1

·INTRODUCTION ·To:-
BuSiNESs ·REsEAiif'.•,
'1
I.----'----"---:...__. J
ME~:
t' ·./ .,
..
HODS

• Meaning of Research

• Nature And Characteristies of Research

• Objectives of Research

• 'Jlfeed And Importance of Business Research

• Types of Research
• Concepts of Research
• Stages in Research Process
• Characteristics of Good Research
• Meaning And Nature of Hypothesis
• Significance of Hypothesis
• Types o/Hypothesis
• Sources Of Hypothesis
• Meaning And Definitions of Research Design
·• Need And Importance of Research Design
• Steps in Research Design
• Essentials 'o f A Good Research Design
• f\reas/ Scope of Research Design
• Types of Research Design
 Business Research Methods (S. Y.B.M.S .: SEM-IV)
2
Meaning of Sample And Sampling
•
Methods of Sampling
Probability Methods Vs. Non-probability Methods
•

Research is a systematic investigation to search for new facts in any .

branch of knowledge. Research helps to a_rrive at n~w conclusions.
It enables to find solutions to certain problems.
Research is often referred to as 'scientific inquiry' into a specific
problem or situation. This is because; the search for facts needs to
be undertaken systematically and not arbitrarily. The systematic
approach to research enables lhe research to search for facts in a
rational manner arid to arrive at logical conclusions, whereas, the
arbitrary approