
(ii) Quantitative
(i) Qualitative description: Studies related to culture, patterns or processes of social and
cultural change and of their elements, like, customs, norms or values, of social structure and
organization or of patterns of human behavior lead to qualitative descriptions of general
nature. Contributions of early sociologists or anthropologists were related to these areas and
were most general and wholistic in nature. They dealt with societies as a whole. Therefore,
these descriptions were too general in nature. The qualitative description is obtained with the
help of historical method or comparative method which looks into the origin and the
processes of development of any culture or society and their parts. Though descriptive studies
are not limited to any one method yet observation method, and especially the participant
42
observation method has been more commonly used for collection of information. Participant
observation is more useful than the non-participant one in providing such detailed description.
Besides the use of participant observation method the researchers have made use of other
methods like, formal or informal interviews, reading diaries or other secondary sources of
data and so on. Information drawn from different sources have been combined together for
drawing inferences.
(ii) Quantitative description: It is achieved through the use of questionnaire method,
structured interview, or through any other structured method. Such descriptions have covered
different areas of human life and can be organized under different categories. These are:
For example, take (i) Description of characteristics of individuals, groups, or communities:
Under such descriptions come the age distribution of individuals, their racial background,
caste, income, occupation, nationality, or religion. These may also include description of
groups or communities in terms of their socio-economic condition, cultural pattern, attitude,
etc.
(ii) Describing facilities: Such descriptions may be about the conditions of living in terms of
average number of persons per room, per dwelling unit, per school, or per hospital bed in a
year. Again, total calories of food intake of persons from different status, number and types of
articles per family from different standards of living etc. also come under this type.
(iii) Description of habits: Many studies have been carried on in sociology to describe
different types of habits of the people, namely, their cinema going habit, their reading habit,
etc. This are studied in terms of the number of times, like, regularly occasionally, or rarely,
one visit or uses them.
(iv) Studies describing attitudes of the people: These are becoming more common since the
emergence of modern democracy increased emphasis on people’s greater participation in
activities of the government or in the development of the society. A large number of
attitudinal studies have been carried on by sociologists and psychologists to assess peoples’
opinion and attitude towards political parties, government, socio-economic systems, and
cultural values and so on. One may estimate on the basis of such attitudinal studies as to what
number or percentage of people from a community or from a given population will be holding
certain views or attitude. As for example one may state that greater number of protestants than
Muslims prefer family planning. Highly specific descriptions help in making predictions. In
43
case of the above example on would be able to predict as to what percent of people from
Addis will vote for EPRDF as against the percentage of persons who will vote for EDP.
(v) Descriptive studies do not only provide picture about variation or distribution of a
characteristics in the population, they also help in discovering how different variables are
associated. For example, the study of students’ political awareness and political behavior may
reveal that their political awareness and political behavior varies with their levels of
education, their family background, and parents’ level of education. Such studies help in the
development of theories.
(c) Importance of descriptive research design
In behavioral sciences as a whole, descriptive approaches have been most widely used. These
days there is greater demand of making our studies as accurate as possible like those in
natural sciences. Studies in natural sciences are highly structured as they are carried on under
controlled conditions through experimentation. This helps in making accurate predictions and
control. Therefore, there is greater emphasis on the use of experimental method behavioral
sciences.
Even then descriptive studies would continue to occupy important place as a part of survey
research work; especially if and when developmental programme would be based upon
assessment of peoples’ opinion and attitude.
Besides, descriptive approaches have great value because of their use in both the preliminary
and final stages of an experimental study. Thus, the descriptive approaches may serve as the
‘reconnaissance’ phase of an investigation in a new area in which the purpose is to identify
factors which are most promising for experimental investigation. The descriptive approaches
may be used to portray the status of a situation after some procedure suggested by an
experimental analysis has been put into effect. A second reason for the widespread use of
descriptive approaches is that the kind of data yielded from such a study is highly useful to
applied workers. It provides descriptive but concrete and factual information about the
characteristics of any phenomenon or event and the situation under which it takes place. This
also suggests the possible factors behind the occurrence of events which help the applied
workers to deal with every day practical problems and to make definite plans for further
action. Finally, the descriptive studies are also very useful, because they apply to a very broad
class of problems.
44
(d) Steps of descriptive research design
Of scientific investigation, which are formulating the objective of the study, designing the
methods of data collection, selecting the sample, collecting information, processing and
analyzing the data, and reporting of the findings. These steps or procedures are followed in
order to obtain complete and accurate information. This provides greater protection against
bias than is possible in exploratory studies.
The first requisite of a good descriptive study is to clearly and specifically state the purpose,
goals and objectives of the investigation. If the purpose is stated in a vague manner, then it
may not clearly direct the appropriate sources of information. The data to be collected must be
appropriate or relevant to the questions raised. The second logical step of a descriptive study
is spelling out in detail exactly the processes, products, conditions, or characteristics of the
things to be examined. Naturally at this stage one has to clearly define the variables in
operational an term, that means defining them in terms of the ways in which they are to be
measured or observed. At the third stage, one decides precisely as to how the descriptive
information will be gathered and analyzed and through which methods. The methods of data
collection has not only to be selected from those available methods like, questionnaires,
observation, interview, projective techniques etc. but has also to be constructed and evaluated
in terms of their objectivity, reliability and validity. One has also to decide about the
structures and procedures of sampling and the size and reliability of the sample. One has to
determine exactly how the data will be categorized, what descriptive indexes will be used to
summarize the information. Furthermore, one has to identify which, if any, statistical test of
hypothesis will be applied, and lastly, plan the charts, graphs and tables that might be
appropriate for the ultimate presentation of the results.
After the above decisions have been made, the next step is to carefully and accurately
implement them.
At the final stage of descriptive study comes analyzing the data and making inferences. The
process of analysis includes coding the interview, replies and observations, tabulating the
data, and performing statistical computations, like, average, percentages, correlations, etc.
Steps are taken to observe economy and safeguard errors at each step of research.
Thus, we have seen that the techniques of data collection in descriptive research design are
more specific. Also that it has more specific hypothesis to be tested. One can make more
45
accurate predictions on the basis of descriptive studies which is not possible in exploratory
research designs.
(e) Similarities and differences between descriptive and exploratory research designs
(i) Specific structure of hypothesis, of methods of data collection and analyses make
descriptive studies much more structured than the exploratory research designs.
(ii) Exploratory research designs develop hypotheses at the end of the study whereas
descriptive research designs begin with hypotheses. This is because exploratory
research designs are applied when one does not have knowledge or clear ideas
about a community. Descriptive research designs are applied when one has ideas
or knowledge about the community which he tries to present in a systematic or
organized manner. Therefore, with the existing amount of knowledge,
development of hypotheses becomes possible.
(iii) In descriptive studies hypotheses are much more specific in comparison to the
exploratory studies.
 Often exploratory and descriptive study designs are used in sequence. Exploratory
study of a problem is done at the initial stage of research and hypotheses are
developed at the later stage. These hypotheses are tested through descriptive study
designs.
 Common advantages and disadvantages of the descriptive approaches
Advantages
(i) One may desire to isolate the effect of one or two variables upon a dependent
variable. To isolate such effect, it is necessary to test them under the controlled
conditions of a laboratory. However, it is not possible to bring many social events
or units from a ‘real life’ situation to the laboratory for controlled study, for
example, ‘locating cause behind social movements.’ Nor is it possible to lift the
findings from the laboratory and apply them directly in the field. In such cases
effects of different variables may be marked out through statistical computation of
data achieved through descriptive studies.
(ii) The different research designs are suited to the study of problems of different
nature. Therefore, there are specific types of problems which can be studied only
through the descriptive method.
46
(iii) Many of our developmental plans or welfare policies are related to a big
population. Before such plans or policies are made and implemented we may need
information about the characteristics, nature, or problems of the people. We may
also need to know the attitude of the people towards specific government plans or
policies. Such attitudinal studies would reflect the possible type of reaction of the
people towards the policies. This would help the government to find out suitable
methods of implementing the policies, like ‘family planning.’ These can be easily
achieved through descriptive studies.
Disadvantages
Defects in the results of descriptive studies occur due to the structure of the research design
itself, i.e., ‘faults of the method per se’, and secondly, because of avoidable faults involved at
the stage of its construction and application.
(i) The first fault involved is in its structure, that is, it appears to be simple on the
surface. People, therefore, tend to apply it without care. In such cases the
descriptive studies remain mainly tools of gathering information and not of doing
research.
(ii) People tend to collect only those evidences that support the ideas of the
investigators or their hypothesis. This leads to over generalization.
(iii) Other disadvantages of the descriptive studies are that they do not provide very
much information about the effects of the variables because in them one cannot
manipulate the variables to go their effects isolated. Hence, no real evidence of
cause and effect relationship is provided. Thus, the descriptions of natural situation
supply us with information about the events in terms of their concomitant
variation. But they do not show causal sequences of events. They provide us with
information or description of the conditions under which an event takes place but
do not provide evidences for inferring causation.
(iv) Faults may also occur because of respondents or investigators bias. Descriptive
studies require the co-operation of respondents over whom the investigator has no
direct control. A great deal of time and effort has to be spent after creating their
interest and co-operation. Even when they co-operate, there are chances of miss-
information and forgetting.
47
4.5.3 Experimental research design
By experiment we refer to that portion of research in which variables are manipulated and
their effects upon other variables are observed.
Thus, experiment is the procedure for gaining knowledge by collecting new or fresh
observation under controlled conditions. When a scientist fails to find out through simple
observation the possible factors operating in a given problem, he resorts to experiment. Thus,
the difference between simple observation and experiment is that in the second one,
observation is done under controlled situation. ‘Experiment may be considered to have begun
when there is actual human interference with the conditions that determine the phenomenon
under observation. The above statements show that the fundamental rule of the experimental
method is to vary only one condition at a time and see its effect on the dependent variable.
Here one creates simple cause and effect situations under which one or more variables, the
causes, are manipulated and their effects upon the dependent variables are observed, while the
other conditions are rigidly kept to be constant.
Experimental research design is different from the previous two research designs, i.e., the
‘Exploratory Research Design’ and the ‘Descriptive Research Design’ in its objective,
structure, and procedures. The purpose of experimental research design is to test a causal
hypothesis which is highly structured. It has been mentioned before that a causal hypothesis is
one that states the cause and effect relationship between two or more variables. The structure
and procedure of experimental research will be discussed later on. To understand its structure
and procedure, it is essential to define first of all some basic concepts related to experiment,
the logic behind experiment, the concept of evidence and of validity.
(a) Concept of variables and its types
It has been said before that scientists set up relatively simple cause-and-effect situation where
one or more variables are manipulated and their effects upon others are observed while other
conditions are held constant.
It can be said that the term ‘variable’ can be defined in simple language as the property of
individuals, groups, community or any object that varies from one unit to another and can be
represented through definite values. According to F.N. Kerlinger, ‘A variable is a property
that takes on different values.’ The variables, most frequently used in sciences, like sociology,
education and psychology, are class, income, occupation, age, intelligence, etc. These
48
properties do not only vary from individual to individual or from group to group but do also
influence and thus create variation in the peoples’ aptitude, opinion, behavior and even their
inter-social relations and interaction.
In an experiment a researcher is to deal with two sets of variables:
Independent variables: i.e., the variables or factors which act as determinant of other factors
or events or say do influence them. These are called cause.
Dependent variables: These cover those characteristics or behavior of units which are
influenced by the first group of factors. This means that whenever there is a variation in the
independent variable or cause there would be a variation in these. Individual’s behavior,
attitude, interaction, etc. can be called dependent variables or the effect when seen as the facts
affected by the individual’s age, income, education or intelligence which are seen as cause or
the independent variables.
The experimenter may be interested to measure the degree of effect of one or two independent
variable upon the dependent variable. Which characteristics will be considered as independent
variable and which one as dependent variable depends upon the nature of the problem under
study and varies from one research work to another. For example in one study, population
growth and population density can be taken as determinant factor of the increasing slum
condition, which is the effect, whereas in another study slum may be taken as a condition that
creates and increases the rate of delinquency. In the first example ‘slum’ is the assumed effect
and in the second example it is the assumed cause.
A third group of variables which come into the picture while conducting an experiment can be
termed as extraneous variables. The extraneous variables are those independent variables
which are likely to influence the assumed effect but since the researcher is interested in
establishing the causal relation between the specific assumed cause and effect, therefore, the
effect of such extraneous variables have to be controlled, eliminated, or at least minimized.
Besides this, the effects of the random variables have to be eliminated.
In his attempt to establish causal relationship between an independent and dependent variable,
the experimenter tries to manipulate or say create variation in the independent variable, and
see its effect upon the dependent variable.
Any variable that can be manipulated or in which variation can be created is called active
variable. Imparting educating by different teaching methods or creating different
49
environments of play, creating anxiety etc. can be called active variable. Contrary to these are
the assigned variables which cannot be manipulated or in which we cannot create any change.
These are like sex, age, caste characteristics of the individuals.
(b) Procedures of experimental research design
The steps or processed involved in experimental research design can be divided into the
following:
(i) Construction of control and experimental groups.
(ii) Pre-test
(iii) Exposure of the experimental groups to the experimental variable.
(iv) Post-test
Step 1. Construction of control and experimental groups: This step consists of the selection of
the units of measurement, i.e., the individuals from a given population, called universe for the
purpose of a specific study. The process or method of selecting the individuals is same or at
least related to the process or method of assignment of individuals to the control and
experimental groups.
Whatever be the structure or type of experimental research design, it involves comparison of
two or more groups of subjects. One group, the experimental group is exposed to the effect of
experimental variable whereas the control group is not. If the two groups are found to be
similar on all other characteristics, except the exposure to the experimental variable, then the
difference in the dependent variable is assumed to have been caused by the experimental
variable. Sometimes more than one experimental group is constructed and exposed to
different experimental treatments in order to find out the effect of different amount of
exposure to the experimental variables or the effect of several but similar experimental
variables. One may make several control groups to isolate the effect of measurement, or the
instrumentation effect from the other effects. As an example of studying the effect of similar
experimental variable, we can take the example of studying the effect of different teaching
methods on the students’ performance. We may assume that the different teaching methods,
viz. (1) Lecture, (2) Lecture and demonstration, and (3) Lecture and class participation
through seminars – will determine the students’ class performance differently. For this we
will need three groups of students who will be taught by the different methods and then be
compared on the basis of their performance.
50
In order to test the effect of experimental variable, we need to control the effect of other
variables, which may create difference in the results. In case of the above example, besides
the method of teaching, class performance of the students may vary from one to the other
according to their varying economic position, family background, occupation of the parents,
pre-university environment, and so on. These variables can be called extraneous variables for
they are likely to determine the effect but we are not interested to measure the degree of their
influence on the dependent variable, because we want to measure the effect of methods of
teaching by isolating them from the above mentioned factors. The way to culminate or control
the effects of such extraneous variables is to make the control and experimental groups to be
equal or similar to each other on these characteristics. Control of social conditions cannot
always be obtained by manipulating the subjects or by exerting any physical force on them or
by creating an artificial condition as is done in the areas of natural or physical sciences. Here
control is achieved by selecting the individuals for different groups in a manner that they are
similar on extraneous variables like, income, occupation, chronological etc.
There are two different processes through which the individuals are selected from the universe
and assigned to the different groups in a manner that they become similar to each other. These
two methods of selection and assignment are called (a) processes of Randomization, and (b)
the process of matching.
(a) Randomization: Process of randomization includes the process of selecting the units of
people at random from the total population and assigning them to the control and
experimental groups at random. According to the principles of randomization any social
characteristics is distributed in the population systematically and not haphazardly. Secondly,
in random procedures every member of a population has an equal chance of being selected.
Therefore, all the characteristics of the population are expected to be equally or truly
represented in the sample if the selection is made on the basis of random principles. Thus, to
study the effect of teaching methods on the performance of the English students we would
select a given number of them from 3rd year and 4th year English classes and then assign them
into groups at random. Randomization is the fundamental process of achieving control in
experiment in social sciences. It is assumed that random allocation of subjects to treatments
51
should make the groups equal on the dependent variable. Whether it does so or not depends
upon chance, which with careful planning will be reliable in most cases.
(b) Matching: Matching is the second process or technique, used to make groups, control and
experimental, equal. The units of the groups or say the individuals are compared and matched
on selected characteristics. One may select the experimental groups in a way that each group
has the same number of members having the same experience, or the same intelligence level
and so on. This sort of matching can be called group matching or frequency matching. Here
the subjects are allocated in a manner so as to create groups with similar distributions on a
few relevant variables, but not necessarily on all of them simultaneously. Thus, two groups
might have the same age distribution and the same sex distribution but not the same age by
sex distribution. This method is less rigorous and can be attained with small number of
subjects. A second way is to match the individuals for these characteristics in a way that if a
member of group ‘A’ has a high level of intelligence comes from a given social class and has
a certain amount of past experience, and then each of the other groups should have one
member with the same characteristics. This sort of individual to individual matching is
difficult. The process here is more rigorous more time consuming and more costly than the
group matching. If we are studying students’ political socialization it would be easy to select
two students who have same age, same level of education and same marks. But if we add the
characteristics caste, I.Q., and school environment, to those mentioned above, then the task
would become difficult. The greater the number of factors on which the individuals are to be
matched, the more difficult is the task of finding identical persons. Since the problem of
matching makes the process a complex one, the researchers use randomization more
commonly.
Step 2. Pre-test: Pre-test helps to establish the original equality and thus comparability of the
different groups. By measuring certain given characteristics of the units of different groups
say for example age, income, I.Q., level of education etc. we can know whether and to what
extent the groups have been equal on extraneous variables.
Secondly, pre-test also helps to measure the initial position of the different groups on the
dependent variable. This helps us to safely conclude that the difference found between the
experimental and a control group is due to the influence of the experimental variable.
52
Step 3. Exposure of the experimental group to the experimental variable: After giving pre-test
the experimental group is exposed to the experimental variables or is given a treatment. For
example, an experimental group of students might be exposed to a new method of teaching or
be shown a picture or be made to use new cosmetic products. But at the same time the control
group is denied of the exposure to the same variable.
Step 4. Post-test: at the fourth stage, observation or measurements on a sociometric scale are
made of each group and are compared with the pre-test measurement available. Comparative
changes or differences in the two measurements, pre-test and after-test, indicate the possibility
of influence of the causal factors. Sometimes changes in the final measurement or differences
between the groups may be due to fluctuations of random sampling or because of extraneous
factors. However, if we are sure that such effects have already been successfully controlled,
then we can safely conclude that the difference between the control and experimental groups
or between the results of pre-test and the post-test are due to the impact of the experimental
variable. Such relationships cannot be considered as proved, unless confirmed and established
by repeated testing of the relationship exactly under similar conditions. In case of
experimenting on human beings, it becomes very difficult to maintain the similarity of the
experimental condition.
Check your progress Exercise
1. What is research design?
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2. What is the purpose research design?
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
3. List and define each types of research design.
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
4. What we thuds do you use for collecting data in exploratory research design?
53
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
5. Why do we prefer one research design over the other?
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
4.6 SUMMARY
Research design is the arrangement of conditions for collection and analysis of data. It is the
plan, structure, and strategy of investigation worked out to obtain answers to research
questions. We have identified a number of types of research designs suited to specific
research problems. In this regard, we made mention of experimental, exploratory, and
descriptive research designs.
1. It is the plan or design used to facilitate conditions for the collection and analysis of
data. This is done with the view to answering research questions.
2. a) To maximize systematic variance
b) Control external variables
3. Minimize error variance
a) Experimental research design – tests relation between different variables
b) Formulative research design – helps to identify research problems
c) Descriptive research design – focuses on offering specific descriptions without
focusing on causal relations
4. a) Review of related literature
b) Survey
c) Analysis of insight stimulating experiences
5. The choice of one design over the other depends among others, on the nature of the
research problem selected for study.
54
UNIT FIVE
SAMPLING TECHNIQUES
Aims and objectives

At the end of this unit you should
 define the term ‘sampling’
 define the term non-probability sampling
 define the term probability sampling
 list and describe the many types of sampling
 describe the advantages of sampling
5.1 INTRODUCTION
The purpose of this unit is to briefly discuss about the concept, advantages, and types of
sampling which is one of the basic components of the research design.
5.2 SAMPLING TECHNIQUES
5.2.1 Sampling Defined
In simple words sampling consists of obtaining information from a portion of a larger group
or universe. Often a researcher has to collect information about universe that consists of vast,
differentiated population spread over a larger territory; and that too within limited amount of
time and money. Measuring or collecting information from each and every member of such a
vast population is, therefore, always not possible. It is known that part of a whole can give
sufficient dependable information if the procedures followed in selection of the part had been
scientific. It is this part of a whole which is called a sample.
While dealing with sample and techniques of selection the first question that arises is ‘what
should be the desired characteristics of a sample? In an answer to this it can be said that: (i) a
proper sample must give a precise but correct picture of the population from which it is
drawn; (ii) the sample must be obtained by probability process. This would permit the use of
statistical procedures to relate the sample to the population from which it has been drawn; (iii)
the sample should be as small as precision considerations permit, and (iv) it should be as
economical as possible and gathered swiftly to be completed within the time schedule.
55
Before we elaborate the advantages aPage | 56nd
56nd disadvantages of sampling in comparison to
non-sampling studies, the different techniques of drawing a sample, and the problem of
reliability of sample, it is essential to define the concepts most commonly used in this
concern, that is: (i) universe or population, (ii) stratum, (iii) elements, and (iv) sample.
Universe
In simple language a population or universe can be defined as any collection of persons or
objects or events in which one is interested.
Universe or population differs for each research problem depending upon the nature and type
of information sought. In other words, population consists of the people who are related to the
specific problem under investigation. “A population is the aggregate of all the cases that
conform to some designated set of specification”. Thus for example, if we are studying the
relationship between the class achievements of the university students and the methods of
teaching, then the students of any place and of any time will come under our population. For
another example, if we are studying the voting behavior or political participation of the
citizens of Ethiopia then all the adult citizens of Ethiopia living in Ethiopia or outside will
come under ‘population’. In research we often speak in terms of population characteristics,
examples of which are age, sex, income, place of residence, caste, occupation, population,
size, density, migration rate and so on. At a time all of these characteristics are not measured.
Which characteristics are to be measured depends upon the nature and type of problem under
investigation. According to the number of characteristics in which we are interested at a time
or which we intend to measure at a time, we can divide the population into several types.
(a) Univariate population.
(b) Bivariate population.
(c) Multivariate population.
(a) Univariate Population
Univariate population is one in which one considers only one characteristics at a time. This
characteristic may be any one, for example, age, income, sex, T.V. listening habit, or cinema
going habit, and so on.
These doses not, however, mean that the population has one characteristic rather that we are
considering only one out of the several at a time.
56
(b) Bivariate Population
Our population can be defined as a bivariate type when we are measuring two characteristics
simultaneously of each member. As for example, we may want to know-how cinema-going
habit varies from urbanites to ruralites or how political participation is determined by degree
of political awareness, or by political orientation.
(c) Multivariate population
A multivariate population is one in which we consider observations on three or more
characteristics simultaneously. A car accident on the road is often caused not only by the
mechanical failure of the car but also by one or several of other factors, like, the driver’s
mental and physical condition, pedestrian’s behavior, traffic volume, slipperiness of road,
improper signals at cross section, disregard of traffic rules and regulations, careless driving,
mechanical failure of vehicle etc. Take another example ‘poverty’. Poverty is caused by many
factors, such as big and fastly growing population, lack of proper industrialization according
to the growing need of the population, discriminate distribution of wealth, religiosity, and so
on.
Stratum
When a population is divided into several groups on the basis of one or several characteristics
we call each group a stratum. Thus, stratum can also be called a sub-population. A stratum
may be defined by one or more specifications that divide a population into mutually exclusive
segments. For instance, a given population may be divided into different stratums on the basis
of the cinema going habit of the people, viz., (i) males who visit cinema frequently, (ii) males
who visit cinema occasionally, (iii) males who rarely visit cinema, and (iv) those males who
do not at all visit cinema. Males, females, students, non-students, persons of different age
groups, all can be divided into the above four stratums on the basis of their cinema going
habit. Thus the number of stratums would be increased depending upon the number of
characteristics included for stratification.
Population element
By a population element we mean the units that make the population. Such unit may be an
individual, an object, or even a small group. We often want to know-how certain
characteristics of the elements are distributed in a population.
57
Sample
By sample we mean the aggregate of objects, persons, or elements selected from the universe.
It is a portion or a sub-part of the total population. To collect information about a population
one can follow any of the two methods, (i) Census, or (ii) Sampling.
(i) Census
A census method is one in which one counts of all the elements of a given population. Here
determination of the distributions of their characteristics is done on the basis of information
obtained for each of the elements.
(ii) Sampling
Sampling consists of obtaining information from a portion of a larger group, or universe.
Elements are selected in a manner that they yield almost all information about the whole
universe, if and when selected according to some scientific principles and procedures. Thus,
for example, if we want to know what proportion of people from Addis favor EPRDF, then
we may select a sample of 100 persons form certain part of Addis and ask them about their
preference. Again if we want to know what the university students think about Indira
Gandhi’s 20 point programme and how far this has helped to solve their unemployment
problem then we would select a reasonable number of students from one or two universities
and ask them. However, the proportion of the sample preferring a particular political party or
the population of the students from sample favoring the twenty-point programme may or may
not be the same as their corresponding proportion in the population. The more approximately
the sample presents the true characteristics of the population, the more representative and
dependable it is. Then only the conclusions drawn on the basis of the sample will be more
clearly applicable to the population. All these depend upon how we have drawn the sample
from the population. If and when the sample is drawn on the basis of proper sampling plans,
devised and executed according to the problem under investigation and the nature of the
population, we can be more sure that repeated study on a number of different samples,
selected from a given population, would yield findings that would not differ much from each
other and also from the true population figures by more than a specified amount for a given
confidence limit. Thus it should represent not only the total population characteristics but also
the various sub-classes of the population. The validity of findings, as in above examples,
would also depend upon how people willingly and correctly report of their opinion, attitude,
58
preferences and favorableness. In nut shell it can be said that the validity or dependability also
depends upon: (i) the measuring instrument, i.e., questionnaire, observations, (ii) the
measurement procedures, and (iii) the observer’s qualities, experiences and training.
5.2.2 Why sampling
Advantages
We prefer sampling, while doing research, to the census method for it involves a large number
of advantages over the second one.
1. Sampling helps to collect vital information more quickly. Even small samples, when
properly selected, help to make estimates of the characteristics of the total population
in a shorter time. The modern world is highly dynamic. Therefore, any study must be
completed in short times, otherwise, by the time the survey is completed the situations,
characteristics, or conditions might have changed.
2. Sampling cuts costs. Enumeration of total population is much more costly than the
sample studies. Any research process includes selection of elements or objects for
study, collecting information from them, organizing the data and drawing conclusions
from them. All these are based upon a much smaller number of elements, than in the
census method. Much of time and cost is saved at each stage of research. The money
that is saved by using the sampling method can be further used for more detailed study
of the population, and for a more comparative study of it with other groups or
population and also in tabulation and analysis. Thus, it helps to obtain more
comprehensive data that could not be possible within short time and money, and could
not be available in the absence of comparative study. Often time and money available
is so short that sampling becomes not only preferable but also essential.
3. So far accuracy is concerned; sampling techniques often increase the accuracy of the
data. With small sample it becomes easier to check the accuracy of the data. Some
sampling methods make it possible to measure the reliability of the sample estimates
from the sample itself.
4. Lastly, from administrative point of view also sampling becomes easier. Study through
census method would involve the hiring of a large staff, the task of training and
supervising provides short-cut ways to solve these problems.
59
Disadvantages
1. Sampling is not feasible in a situation where knowledge about each element or unit or
a statistical universe is needed. As for example, estimate or our national income for
the current year 97 cannot be based upon sample records, similarly if any university
wants to make a quick assessment of payable amount as Bivr advances or bonus to its
employees it cannot do so on sample basis to save time or money.
2. The sampling procedures must be correctly designed and followed, otherwise, what
we can call a ‘Wild Sample’, would crop up with misleading results. Census method
should also be followed carefully. But in case of sampling study the sampling error
may be larger than expected if the sampling procedure is improperly designed or
incorrectly carried out. This leads to biased data and incorrect generalizations.
3. There are numerous situations in which the units to be measured are rare and highly
variable. Here a very large sample is required in order to yield enough cases for
achieving statistically reliable information.
4. Most of the sampling techniques require the services of sampling experts or
statisticians. Knowing statistical techniques to determine size and reliability of a
sample and to estimate sampling error is not sufficient. In research one also needs
considerable experience of dealing with and of collecting data accurately.
5. Complicated sampling plans may ultimately involve almost the same amount of time,
money, energy or labor, as is required in the census method.
6. Each type of sampling has got its own limitations.
7. Scientists or researchers are faced with additional problems in sampling because of
scattered distribution of sample units, non-cooperative nature of respondents, and
inaccessibility to respondents and so on. These problems can be solved to a great
extent when researchers are experienced and trained.
8. To know certain population characteristics, like, population growth rate, population
density, etc., census of population at regular intervals is more appropriate than
studying by sampling.
A sample can be representative of population when it has all the significant characteristics of
the population in relative proportions. Different techniques are available and are employed to
achieve representative sample. The different sampling techniques can be broadly divided into
60
two groups: (a) probability sampling techniques, and (b) Non-probability sampling
techniques.
1. Probability sampling techniques
A probability sampling technique is one in which one can specify for each element of
population the probability of its being included in the sample. Every probability can be
expressed in the form of a proportion. We can say that the probability of getting a head in
tossing a coin is ½ or say 1 chance 2 trials. Again in throwing a dice the probability of getting
any one of the six faces or way a specific face with 6 spots is 1 to 6. For certain types of
samples the degree of precision devised from the sample estimate can be stipulated in
advance. Thus, probability samples are characterized by the fact that the probability of
selection of each unit is known.
In the simplest, example each of the elements has the same probability of being included as in
random sampling method. For being a probability sample, it is not essential that all the units
have same or equal probability but that the probability of the unit being included can be
specified.
An essential quality of a probability sample is that it makes possible representative sampling
plans. It also provides an estimate of the extent to which the sample characteristics or findings
are likely to differ from the total population. Here one can also specify the size of the sample
or the size of the components of the sample required to guarantee that the sample findings will
not differ by more than a specified amount from that of the total population. Major forms of
probability sampling methods are
(i) Simple random sampling method, and
(ii) Stratified random sampling method.
2. NON- PROBABILITY SAMPLING
In non-probability sampling techniques, one cannot estimate before hand the probability of
each element being included in the sample. It also does not assure that every element has a
chance of being included. In probability sampling one has to prepare or know a list of all the
elements of the total population from which the sample is to be drawn. This makes the
sampling procedures costlier and more time-consuming, which is saved in non-probability
sampling. The major forms of non-probability samples are (i) accidental samples, (ii) quota
samples, and (iii) purposive samples.
61
5.2.3 Types of probability sampling
(i) Simple and random sampling methods
In day-to-day language the term random is used frequently to mean careless, unpremeditated,
casual haphazard activities or processes. As such a random sampling would mean a sample
drawn carelessly in unplanned manner, i.e., without a definite aim or deliberate purpose. This
concept is not correct in case of random sampling.
By random sampling is correctly meant the arranging of conditions in such a manner that
every item of the whole universe from which we are to select the sample shall have the same
chance of being selected as any other item. It also makes the selection of every possible
combination of the desired number of cases equally likely.
Random sampling, therefore, involves careful planning and orderly procedure. ‘Random
selection is often sufficiently secured by the process of spreading out a consignment of goods
etc., and marking one here and another there, avoiding the first, the last and the most obvious
ones and testing the objects marked.’ When random sampling is applied to material objects,
cards or capsules objective selection is easily achieved. But there are occasions when in
selection of individuals or goods subtle, unconscious preference or prejudice may take place
and make the sample a biased non-representative one.
Among all the probability sampling procedures random sampling method is the most basic
and least complicated.
Steps of a simple random sampling
1. It involves listing or cataloguing all the elements in the population and assigning them
consecutive numbers from I to N or putting symbols to them according to the
requirement.
2. Deciding upon the desired sample size.
3. Using any method of selection, discussed later on, a certain number of elements from
the list of elements I to N is selected.
4. One can then interview the people or measure the items designated by the sample
elements.
62
Advantages of random sampling method
1. It is the most basic, simple and easy method. It, therefore, requires less money, time
and labor.
2. It provides a representative sample. As the sample size is enlarged, it becomes
increasingly representative of the universe. Only when the random sample contains a
small number of cases it may deviate considerably from the universe.
Disadvantages of random sampling method
1. It is in most of the cases difficult to find an up-to-date list of all the units in the
population to be sampled. For random sampling we need a completely catalogued
universe.
2. The task of numbering every unit before the sample chosen is time-consuming and
expensive, except in regular interval samples.
3. The units need not only to be numbered but also to be arranged in a specified order.
4. As the size of the sample required is often large enough in order to achieve statistical
reliability and representativeness, the cost and time consumed becomes equal to the
stratified sampling method which is much more reliable.
5. From the point of view of field surveys, it has been claimed that cases selected by
random sampling method tend to be too widely dispersed geographically and that the
time and cost of going from one address to another is too large.
6. The possibility of obtaining a poor or misleading sample is always present when
random selection is used.
Methods of drawing sample in random method
(a) Lottery method – in this method the numbers of all the elements of the universe are
written on different tickets or pieces of papers of equal size and shape and color which are
then shuffled thoroughly in a box bag or pot. Then the tickets are drawn at random, their
numbers noted and the corresponding individuals or objects are studied. The numbers of
tickets drawn are equal to the sample size desired.
This method is followed in our day-to-day life like in allotment of questions in the practical
examinations or in drawing lottery tickets for money.
The process is thus designed to maintain objectivity and avoid any type of preference or bias
towards any one unit.
63
(b)Tippet’s Numbers – This method of drawing samples was first developed by Prof. L.H.C.
Tippet and since then is known by his name. He developed a list of 10,400 sets of numbers
randomly, each set being of four digits. These numbers were written on several pages in
unsystematic order. Examples of such 30 sets developed by Tippet are given below.
2370 7483 3408 2762 3563 1089

0560 5246 1112 6107 6008 8126
2754 9143 1405 9025 7002 6111
6641 3991 9792 7979 5911 3170
9524 1545 1396 7203 5356 1300
This method was developed and suggested by Tippett to increase the accuracy of random
sampling so that the researcher selects the elements by not seeing the elements directly but
from the list. Suppose we want to select a sample of 30 students from a population of 5000
students of a high school who matriculated this year. For the purpose of selection through
Tippet’s method we have to follow the steps given below:
(i) First of all, prepare a complete list of the universe. Every name has to be written in
systematic order. In case of our example we can consider the examination roll number
of the students as their serial number, since list of roll number is prepared
systematically and alphabetically and is available.
(ii) Select any thirty number from any page of Tippett’s list systematically. Often the size
of the universe is smaller than that expressed in Tippett’s list. In such case we can pick
up the first two or first three or last two or last three digits of each of the 30 sets
selected from the Tippett’s list. Comparing our prepared list with these selected lists
we can select those persons who carry the same number.
(iii) Suppose for our universe of 5000 students we decide to make selecting on the
basis of first three digits. Then the following process would be followed to select
numbers:
64
237 0 748 3
056 0 524 6
275 4 914 3
664 1 399 1
952 4 154 5
Then the persons carrying the number like 237, 056, 275, 664, 952, 748, 524, 914, 399, 154
and so on would come under our sample.
Tippet prepared this list for selection of elements from population. It is not essential to use his
specified list of number. One can develop his own list. The basic principle is selecting the
number at random.
In selection of numbers form such types of lists two criteria have to be kept in mind; viz., 1.
Set of number can be considered or selected once only and 2, zero number cannot be selected.
As for example if Tippett’s sets are 9772 and 9734, and we are considering first two digits
only, then in both cases the numbers would be 97(92) and 97(34). Thus, the elements carrying
the number 97 has to be included but should be included only once.
(a) Grid method – this method is applied in selection of areas. Suppose we have to
select any number of areas from a town or any number of towns from a province for survey.
For selection, first a map of the whole areas is prepared. The area is often divided into
different blocks. A transparent plate is made equivalent to the size to the map that consists of
several squared holes in it which carry different numbers. By random sampling method it is
decided as to which number are to be included in the sample.
(ii) Systematic sampling method
In this method first of all a list is prepared of all the elements of the universe on the basis of a
selected criterion. A list may be prepared in alphabetical order, as given in the telephone
directory. Then from the list every third, every tenth, every twenty or any other number in the
like manner can be selected. We can start with any number, say five, then every tenth number
will be 5, 15, 25, 35, 45, 55 and so on. For the application of this method, preparing a list of
all the elements and numbering them is essential. Secondly, the population need to be
homogenous in nature since social phenomenon is variable in nature and the individuals are
heterogeneous in their social characteristics on which they are homogenous. Thus we may
65
decide to cover only students, professors, slum-dwellers and so on. The characteristics to be
selected for this purpose must be relevant to the problem under study. In other words, the
research worker decides upon a sampling fraction or percentages of the population which is to
be included in the final sample.
(iii) Stratified random sampling method
We have seen before that in case of simple random sampling the accuracy of the sample is
increased by increasing the size of the sample. The same result can be achieved without
increasing the size of the sample through the application of stratified random sampling
method. The stratified random sampling method can be applied only when the population
characteristics are known.
1. Definition: When the population is divided into different strata or groups and then samples
are selected from each stratum by simple random sampling procedure or by regular interval
method we call it stratified random sampling method.
The characteristics and utility of stratified sampling method may further be clarified by
illustrating the steps or procedures followed.
2. Steps: (a) according to the nature of the problem relevant, criteria are selected for
stratification. Among possible stratifying criteria; come age, sex, family income, number of
years of education, occupation, religion, race, place of residence and so on.
(b) On the basis of the characteristics, universe can be divided into different strata or
stratum. Each stratum has to be homogenous from within. Such a division can be done on the
basis of any single criterion. As for example on basis of age we can divide people into below
25 and above 25 groups, and on basis of education into matriculates and non-matriculates.
Again, stratification can be done on the basis of a combination of any two or more criteria,
viz., on the basis of sex and education we can divide the people into four groups, (i) educated
women, (ii) uneducated women, (iii) educated men, and (iv) uneducated men.
To stratify, therefore, simply means to place every member of the population into its proper
stratum as determined by the criteria used for stratification. In case of the above example, sex
and education of each member of population are to be used for developing categories and then
each member could be placed into any one of these categories.
66
(c) Elements are then selected from each stratum through simple random sampling method.
An estimate is made for each stratum separately. These estimates are combined to provide an
estimate for the entire population.
1. Purpose: The primary purpose of the stratified sampling method is to increase the
representativeness of the sample without increasing the size of the sample on the basis
of having greater knowledge of the population characteristics.
(iv) Cluster sampling
In cluster sampling the stratification is one in a manner that the groups are heterogeneous in
nature rather than homogenous. Here elements are not selected from each stratum as is done
in stratified sampling, rather, the elements are obtained by taking a sample of groups and not
from within groups. This means that out of several clusters or groups one, two, or more
number of clusters are selected by simple or stratified random method and their elements are
studied. If all the elements in these clusters are not to be included in the sample, the ultimate
selection from within the clusters is also carried out on a simple or stratified random sampling
basis.
Suppose for example, we want to study the socio-economic background of the secondary
school teachers of Addis Ababa and its impact on their teaching motives. For this all the
teachers of secondary school level from Addis will constitute the universe. For clustering
purpose, we can then make a list of all those sub-cities of Addis state which have high or
secondary schools. These school sub-cities would vary from each other in the number of
schools and their status. Each school district would be heterogeneous form within, for each
would contain schools of different status, standard and affiliation. The teachers would vary
more among themselves. Each school district can be considered as a cluster and be assigned a
cluster number. Then we can select a certain number of clusters from the list either by simple
random method or by systematic techniques or by stratified method. All the elements from
these selected clusters can be included in our sample, or from within the clusters a certain
number of elements can be selected on a simple or stratified random sampling basis.
Goal or purpose – the purpose of a cluster sample is to reduce cost and not essentially to
increase precision. However, it does not mean that cluster sampling method does not take care
67
of precision. On the other hand, it seeks to maximize the precision per unit cost by reducing
the cost as much as possible, while engendering only a small loss in precision.
These two goals can be fulfilled only when natural groupings are used for clustering. These
natural groupings are not too homogenous. Examples of such natural groupings are school
districts, voting districts, city blocks, natural geographical boundaries and so on.
v) Multistage sampling
Normally a multistage sampling procedure is one that combines cluster and ratified sampling
methods.
For example, if we want to study the socio-economic background, attitudes and motivations
of slum-dwellers, we can first make a list of the cities which would thus make our cluster.
From these clusters we can select any number of cities. Then each city or cluster would be
stratified into different slum areas. A few slum areas can be selected on the principles of
cluster sampling and then further from them we can select a sample of households.
Thus, our cities can be called primary sampling units and the slum areas secondary sampling
units. This process of creating sampling clusters within clusters can be continued according to
need. Such a process of continuous cluster sampling is called multistage cluster sampling.
5.2.4 Types of non-probability sampling
It has been mentioned before that a non-probability sampling is one in which one cannot
estimate beforehand the probability of each element being included in the sample. Also that it
does not assure that every element has a chance of being included. The major forms of non-
probability samples are (i) accidental samples, (ii) quota samples, and (iii) purposive samples.
(i) Accidental samples
Accidental sampling means selecting the units on basis of easy approaches. Here one selects
the samples that fall to hand easily. Suppose one is studying political socialization and
political participation among the university and college students of Unity University and that
his sample size in 100. He would go to the university campus and would select the first
hundred students whom he happen to meet, whether in the class room, or in the students’
common room or in the football field.
68
Such type of sampling is easy to do, and saves time and money. But the chances of bias are
also great.
(ii) Quota sampling

In quota sampling the interviewers are instructed to interview a specified number of persons
from each category. In studying peoples’ status, living conditions, preferences, opinions,
attitudes, etc. the relevant basis of stratification usually used are age, sex, education, place of
residence, socio-economic status, religion and so on. The required numbers of elements from
each category are determined in the office ahead of time according to the number of elements
in each category. Thus, an interviewer would need to contact a specified number of men and
specified number of women, a certain number of persons from different age categories, from
different religious or racial groups and so on.
The basic goal of quota sampling is the selection of a sample that is a true replica of the
population about which one wants to generalize.
This means that the purpose of quota sampling is to provide a time representative sample of
the population.
Purposive sampling
In purposive sample one picks up the cases that are considered to be typical of the population
in which one is interested. The cases are judged as typical on the basis of the needs of the
researcher. Since the selection of elements is based upon the judgment of the researcher,
purposive sampling is called judgment sample. The researcher tries his sample to match the
universe in some of the important known characteristics. It is assumed that if the sample is
representative with respect to known population characteristics, it will also be representative
with respect to the unknown characteristics. Such assumption is wrong.
The defect with this method is that the researcher can easily make error in judging as to which
cases are typical. Any type of judgment sample precludes the calculation of estimates of the
likely range of errors, and thus is not dependable for any scientific work.
Check your progress
1. What is sampling?
69
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
2. Why do we sample research subjects?
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------
1. Distinguish between non-probability and probability sampling techniques?
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
---------------
2. List and define the different types of probability sampling?
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
---------------
3. List and define the different types of non-probability sampling?
70
5.3 SUMMARY
Sampling is defined as the process of obtaining information from a portion of a larger group,
or universe. We have seen two major forms of sampling: Probability and non-probability
sampling techniques. We have also seen specific divisions of each type of sampling
technique.
1. Sampling consists of obtaining information from a portion of a larger group.

2. It is not possible to include all subjects in a study for cost, energy, and resources.
3. In probability sampling each element of population has an equal chance of being
included in a study, whereas this cannot be estimated before hand in non-probability
sampling.
1. Probability sampling includes
a) Simple random sampling
b) Systematic sampling
c) Stratified
d) Cluster sampling
2. Non-probability sampling includes: a) accidental samples, quota samples, purposive
samples etc.
71
UNIT SIX
TOOLS OF DATA COLLECTION
Aims and Objectives
At the end of this unit you should be able to
 list the major tools of data collection
 describe the specific purposes of the tools of data collection
 describe the advantages and disadvantages of the tools
6.1 INTRODUCTION
Once the researcher has identified the subjects to be involved in the study using a sampling
technique appropriate to the study, he/she devises a method or methods of obtaining the
essential data from the subjects. These thus are usually referred to as tools of data collection.
Three tools of data collection most commonly employed include questionnaires, interviews
and observation.
6.2 TOOLS OF DATA COLLECTION
6.2.1 Interview method
The different methods of data collection used in research help us to gain accurate and
scientific knowledge about the individuals and their society as far as possible. The important
methods of data collection are observation, case study, content analysis, questionnaire,
schedule and interview. Among these the first three methods, that is, observation, content
analysis and case study can be termed as qualitative methods in the sense that they provide
information about the phenomenon in qualitative, descriptive and unstructured form. The
questionnaire method and the schedule can be called quantitative methods as they provide
information in most structured form. The interview method varies from being most highly
qualitative method, as in case in the depth interview and non-directive interview methods to
the most quantitative ones, based upon structured schedule. The less structured it is the more
quantitative would be the information. On the basis of its structure the interview methods can
be divided into several types. The different methods of data collection are very similar to each
other in one sense that they all are related to specific sets of enquiry of their own, that is they
are purposive. Secondly, all of them are systematic and aim to be objective. They differ in the
type of information they can provide, in the nature of the problem to which they are suited,
and also in the extent of their scientific validity.
72
The observation method is mainly suited to the study of culture, institutional process or
human behavior. It cannot be used to gain information about a person’s perceptions, beliefs,
feelings, emotions, motivations, anticipations, future plans, etc. Such types of information can
be collected through questionnaire, interview and projective technique methods. These three
methods differ among themselves in their directedness and structuredness. The questionnaire
method is the most direct method in which a set of questions are put before the respondents
for specific information. Projective method is the most indirect method in which the
respondent is given some ambiguous stimulus, like a blurred picture, a blot of ink, or a vague
question. The respondent is asked to give his impressions of the stimulus. In doing so he gives
the needed information without being aware of doing so. Interview method uses both direct
and indirect techniques.
1. MEANING
In simple words interview can be called the process of talking in more purposive and more
systematic manner than our day-to-day gossiping with each other. The purpose and process of
interviewing is clear from the following definitions.
“Interview may be regarded as a systematic method by which a person enters more or less
imaginatively into the life of a complete stranger. This definition clarifies the process. In
interview method the researcher tries to penetrate deeply in his imagination into the
circumstances presented by the subject. He tries to enter into the feelings of the respondents
but at the same time tries to maintain the objectivity of research while studying about the
individual. By studying each individual’s behavior or actions and reactions he tries to locate
the feeling, thinking, or motivation of the people that guide their behavior, actions,
interactions, and interrelations. The purposes of interview are if we want to know how people
feel, what they experience and what they remember, what their motives and emotions are like
and the reasons for acting as they do why not ask them?
The definition by F.N. Kerlinger points to the relationship between interviewer and
interviewee and the purposes of interview. To him “The interview is a face-to-face
interpersonal role situation in which one person, the interviewer, asks a person being
interviewed, the respondent, questions designed to obtain answers pertinent to the purposes of
the research problem.”
73
Advantages of interview method
1. There are some kinds of information which are virtually impossible to be obtained by
any other means. For example, information about a person’s past experiences, his
reactions to such experiences, his anticipated future behavior, and a report of his
thoughts while carrying out an activity, can be obtained only by asking for a verbal
report from the persons concerned. Also, where the intention is to study perceptions or
attitudes then asking persons to describe what they see and how they feel is the only
possible way to obtain the information.
2. A second advantage is the directness involved. As long as the researcher is assured
that the respondents do not distort their descriptions of attitude and perception, the
interview method is the most simple, direct and valid approach to use. It avoids the
difficulties and disagreements which occur in an attempt to infer the attitudes and the
perceptions from the behavior through observation.
3. The interviewer can modify the situation whenever necessary. He can clear up a
misunderstanding about a question and keep the respondent on the track of providing
only essential information. This becomes very important when working with the
children. Also, the interviewer can note special happenings and pick up clues which
might prove to be highly valuable in interpreting the results and for future study.
Finally, in ‘action research’ where the interest is as much to bring about change in
those involved as is in finding out information, the interviewer can set the stage for the
development of certain attitudes in the respondent.
Types of interview
There are mainly two types of interview: (i) structured, and (ii) unstructured. The type of
interview method to be used depends upon the nature of the problem being investigated and
the type of the information wanted. The different types of interview differ from each other in
their structure, process, problems, and the quality of validity and reliability. These will be
discussed one by one.
A. Structured interview
Structured interview method is the one which is based upon structured set of questions and is,
therefore, highly standardized in form or content. Unstructured interview method is one which
does not make use of any set of structured questions but only of semistrucutred interview
74
guide. Therefore, it may vary from being semi structured to the most qualitative one. In the
structured interview method, information is collected by the investigator by directly asking
the respondents on the basis of a schedule. In it not only the type of information sought, but
also the specific questions to be asked, the language of the questions and their order is pre-
fixed. Therefore, the interviewer does not have the freedom to ask extra questions. He can
only, form time to time, make use of probe questions and thus encourage the respondents to
answer the questions. Here the interviewer’s job is to see that the respondents. The
interviewer can clarify the questions to the respondents only by repeatedly reading out the
questions or interpreting them in specified language.
Different types of questions (already discussed under questionnaire) are used in structured
interview method, like the fixed alternative type in which not only the language of the
question is structured but the answer is also structured either in the form of yes/no alternative
or in the form of categorical answers, as given below:
Show your agreement or disagreement with the statements given below by putting a tick mark
before the most appropriate category –
1. Are you married? yes no
2. Democracy in present socio-cultural and economic system of Ethiopia is more harmful
than advantageous. yes no
3. A democratic undertaking by the government is defective.
4. Yes no
5. Second type of questions used are the open-ended questions in which the questions are
written in specific language but the respondents are free to give answers in their own
language, as given below.
1. What advantages do you expect from our present government?
2. Do you think that the democratic form of government can maintain socio-political
stability in our country under the present social structure? If yes, how?
The set of questions used in the structured interview method may be any of the above types or
a mixture of them. The types of questions used determine the structuredness of the interview
method.
75
Structured interview method is used when data is to be collected from a large number of
persons. It helps in systematic collection, comparison, organization and analysis of data
within a limited time period. Comparability is lost if questions are not asked in the same
language and in the same study quite different. For example, given below are two different
questions on democracy. Both the questions aim to assess the people’s attitude towards
democracy but ate different in their language, which makes the meaning quite different.
Question asked by interviewer No. 1 –
Show your agreement with –
Democracy is the best form of government,
Agree Disagree Neural
Questions asked by interviewer No. 2 –

Show your agreement with.
Democratic government is considered as best by the people, as it provides full opportunity of
political participation and freedom of expression.
Agree Disagree Neutral
Similarly, change in the order of questions also creates different amount of emphasis on
different questions and hence the answers are not comparable.
It is always better to use open-end question, unless impossible, because respondents are likely
to give answers, more freely according to their real attitude. They are not indirectly compelled
to place themselves in any one of the given categories.
If the above rules are followed in establishing rapport, in asking questions and in recording of
answers one is most likely to get reliable and valid data. However, there is another type of
problem which might affect the success of any survey work.
Usually 88% to 90% of the selected units in the sample produce completed interviews. This
leaves between 10 and 20 percent of the address that result in a non-response of some kind.
There are two types of non-response that result in a non-interview and the non-sample. This
can happen in case of both structured and unstructured interview methods.
76
(a) The non-interview – if the interviewer does not get an interview with the
eligible respondent it is a non-interview. A non-interview may result from
conditions beyond the interviewer’s control. For instance, the eligible
respondent may be too ill to be interviewed or unable to speak, or not available
at home. Such non-interviews affect the response rate because there is an
eligible respondent who was not interviewed.
(b) The non-sample – when a selected dwelling unit is found to be vacant on when
in the original address an eligible person by study definition does not live, or
when an eligible respondent refuses to answer it is known as non-sample non-
response situation. One has to develop better means and ways to overcome
such situation.
B. Unstructured interviews
I. Purpose
Unstructured interviews are used mostly in exploratory situations, as follows –
a. For providing background for defining a conceptual model and stating
an objective in an exploratory study.
b. For serving as the basis for constructing more highly structured
questionnaires by showing the general framework within which
respondents think, the kind of words they use and the types of incidents
which have meaning for them.
c. For explaining an ambiguous finding from an extensive study.
II. Nature
In this method the questions are not essentially preplanned, structured, or ordered. The
investigators are given some broad topics upon which they have to collect information. The
respondents are free to narrate their experiences and incidents of life. They can express their
attitude or feelings towards a given event or narrate their personal experience in their own
language. This sort of interview is, therefore, known as narrative interview. In unstructured
interview the investigator must be able to handle lengthy conversations. He should be capable
of collecting information on highly personal matters. Often the arrangement and working of
the questions are to be determined on the spot by the investigator. The investigator must also
be able to analyze quickly what the respondent says, so that he can probe for further
77
information. He must know how to deal with both types of respondents, i.e., who do not ‘open
up’ in an interview situation and also respondents, i.e., who do not ‘open up’ in an interview
situation and also with those who wander off the subject with a flood of irrelevant
information.
6.2.2 Questionnaires
Meaning
Introduction
By questionnaire is meant a set of questions developed in an organized and ordered manner
for gaining information from the people in relation to a given problem. Questionnaire method
of data collection is the tool which is more frequently used in mail survey research than any
other method of data collection. Mail survey research is that branch of scientific investigation
which studies the universe by selecting a sample from it. In survey research one attempt to
discover the relative incidence, distribution and inter-relation of sociological and
psychological variables. Though the approach and technique of survey research can be used
on any set of objects that can be well defined, survey research mainly focuses on people, the
vital fact of people, and their beliefs, opinions, attitudes, motivations, memory, behaviors,
actions, interactions and even future plans. The questionnaire is the best tool for collecting
information in all above areas, though besides is the manner through which information is
collected. Secondly, in mail surveys the percent of return is small-i.e., between 20 and 40
percent on the average, as compared with interviews which typically result in about 95
percent return. The users of mail survey are almost always faced with the question of how to
estimate the effects the non-respondents may have had no the results.
Functions of questionnaire
Questionnaire method, as a tool of scientific enquiry, fulfils several purposes or functions:
measurement, description, comparison, and inference. These will be discussed one by one.
(a) Measurement
Similar to the other methods of data collecting the questionnaire method also aims to measure
different variables. Through it we collect information about individual or group
characteristics, like, sex, age, height, weight, years of education, occupation, income, political
affiliation, religious preferences, membership of different groups and organizations, food
habits, picture-going habits, and so on. On the basis of measuring these characteristics
78
through questionnaire, we are able to derive conclusions about more abstract and broader
concepts, like, social distance, attitude, perceptions of group cohesiveness, degree of social
prejudice, alienation, etc.
It is much easier to measure accurately the more concrete concepts like, age, sex income,
height, weight, etc. than to measure the abstract broader sociological concepts, mentioned
above, which therefore need to be operationally defined with the help of observable concrete
concepts or indicators.
(b) Description
On the basis of the measurements achieved of the above mentioned characteristics, we are
able to provide description of a given population. Since through questionnaire we are able to
achieve more concrete numerical measurement of characteristics, it becomes easier to
describe the population characteristics in more specific terms. This helps in more specific
comparison of different samples of related populations.
(c) Inference
It helps to infer about the total population on the basis of sample studies. Descriptions
achieved through it help not only in better understanding of the people but at the same time it
also helps in explaining why and how peoples’ attitude, beliefs, motivations, behaviors, etc.
vary from one group to another. It means that on the basis of such descriptive information we
are able to find out association or correlation between different factors. We can see whether
the ability for job efficiency of factory workers vary with difference in their age, education
and sex.
There are several factors which create difficulty in the measurement of social characteristics
as accurately as found in natural sciences. This hinders the accurate scientific description,
comparison and inference. Therefore, the methodologies are constantly attempting to improve
the questionnaire by structuring and restructuring the statements or scales used in them.
Types of questionnaire
Questionnaire can be divided into different types on the basis of the type of response required.
Type of response required
Response may be (a) fixed or closed type, (b) open-end type, and (c) mixture of both a and b.
(a) Fixed or closed response
79
In the fixed or closed response type of questionnaire not only the questions but also the
answers are already given in the questionnaire. A respondent is asked to check the answer that
fits him best. The fixed or closed type of questions can again be sub-divided depending upon
the number of alternative answers given, into dichotomous alternative type, multiple choice
type, and rank order type.
Use of fixed alternative questionnaire depends upon the degree of the researcher’s knowledge
of the people in the sample, upon the people’s knowledge of the subject matter under
investigation, and upon the assumption that the people would be capable of reflecting their
attitudinal positions of varying intensities on the issues mentioned in the questionnaire, by
placing them in any of the categories that reflect their attitude and opinion in the best manner.
Advantages of fixed questions
1. Fixed-response items are easy to reply, to score and to code.
2. No writing is required on the part of the respondent. Thus, it helps in acquiring answer
quickly.
3. In cases where respondents cannot adequately express themselves verbally, the fixed
response item is an advantage.
4. Fixed response items facilitate completion of the questionnaire.
5. In lengthy questionnaire fixed response items are completed more rapidly than the
questions with open end responses.
6. In case of mailed survey, percentage of return is greatest when the questionnaire
consists mainly of fixed alternative answers.
Disadvantages
1. There is a possibility that the researcher would fail to provide with all relevant
response alternatives or exhaustive categories. Sometimes respondents may be forced
to make a choice among alternatives that do not fit them well. This leads to wrong
information.
2. In fixed response items used in attitudinal studies the respondent may lapse into a
response set. In order to finish it the respondent may not go through each question and
its possible alternative answers patiently but may answer all the questions in the
pattern of response to the first questions. This is again misleading.
3. Open-end response
80
4. This type consists of questions which are written in specific language but the
respondents are free to give answers in their own words, for example,
5. ‘Do you think the present government will be able to solve the problem of price rise’?
Advantages of Open-end types

1. These are useful when the researcher has little or no knowledge of the subject matter
or the topic being studied. This is very useful when a detailed /deep study is needed.
2. The respondent is less restricted or is not forced to reply in terms of any of the given
choices.
3. It helps in gaining insight into the opinion and behavior of the people.
4. It may elicit unanticipated and insightful replies from the respondents which help the
researcher to have a better understanding and greater knowledge of the people and
subject.
Disadvantages of open-end responses
1. It involves difficulties of coding, classifying, grouping and comparing of the

responses.
2. A bias exists in open-end questions that stem from different sources. For example,
persons who cannot express themselves adequately on paper are combined with fluent
persons. Moreover, persons from different socio-economic status do not see things
similarly and do not use same vocabulary to express ideas.
(c) Mixed type response
Most of the questionnaires are of mixed type. This means that they consist of both fixed and
open-end type of questions. A questionnaire may consist of varieties of enquiries. For
example, it may consist of factual enquiries about an individual’s social, economic, marital,
and occupational status. It may also consist of attitudinal questions related to individual’s
beliefs, feelings, motivation, consciousness, attitude and opinion, etc. Such factual and
attitudinal questions can easily be of fixed types. On the other hand, many questions are
related to mental process and need to be of open end-type being narrative or descriptive in
nature. This is also applied when the researcher is not able to anticipate probable replies from
the respondents, or is not familiar with the population under study.
81
Construction of questionnaire
Questionnaire construction may apparently seem to be a simple task but in reality is not. This
is evident from the processes and the considerations involved in the construction of a
questionnaire.
The development of a questionnaire, whether it is to be used in a personal interview, it to be
self-administered, or is to be mailed usually requires to be processed through at least three
progressive steps, namely,
A. Examining the conceptual model.
B. Framing of standardized questions.
C. Pilot study or pretesting.
D. Printing of the questionnaire.
E. Types of questionnaire administration.
Although no two research problems are exactly alike, yet the above mentioned steps are
necessarily to be followed.
(A) Examining the Conceptual Model
It is done through literature survey, exploratory interviews and analyses. This is done for
explaining the model and for locating the relevant variables. When the variables are not
known and properly defined, it becomes difficult to decide as to what information is to be
sought.
(B) Framing of a questionnaire
In framing of a questionnaire one has to keep in mind the following considerations.
(a) Form of questions and answers – first of all one has to decide about the general strategy of
the questionnaire. One has to decide whether structured or unstructured, direct, indirect or
projective questions would help to achieve research objectives. Such decisions have to be
made early at the exploratory stage. Thereafter the form of each question has to be decided
whether they should be open or closed, loaded or non-loaded, be single or be in sets, direct or
indirect. Similarly, structure of answers or form of response has to be decided. Should they be
dichotomous, multiple choice, scaled, short answered, (limited to a word or two) or of open-
end type? The form or structure of the questionnaire depends upon the nature of quality of
information wanted and the type of the respondents. The respondent may consciously or
unconsciously, be reluctant to express his orientations, or may report it wrongly or may be
82
incapable to report some of the actions or orientations towards political, social or religious
values. Hence the researcher should provide indirect questions in the questionnaire in order to
prevent concealment or distortion.
Whatever be the format of the question they must be easily understandable.
(b) Wording of the questions – (i) Language: Questions should be framed in the common
language of the respondents so that they are easily understandable. In case the questionnaire is
to be administered to educated or technical persons, then the language of a higher level or of
technical type can be used.
Ideally only those words should be used within the questionnaires which carry same meaning
to every person. The use of structured questions assumes that all respondents will understand
the same thing by the same word. Unfortunately this does not happen always. Ambiguous or
loaded words and long questions should be avoided as they are hard to follow and yield
biased answers. They can be used only when they serve some specific information.
(c) Level of information – it is not essential that the respondent will be acquainted or familiar
with the topic under investigation and that he will be able to give some or to answer or
information on the same. The respondent may quite naturally pretend to know and provide
information on the topic rather than say that he is not well informed. This can be checked by
using probe questions.
(d) Specifying the questions – unless one wants to explore the ways in which the respondent
will interpret a question, one should establish the frame of reference one has in mind. For
example, if one is investigating into family disagreements it is essential to differentiate
between situational causes and role conflicts. Specifying the questions in terms of time and
place of the occurrence of an event helps in avoiding the bias that results form loss of
memory.
(e) Predisposing the answer – frame of the question should suggest all possible alternatives
for the respondents so that the respondent selects the category which represents his views to
the nearest extent and does not select answers suggested by the investigator. The frame of
questionnaire should not be one that suggests any particular answer.
83
(f) Protecting the respondent’s ego – if one wants to study the respondent’s unpleasant
orientations or his negative attitude then one should in the beginning give the respondent a
chance of expressing his positive feelings first and then the respondent should be asked to
express his negative attitude or feeling.
(g) Asking direct/indirect questions – respondents may consciously or unconsciously give
wrong or incomplete answers. Therefore, both direct and indirect questions should be
included.
(h) Sequence of questions – the questions is put in a particular sequence in order to protect the
respondent’s ego. This means that they do not hurt or discourage the respondent to give
accurate answers. If personal and intimate questions are asked later, there is least resistance
from the respondent. Asking more impersonal questions first gives the respondent a chance of
feeling at ease. If later on the respondent becomes excited on asking personal questions, it
does not interfere with the quality of answers to the previous questions.
The opening question is important. One should start with easy questions, so that the
respondent would enjoy answering. One should not condition answers to subsequent
questions by preceding ones. This may be accomplished by funneling. This means that in the
beginning more general and easy questions should be asked and later on one should slowly
switch over to more specific and difficult questions. This procedure establishes the frame of
reference gradually and helps the respondent to think more explicitly at successive stages.
Funneling may also be accomplished by starting out with open questions and then proceeding
to closed ones. In special cases one may use it in the reversed way, which means that the
closed questions should be asked first and open-end questions later on.
(i) Categories for recording and coding – the categories to be used for recording of answers and
for classification have to be determined in advance. If one waits until the data are in before
deciding on categories one may fit the concepts too closely to the obvious data and loose
some of the significance of the priori conceptualization.
If the categories are developed and the tables are planned in the beginning then it is helpful to
the researcher in numerous ways. It helps to
(i) Visualize the relationship one seeks to,
(ii) To check whether the questions have adequately covered the variables needed,
and
84
(iii) To check whether the sample is large enough to fill the cells.
Use of closed answers saves time and reduces mechanical errors of preceding the answers,
i.e., printing the IBM code numbers alongside of the various answers.
In case of self-administered questionnaire this type of marking may confuse the respondents
and therefore should be avoided or the answers may be numbered.
(C) Pilot study or pretesting
After the drafting of the questionnaire it is pretested through what is called as pilot survey. At
this stage the questionnaire is finally edited to ensure that the content, form, sequence of
questions, spacing, arrangement and appearance etc. of the questions are checked. The object
of pilot survey is to make the questionnaire easily understandable to the people and to
eliminate the probability of misunderstanding, confusion and bias. Such confusing and
problem questions are reformulated and rechecked.
For pilot survey a very small sample is selected from the same population upon which the
actual study is to be made and the questionnaire is applied on it.
It is always better to do the pilot by interviewing the people on the basis of the questionnaire,
so that through experience the interviewers are able to locate the weak points of the
questionnaire.
(D) Printing of the questionnaire
After the pretesting and final correction we come to the printing of the questionnaire. While
printing the questionnaire we have to keep the following in mind. For example,
i. The physical format must be attractive and unambiguous.
ii. The auspices and statement of purpose must be crystal clear.
iii. There must be some inducement for the respondent to co-operate. The subject matter
should preferably be of demonstrated interest to him.
iv. The guarantee of anonymity must be clearly specified and systematized through the
use of sealed envelops and other such devices.
v. The set-up of questions and answers spaces must be very clearer and consistent, with
explicit instructions on how to answer. Whatever be the type of the questions they
must be easy to answer.
85
vi. The entire form should be attractive and realistically spaced so as to maintain interest
and minimize boredom.
(E) Types of questionnaire administration
When the questionnaire has been printed finally and is ready for collection of data, the
researcher needs to make preparations for its application, depending upon the method of
administering the questionnaire. The different methods of administering the questionnaire are
i. Interview method;
ii. Distribution method; and
iii. Postal delivery method.
(i) The interview method
In this method with the help of a trained interviewer, data is collected from the respondents.
This method has the advantages of depth study with a very low probability of the
interviewees’ (respondents) giving incorrect or inadequate answers due to misunderstanding
of the questions. A high rate of questionnaire completion is expected.
If data is to be collected through interview or through distribution method, the field

investigators must be given the following instructions.
i. What are the objectives of research?
ii. What should be the manner of asking the questions and recording the answers?
Whether the interviewers have to follow the exact wording and order of the questions
or can change it if needed? and
iii. What should be the manner of distributing and collecting questionnaires?
(ii) Distribution method
This method consists of distributing the questionnaire to the respondents at their households,
work places, or in any other setting, depending upon the nature of the problem to be studied
and the nature of the universe. The completed questionnaires are collected by the investigator
later on. This method has the advantage that the respondents can fill in the questionnaire in
privacy without the fear of replies being overheard. Moreover, the time delays and the
expense associated with follow up interviews are avoided in it. The great disadvantage is that
since the questionnaire are self-administered, there is the risk that the respondent may
willingly or unwillingly provide wrong information. This disadvantage can be removed only
86
if the distributor of the questionnaire is available to help the respondents in case of
difficulties.
(iii) Postal delivery method
This method, often called mailed questionnaire method, constitutes distributing the questions
through postal delivery along with a letter of request for postal return in pre-paid envelops.
Then a follow up letter is sent for return of the questionnaire. The disadvantages of this
method are those associated with self-administered questionnaire. Besides, it is not applicable
to a sample with a low level of education. It also involves uncertainty of response or very low
percentage of returns. The only advantage is the speed of distribution. Another difficulty is
that the researcher cannot know whether the questionnaire was completed by the particular
respondents or by other persons known to them. Here also the possibility of misinterpretation
of question by the respondent is always there. Respondents may not like to reply the questions
which are highly personal private, or sensitive. The main advantage of this method is the
privacy in which the questionnaire is filled in. It involves low cost of data collection as the
cost of appointing and training of the interviewers and of traveling is eliminated.
Among the three methods of administering the third one is more commonly used. Use of self-
administered questionnaire method, neither through distribution nor through mail, has created
new type of research roles, namely, (a) questionnaire administrator, and (b) supervisor.
6.2.3 Observation method
Meaning: scientific and non-scientific observation
We all, in our day-to-day life, are constantly observing and noticing the events going on
around us. Often such observation may lead to some decision. As for example, while walking
on the street we notice people and vehicles moving towards different directions. On the basis
of such observation we may make a decision of crossing or not crossing the street. According
to such decision we may act also. Another example of simple observation of daily life is like
observing the events going on in the street, in the market place, at home, in the exhibition, etc.
We also make observations, while driving. We have to see whether there is green signal or
not. Besides this, we have to watch the street conditions, other moving vehicles, and also the
persons moving on foot, and on cycle, and drive them only, so that no body jumps before the
vehicle all of a sudden and creates accident.
87
But such observations cannot be called scientific observation because they are not essentially
related to a set of systematic questions. Here we are observing several events simultaneously
but not categorically. Also we do not aim to make a general conclusion in relation to the
factors determining accidents, or the driver’s behavior. Here we are simply interested to make
a decision of crossing the street or further driving the vehicle. Scientific observation has to be
systematic, planned and categorical in providing description, in establishing relationship
among the events, in locating the causes behind them and thus in testing of the hypothesis.
Examples of scientific observation are also many. One may want to study the manner in
which the mothers rear their children in different cultures, or say how child rearing practices
are function of culture. It has also been used to study intergroup behavior or intergroup
relations, differences in ceremonies or rituals, or the behavior of different religious and racial
groups, play behavior of children, and development of child personality and leaderships
quality in them.
Essential characteristics of scientific observation
There are certain essential characteristics which must be fulfilled for an observation to be
scientific. These are given below.
(i) It must serve a formulated research purpose
(a) Observation may serve a variety of research purposes. It may be used in exploratory
research to gain insights into the problems which can be later tested by other techniques of
data collection. It may be used to gather supplementary data to interpret the previous findings,
obtained by other methods. It may be used as primary method to obtain description of a given
phenomena or to provide explanation of specific problem or events in terms of their causes,
that is to test a casual hypothesis.
(b) Like other methods observation method can also be used to verify and extend the
generalizations or theories formed on the basis of other studies. Thus, through verification,
observation method helps to establish the validity of such generalizations and theories.
(ii) Observation must be planned systematically
Planning systematically means to spell out the steps. It consists of planning the area of
observation, deciding the processes of observation, selecting the settings where observation is
to be made and deciding the type or structure of the observation to be followed. For any
observation both the observation and recording are simultaneously done with the occurrence
88
of the events. Settings are places where people from different stratums of the society, namely
form different religious, racial or other groups live together or are involved in action and
interaction or in which the events take place. Examples of such settings are neighborhood,
social institutions, places of work, restaurants, drug stores, cultural centers, place of meetings,
club, church, etc. After the selection of the setting we decide about the specific types of
behaviors, interactions, or interrelations that have to be observed. An observer selects form
the numerous events going around himself (those which are of his interest, i.e., relevant to a
specific study). He then plans their systematic observation and categorical recording. How
observation is systematically planned and executed, can be clearly explained on the basis of
an example. Suppose we want to study the development of leadership personality among the
children as a function of environment. Steps of observation for this may be the following –
The first step would consist of selecting children from different family environments, namely
the democratic environment and the autocratic environment. The families that provide equal
opportunities of decision making and participation to its members and thus encourage,
leadership qualities would come under democratic type, whereas those which discourage such
free expression of opinions or participation would come under the second type. Family
environment is treated here as the factor or variable that determines leadership personality.
At the second stage one will have to decide about the settings where expression of leadership
qualities is possible. Settings for children’s activities are school playground, club, and
neighborhood. These are examples of natural settings. Another type of setting may be an
artificial one. For example, one may create a laboratory setting where observation can be done
under controlled conditions in order to establish the effect of one or more variables.
Then one will have to decide as to which activities of the children are to be taken as indicators
of or indices of leadership personalities. After deciding all of these one has to observe and
record the events systematically.
The above mentioned steps are taken to make an observation an objective one. This is because
a scientific observation has to be accurate, complete, reliable and valid. Only then verification
is possible. This makes the scientific observation different from chance observation or day-to-
day observation.
89
Such characteristics will be further examined and discussed separately under “types of
observation” to see how in different types of observations different steps are taken to fulfill
the above requirements. Whatever be the type of observation, the questions that generally
arise in observation are, what to observe, how to observe and how to record. Answers to these
vary with difference in condition are dealt with differently in different types of observations.
Types of observation method
On the basis of structure and process, the observation method can be divided into the
following types –
(i) Systematic, Structured and Controlled observation method.
(ii) Unstructured and uncontrolled observation method: This can further be subdivided on the
basis of degree of participation by the observer in the activities of the group being observed to
the following types:
(a) Non-participant observation
(b) Participant
(c) Quasi-participant.
Any particular observation method, when applied in the field does not remain to be exactly or
purely of any one type, rather it becomes a mixture of two or more types. The different types
of observations differ from each other in the degree of their structuredness, degree of control,
degree of participation and so also in the degree of their validity, reliability and objectivity.
The different types of observations will be discussed in terms of their structural differences,
content of observation, manner of recording, relationship between observer and observed, and
the methods of increasing the accuracy of observation.
Systematic, structured and controlled observation method
This method, most commonly, called structured observation method, is one in which the
observer knows in advance as to what aspects of behavior or activities of the people are
relevant for his research work and, therefore, he plans the manner of observing and recording
of the events much before the actual data collection begins. This means that in it the things to
be observed, the steps of observation, the conditions under which the observation is to be
made, the manner of recording and the items under which the observation is to be recorded,
etc. are prefixed. Even the detailed outline of possible outcomes, as in the case of
experimental research, can also be prepared. These make the observation highly structured
90
and systematized. It involves immediate recording of events as the observation is being made
rather than the preparation of reports of what has happened, based upon memory.
Structured observation method is used when the researcher is well acquainted with the
problem under investigation and also the population which is to be studied. Hence, one can
frame and also test more precise concepts and hypotheses in it. These criteria make the
structured observation method different from the unstructured observation method which is
used in exploratory research in which neither the problem nor the population is known.
In structured observation method one makes use of more carefully drawn schedules and
questionnaire. He may even use the audio-visual aids, scales, maps, tables, and diagrams, etc.
These help in categorical observing and recording of facts, in comparing the current studies
with the previous ones and also in verifying and correcting the previous measurements
through repeated studies.
Being standardized method the structured observation can be easily repeated and the
limitations of the study can be easily assessed.
In structured observation method it is easily possible to achieve structured precise, reliable
and valid data as it avoids inclusion of observer’s biases and his selective perception of the
events.
The degree of validity achieved in an observation method depends upon the degree of its
structuredness, the condition under which it is carried on, and also the efficiency of the
observer.
Structured observation is carried on either in the field situation under a natural condition or in
the laboratory setting under artificial but controlled conditions. Whenever possible,
observation should be done under natural conditions because that presents a more accurate
and true picture of the events before the observer. Naturally the conditions under which
observation is made, depends not only upon the setting of the observation and upon whether
the persons being observed are aware of the observation being made but also upon the
relationship between the observer and the observed. For example, suppose we are studying the
behavior and activities of the children during their play to study the development of their
personality, leadership qualities, manner of communication, pattern of play, etc. in such a
study so long the children being observed are unaware of the study going on their behavior
91
remains natural. Awareness of the observation will not only kill the neutrality of the situation
but may even stop the children from continuing their behavior or activities.
To find out the effect of different variables upon the children the investigator may create new
situations by introducing new stimuli without the knowledge of the observed. Thus, structured
observation can be made under controlled and uncontrolled natural situations.
For achieving reliable valid data through this method, the most important step would be to
develop most standardized techniques for observing and recording the events.
In structured observation method accuracy is achieved by using precise and exact measuring
instruments, which measure the variables. Here also use of two or more observer’s like that in
non-structured observation becomes most useful to maintain validity and reliability.
Sometimes awareness of the observation by the observed kills the natruality of the situation
and this leads to unreliable data. On the other hand, even if the observed persons are unaware,
reliable data cannot be achieved if the observational techniques are loose or unstandardized.
Distortion in observation may also occur due to the observer who is a mediator between the
actual situation and the data. He may affect the actual situation by interfering actively with the
occurring events. He may change the behavior of the persons being studied or may affect the
data because of lack of skill in him of observing and recording. Finally to say, as far as
possible, the observations should be made quantitative. For example, the observer might count
the number of times an event occurred or rate the event on a numerical scale, rather than
portray it in terms of verbal anecdotal description.
Simple, Unstructured, Uncontrolled Observation
Observation method is said to be simple and unstructured when things to be observed are not
prefixed. The observer is more or less in a free situation. He has to decide in the field at the
spot of observation about the things to be observed. This gives the observer greater
opportunities of decision-making. Opportunities of decision-making make the role of the
observer more significant. Hence the success of investigation depends upon the quality of the
observer, i.e., his capability of understanding the situation and observing. This capability
develops through training and experience. On this basis, it can be said that the observers are
not born but made.
92
It is considered as an uncontrolled method because there is no restriction over the observer as
to what to observe, how to observe, how to record, and so on. Also there is no control over the
observers. The observers are allowed to act or behave according to their own wishes. Controls
in this case refer to the standardization of observational techniques, if and when possible.
Most of our knowledge about social relationship is derived from such uncontrolled
observation. The greater problem in the case of this type of observation is that due to lack of
control over the observer there are chances of mixing reporting of observation with
interpretation of the observed facts or events. Due to lack of precise instruments of observing
and recording, which act as control over observation, the same events may be reported
differently by different observers. Simple, unstructured and uncontrolled observation may
further be sub-divided into three sub-types, on the basis of the degree of participation by the
observer and on the basis of the nature or relationship between observers and observed. These
are as follows:
(a) Non-participant observation method.
(b) Participant observation method.
(c) Quasi-participant observation method.
(A) Non-participant observation method

This type is applied when the observer collects facts in the most formal manner. Often the
observed persons are aware of the investigation going on. It is often used in small group
research situation. Complete non-participation becomes both difficult and uncomfortable for
both the observer and observed. Because the observer is always present but there is no
standard set of relationships or role patterns for the non-member observer to follow.
Sociologists and psychologists have adopted the method of limited participation. They have
maintained the position of most objective observer or interviewer and at the same time have
participated in some of the activities of the community.
(B) Participant observation method
Participant observation method is one in which the observer takes on the role of a member of
the group in disguise, participates in the different activities or functions of the group, and at
the same time observes the behavior or activities of the other members of the group, who are
completely unaware of the procedure. For this type of observation it is essential that the
93
investigator introduces himself to the group in such a disguised manner that he is accepted as
a member of the group and is at the same time able to fulfill his real purpose of observation.
Sociologists have, therefore, adopted different role patterns, as a laborer, as a porter, as a
criminal and so on. It is essential that the observer selects a role in the group which do not
disturb the usual pattern of behavior of the observers.
(C) Quasi-participant observation method
The difficulties involved in both non-participant and participant observation methods have
been examined in the previous pages. The ‘Quasi-Participant Method’ involves the merits of
both the previous types and excludes their defects. This type of observation method involves
limited amount of participation in the community activities. It is difficult for the observer and
the observed to maintain the quality of complete non-participation. In non-participation
observation method the observer is all the time present as an outsider and as non-participant in
the group activities. This makes both observed and the observer uncomfortable. It is also
impossible for the observer to be a genuine and full participant in all the ways. Full
participation may become sometimes essential or even unavoidable in participant observation
method. In quasi-participant observation, the investigator can play the roles of both observer
and participator in a simple and easy manner. It depends upon the initial introduction and
explanation of the observer’s entrance and presence in the community of investigation.
Check your progress
1. List and define the three major tools of data collection
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
----------
2. What is the advantage of the interviews over the questionnaire?
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
----------
3. List and describe the problems of observation.
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
----------
94
4. List and describe the steps involved in using the questionnaire as a tool for data
collection.
-----------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------
6.3 SUMMARY
The list we have seen is not complete, and the treatment is far from exhaustive. However, I
hope that you have got the necessary information as far as tools of data collection are
concerned. In order to have a better understanding of the tools, you should find opportunities
of practically using them for gathering data for research purpose.
1. a) content and document analysis – the use of published materials such as newspapers, textbooks,
novels, etc to obtain relevant research data.
b) Questionnaire is a list of open-end and closed terms written to secure relevant data.
c) Interview is a face-to-face approach to obtain research data.
d) Observation – is the act of observing/seeing the actions as it is face to face.
2. It is flexible as it is face to face.
3. Bias due to researcher’s perception
4. Item construction; pilot administration, receiving and tabulating; analyzing it.
95
UNIT SEVEN
PRESENTATION AND ANALYSIS OF DATA
Aims and Objectives

After going through this unit, you should
 list ways of presenting research data
 define the terms classification, tabulation, and approximation
 list and describe measures of central tendencies
7.1 INTRODUCTION
After research data is obtained/gathered, the next step is to present or organize the data into
some shape, and use statistical tools to analyze or interpret it.
7.2 STATISTICAL METHODS
Human mind has a peculiar weakness. It fails to assimilate a lot of ‘things’ or information at a
time. To remove this difficulty and to make “a large mass of data” intelligible, certain
statistical methods are adopted. Classification is such a method. Whenever an enquiry is
made, the enumerators collect a large mass of data. Mostly these data are of different types
and forms. All these data (in the forms of raw materials of the enquiry) are given to the
sponsoring authority for statistical treatment and for the preparation of the final report. In the
absence of their proper classification, they might not be used most profitably. Hence, there is
a necessity for proper classification of data.
Meaning and Definitions of ‘Classification’
Classification is a process of condensation. Statistical material-in its raw form-being
unprocessed, is mostly unwieldy and non-intelligible. For a proper utilization of the collected
data, various statistical methods and techniques are adopted – classification is such a method
of statistical representation. As the human mind fails to assimilate (or use properly) a large
mass of data so they are condensed through the adoption of the system of classification to be
put to a more systematic and proper use. Thus, classification is not only a tool of
representation, but in reality, is a vital part of statistical study.
96
“Classification is the process of arranging things – either actually or notionally – in groups or
classes according to their resemblances and affinities…”
Suppose, an enquiry is made to ascertain the extent of unemployment amongst the educated in
a particular locality. In all, say, there are 10,000 such educated unemployed-about whom the
enquiry is to be made. For the statistician or the sponsoring authority, dealing with the
enquiry, it would be next to impossible to reach a conclusion from the data “so collected”. To
make the things clear, these pieces of information are to be classified into various categories.
One classification may be dealing with fully unemployed and partially unemployed. Another
basis of classification may be the extent of education and the magnitude of unemployment. It
may also deal with age and the impact of unemployment, and so on. Thus, making a few types
of classifications of the “collected data,” the things may be made clear and intelligible. In the
absence of this simplification and condensation through the system of classification, the
enquiry may not yield the desirable result; nor the report may be given the final shape.
Main objects of classification
The main objects of the adoption of the system of classification are
1. To bring out the unity of attributes out of the diversities persistent in the collected
data.
2. To condense the universe and to make the things more and easily intelligible.
3. To make the study and comparisons easier.
4. To make proper use of the collected data.
5. To give prominence to the important information gathered, while dropping out the
unnecessary elements.
6. Putting the ‘collected material’ to statistical treatment.
7. To help the drafting of the required report.
8. It simplifies the complexities of the unwieldy data and makes it possible to draw
statistical inferences.
Sometimes, a difficulty arises about the practicability or otherwise about classification. Of
what nature classification should be, and what should be the form or intensity of
classification? In reality, there is no law which regulates or controls the type of classification
to be adopted. It differs from ‘man to man’ or from ‘purpose to purpose’. It may also differ
according to the accuracy required.
97
Types of classification
“The method of classification depends on the type of data which have been collected in an
investigation. For some type of data the classification. It requires special skill and it is only
after a lot we must remember that the nature and accuracy of the classification depend not
only on the type of classification followed, but also on the type of series taken into
consideration. Some of the types of classifications are:
(a) Classification according to attributes.
(b) Simple classification.
(c) Manifold classification.
(d) Classification according to class-intervals or ‘numerical classification’.
Classification according to attributes
Various enquire deal with qualitative phenomena which cannot be reduced to numerical
characteristics. This are to be classified according to attributes and is technically known as
‘classification according to attributes’. Suppose, in a particular enquiry, we have to deal with
blindness, baldness, deafness, etc. These are ‘qualities’ or ‘attributes’ which cannot be
directly ‘measured’. In this type of enquiry and their classification, the only procedure that
can be adopted is about their presence or otherwise. Naturally, the classification may be based
on the ‘units’ having these attributes or not-possessing these attributes.
So far as the study of statistics is concerned, and particularly dealing with the qualitative
phenomena, they are put under simple classification. Their use, however, is rare as compared
to other types of classification. So far the accuracy of this type of classification is concerned,
we may easily mention that they are not so accurate and reliable as other types of
classifications.
Simple classification or classification by dichotomy
Simple classification or classification by dichotomy is another method of classification. If the
‘object’ of classification is a simple one and can easily be put under two columns, it is known
as simple attributed. Suppose, the male population of a particular area is to be classified into
educated and non-educated only, then the type of classification adopted would be known as
simple classification. Under this type of classification, under normal circumstances, more than
two groups (sub-classes) are non-existent. If we require intensive classifications or
98
classifications under different sub-groups or at least into more than two sub-classes, it is
known as manifold classification.
Manifold classification
Suppose that we are required to classify the data obtained from an enquiry relate to the total
population – to be sub-divided into males, females and children; the figures may also relate to
education-literate and illiterate, etc. Similarity, their employment and civil conditions may be
enquired. When all these related data are to be classified; they are to be put under a system of
classification which is known as manifold classification. For, in the absence of manifold
classification, all the information’s may not be represented.
Classification according to class-intervals
Classification according to class-intervals or numerical classification is also known as
“frequency distribution”. Under this system, quantitative measurements of all the items under
enquiry are taken into consideration. It is a common experience that a large mass of individual
data and measurements may not be easily expressed unless some sort of classification is
made.
(B) TABULATION
Tabulation is a process of orderly arrangement of data into series of rows and columns where
they can be read in two dimensions. Tabulation is essential to represent a particular result of
enquiry or investigation. As has already been mentioned, a mass of data termed as ‘raw
material’ can express no meaning unless they are properly classified and tabulated. In reality,
tabulation is a process of representation of data.
Sometimes a question is asked, why tabulation is made? The reasons are manifold: (i) it
simplifies the comparison of data, (ii) it makes the study complete and accurate in every
respect, (iii) it makes the comparisons easier together with saving a log of time and energy in
following them; (iv) it makes the statistical treatment a possibility, and (v) finally, tabulation
makes the ‘affairs’ easily intelligible; which, in the absence of tabulation, becomes not only
difficult to understand, but almost impossible to give a practical shape.
Various forms of tabulation
Tables may be classified in various forms.
99
(A) Tables
General Tables Summary Tables
(B) Tables
Simple Manifold
(C) Tables
Of the first of the second of the treble of the quadruple manifold/

Order order (third) order (fourth) order complex
Table should never be too lengthy or ambiguous. It must make the study as clear and
‘objective’ as possible. One table should consist of only one title.
Example (Simple Classification)
Index Number of Industrial Production

(1950-51 = 100)
Group 1955-56 1960-61
General Index . . 139 194
Cotton Textiles . . 128 133
Iron & Steel . . 122 238
……. . . .. ..
Source: Third Five-Year Plan, p.39
Approximation
It is another concept about which the statistician and the sponsoring authority must have a
clear idea. The method of approximation not only simplifies the complexities of large number,
but it also makes the job of the statistician much easier than what would have been in the
absence of this system. With the help of the concept of approximation, the jobs of
classification, tabulation and calculation is very much simplified.
Methods of approximation
100
Some of the important methods of approximation are:
(1) By raising the actual figure to the next higher whole number:
e.g.
27,335 would become 28,000 approximately.
27,942 would become 28,000
(2) By approximating to the nearest whole number:
e.g.
(3) Dropping certain digits entirely to round the numbers:
e.g.
(4) By omitting the decimal points and taking the suitable whole number: e.g.,
9.21 would become 9 approximately
9.74 would become 10 etc.
There are certain other methods also that are taken into consideration for the process of
approximation. As in the case of accuracy, so in the case of approximation also, the method it
be followed is to be pre-decided. During the enquiry and when the enquiry is complete the
pre-decided method of approximation should be put to use.
The statistician is provided with a ‘wide discretion’ so far as the process of approximation is
concerned. It is on account of it that very little uniformity is to be found or seen in the method
of approximations adopted by different agencies or statisticians to achieve different types of
ends.
7.3.2 Measures of the central tendencies
Meaning, uses and objects of study
A statistical average is a single value which can be used to represent many divergent items for
some specific purpose.
In general, averages are said to be a single unit or number, representing a particular result.
The average is extracted out of a large, mass of varied data through the application of
statistical methods. It is a process of condensation very common used in the study of almost
101
all the branches of sciences and humanities. Averages are used to simplify the complexities of
large mass of unwieldy data.
The concept of averages is so important in the study of statistics that sometimes it is said that
“statistics is the science of averages”. It is also said that an average is a single, simple
expressing in which the net result of a complex group of large number is concentrated. An
average, thus, is a process of condensation. The average brushes off the irregularities of a
series, levels all differences of the individual items and presents complex and unwieldy data
by a “significant” number. It, thus, gives a bird’s-eye-view of an aggregate and can be
substituted for individual items in further calculations regarding the series.
A few generalizations
From the analysis of all the above, we may draw a few generalizations about the meaning and
objectives of an average:
(a) An average is a process of condensation. As it is not humanly possible to deal with the
large mass of data easily, so this method is used to reduce the figures into a smaller
one or in one single result.
(b) Averages are commonly termed as a ‘type’ because they represent the frequency
distributions in a simpler form.
(c) An average is also said to be a method of representation. The entire mass of data can
neither be successfully ‘represented’ nor can they become intelligible if placed ‘as
they are’. They must be processed to give a clear-cut and precise picture. The
‘Processed’ picture is an ‘average’ in most of the cases.
(d) An average is said to be the basis of comparisons with other series or groups of data.
In the absence of an average, all the ‘universe’ is to be compared. The adoption of the
use of averages as a tool of comparison, thus, is much more precise and simplified.
(e) Averages are commonly used for the further treatment of statistical derivatives and
series.
7.3 PRESENTATION OF DATA
102
7.3.1 Graphs and diagrams
On the completion of the process of classification and tabulation, sometimes it becomes
essential to arrange for their proper representation. Out of the methods of statistical
representation, an important one is the representation of facts and figures by the use of
diagrams and graphs. Hence, the significance of the study of diagrammatic and graphical
representations in the study of statistics.
(A) Diagrammatic Representation
It is commonly accepted in principle that the human mind is so constituted that it fails to
remember a lot of information’s, or some information for a very long period. On the contrary,
while studying statistical methods and other allied subjects, we have to deal with large mass
of data; or with a large ‘universe’. So, there is always a need of representing these data in a
simplified manner – which is easy both to understand and remember for a longer time without
any difficulty. Diagrammatic and graphical representations satisfy this basic requirement.
Forms of diagrams in common use

There are several types of diagrams which are used to represent statistical data and
infroamtions. In this connection, we must remember that all the types are not given equal
importance; neither that is possible. Some forms of diagrams are very frequently used, while
there are a few other types which are rarely drawn or put to use. In some cases, they are not
drawn because they are not considered as desirable or useful for comparisons. Some other
types of diagrams are not commonly used, for they are difficult to be drawn. However, there
are a few types of diagrams which are very commonly used.
UNIT EIGHT
THE RESEARCH REPORT
AIMS AND OBJECTIVES
103
 define the term ‘report’
 describe its function
 distinguish between ‘research’ and ‘report’
8.1 INTRODUCTION
The last phase of the journey in survey is the writing of report. After the collected data has
been analyzed and interpreted and various generalizations have been drawn, the report has to
be prepared. The report of a survey is thus, the statement that contains in brief the procedure
adopted and the findings arrived at by the investigator of a problem. A report is not a
complete description of what has been done during the period of survey. It is only a brief
statement of most significant facts that are necessary for understanding the generalizations
drawn by the investigator.
8.2.1 Purpose of a Report
The purpose of a report is to convey to the interested persons the whole result of the study in
sufficient detail and so arranged as to enable each reader to comprehend the data and so
determine for himself the validity of conclusions. The report may be meant for the people in
general when the investigation has not been carried at the instance of any third party. At such
times the purpose of the report is to invite others to study the generalizations and offer their
own comments regarding it. It is only in this way that a perfection can be achieved in any
branch of study. Research is essentially a co-operative venture and it is essential that every
investigator should know what others have found about the phenomena under study. The
purpose of report is thus dissipation of knowledge, broadcasting of generalizations so as to
ensure their widest use.
Report also creates ground from hypothesis and leads to further research on the same or allied
problems. It may be remembered that the basis of the development of physical sciences is this
co-ordination of work. Various scientists have worked on small aspects or a problem, and
their findings have been consolidated into a theory. All this co-ordination is not possible
unless a complete report is available about the procedure adopted and the finding made in
different researches.
So are we have described the purpose of a report when investigation has been carried on
purely for the purpose of research. The survey is sometimes conducted at the instance of third
party that has some stake in the problem. Such surveys deal with market research, polls
104
conducted by magazines or surveys conducted at the instance of the government. In these
cases the report is not meant for general public and need not contain too many procedural
details. The person or authority under whose instance survey has been conducted may be
interested exclusively in the finding of the survey.
8.2.2 Contents of report
Although no hard and fast rules can be laid down, yet a report should generally contain the
following information: -
(1) Purpose of study
The report should start with the problem undertaken in the investigation. It should give the
background of the problem, need for research and the hypothesis formed for this purpose. If
the study has been conducted at the instance of some third party the scope of investigation
should be mentioned. The importance of the problem for the advancement of knowledge or
removal of some evil may also be explained. If the problem has some historical background,
the same must be mentioned giving at the same time various other studies undertaken and the
findings arrived at.
(2) Methodology
The second part of the report should give the methodology adopted for the purpose. It should
give the various aspects of the problem that had to be studied for valid generalization about
the phenomena. It should give the exact meaning of various units of measurement or terms
used for the purpose. If sample study was made the method of drawing out the sample should
be mentioned giving reasons if any for the selection of particular sample. Other procedural
things like definition of universe, preparation of source list etc. should also be given so that
necessary idea may be formed about the validity of the sample.
Various methods are used for social investigation, e.g., case history, questionnaire, schedule,
qualitative interview, etc. The report should clearly mention the particular method adopted
giving reason why it has been selected and how far the information based upon the method
could be considered a reliable. The report should also contain information regarding the
method of collecting data. It should mention clearly how far the data was primarily collected
for the problem under study and how far the secondary data already in existence was used. In
the latter case proof of reliability of the secondary data should also be given if necessary.
105
Generally, questionnaire or schedules are used for the purpose of study. A copy of all such
questionnaire, forms, statements, etc. must be attached to the report, so that other persons
going through the report may form an idea regarding the nature of validity and reliability as
also the meaning of various terms used in the report:
(3) Organization of survey
The third part of the report should contain the organization of survey. It should give the
following information. If field survey was organized, what was the method of selection of
field workers, their training etc.? What was the actual procedure or interview of observation?
What was the arrangement for the inspection of the work of field workers, how were the
schedules or questionnaires collected and what methods were adopted for their tabulation and
classification?
(4) Analysis and inferences
This is the most important portion of the report. The collected data is analyzed and various
inferences are drawn from it. Inferences are based upon logical as well as statistical reasoning.
The report should contain not only the generalization of inference but also the basis upon
which the inferences have been drawn. All sorts of proofs, numerical and logical, must be
given in support of any theory that has been advanced. Various suspicions and doubts that are
likely must be mentioned and a suitable explanation given for them. It must be remembered
that it is the report that will go to the persons and not the actual data. The writer of the report
should take this weakness into consideration and should avoid all chances of misconception
about the theory formulated or inferences drawn from the research.
(5) Appendices
The report generally contains the statements or original documents on the basis of which the
generalizations have been formed. The questionnaire or other forms of methods and
statements are included in the appendix. If help has been taken from any report written,
document or any other paper it is good to give its relevant portions in the appendix. In short
all those facts that need a special elaboration, but cannot be given in the main report for the
fear of killing the lucidity and sequence should be given in the appendix.
1. Foot-notes
106
Foot-notes serve the same purpose as appendices. They are generally used as reference guide
or short explanations to the points under discussion. They are often used to throw light upon
the topic under opposition to the one being discussed. Foot-notes should be given whenever
necessary but their too frequent use would make the report very complex and lucidity would
be destroyed.
2. Suggestions
Besides the inferences, sometimes suggestions are also required from the investigator.
Suggestions are invariably given when the investigation is not for purely research purpose and
has been set up by a third party either government or non-government body to probe into
some social problem needs a cure. Even in case of purely academic investigations sometimes
suggestions are needed for reformative or curative steps. These suggestions if at all needed are
to be given at the end of the report. Suggestions must be practicable and based on logical
reasoning. They must be exhaustive and should take into consideration all the difficulties that
may arise in implementing them.
8.2.3 Problems of communication
The major problem in writing a report is as to how the matter is to be communicated, what
form of language and expression is to be used? There is a general complaint regarding all
such reports that they are essentially too much technical. Really speaking every research is a
technical matter and is not capable of being followed by every layman unless he has some
prior basis. The investigator while writing the report has to assume that the reader does
possess certain basic knowledge about the problem, various technical terms used and
expressions made.
If he tries to simplify it too much, the report may become other too vast or too simple and
unable to carry the required impression. Every branch of knowledge has not only its technical
terms but also its own way of expression. Unless it is followed the report would appear simply
childish and would lose its research flavor. Therefore, it is necessary that in order to
understand the report the reader himself has to rise higher, rather than let the researcher stoop
too low for him.
Before writing a report, therefore, the researcher has to take into consideration two factors.
Firstly, for whom the report is meant, and secondly, what is his level of knowledge. All
reports are not meant for research scholars. There are some surveys that are conducted for
107
general public or for such persons or institutions which have no technical knowledge. Thus,
for example, the report submitted must be non-technical in nature; otherwise, them would not
be able to make any use of it.
The writer of the report has, thus not only to look to the subject matter of the report but also
the fact as to the level of knowledge of the persons for whom it is meant. The report must
avoid ambiguity at all cost. It may be technical, but it cannot afford to be incomplete or
ambiguous, or give sweeping remarks without giving valid reasons for them.
The drafting of report is an art. It not only requires a grasp over the subject, but also a
command over the language. The writer should have a control over himself to avoid and
tendency towards exaggerating facts. The report should be lucid, and clear-cut so that it may
not be difficult to follow the sequence of the points mentioned therein. It should not be merely
a collection of stray facts and inferences, but those facts and inferences should be interwoven
into a theory or hypothesis, thus leading to come concrete results.
8.2.4 A sample research report format

CHAPTER ONE
I. Introduction
1.1 Statement of the Problem
1.2 Objectives of the Study
108
1.3 Scope of the Study
1.4 Limitations of the Study
1.5 Significance of the Study
CHAPTER TWO
II. Literature Review
CHAPTER THREE
III. Methodology
Sampling
Subjects
Tools of data collection
Procedures of the study
(What steps did you go through as a whole?)
CHAPTER FOUR
IV. Data Presentation, Analysis and Interpretation
CHAPTER FIVE
V. Conclusions and Recommendations
Appendixes
(Attach sample tools of data collection)
Bibliography
Check your progress
1. What is a repot?
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
---------------
2. What are the functions of a repot?
------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------
---
3. What is the difference between research and report?
109
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
---------------
8.3 SUMMARY
A report we have said is a presentation of the major outcomes of findings of a study. It serves
the purpose of communicating findings to pertinent stakeholders. It also lays the foundation
for further research or study.
1. It is a means of reporting the major findings of a study.

2. It conveys the result of study in a comprehensive manner; it also invites others to
comment on the findings of a study, it lays foundation for further research.
3. Research is the investigation itself. It involves identifying and defining problems,
selecting relevant data, analyzing it, drawing conclusions, etc among other things.
Report, on the other hand, is the presentation of the findings of the study orally or in
written form.
110

Business Research Methods

Uploaded by

Copyright:

Available Formats

Business Research Methods

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Business Research Methods

Uploaded by

Copyright:

Available Formats

UNIT ONE

THE CONCEPT OF RESEARCH

AIMS AND OBJECTIVES

At the end of this unit, you should be able to

1.7 ANSWERS TO CHECK YOUR PROGRESS EXERCISES

Characteristic Quantitative Research Qualitative Research

Approach Deductive Inductive

Theory testing, prediction, Describing multiple realities,

Isolates variables, uses large Examines full context,

Developed before study is Begins with an initial idea

Mainly statistical, Mainly interpretive,

Check your progress exercise 8

Aims and objectives

2370 7483 3408 2762 3563 1089

(ii) Quota sampling

5.4 ANSWERS TO CHECK YOUR PROGRESS EXERCISES

1. Sampling consists of obtaining information from a portion of a larger group.

Questions asked by interviewer No. 2 –

Advantages of Open-end types

1. It involves difficulties of coding, classifying, grouping and comparing of the

If data is to be collected through interview or through distribution method, the field

(A) Non-participant observation method

Aims and Objectives

General Tables Summary Tables

Of the first of the second of the treble of the quadruple manifold/

Index Number of Industrial Production

7.3 PRESENTATION OF DATA

Forms of diagrams in common use

8.2.4 A sample research report format

1. It is a means of reporting the major findings of a study.

You might also like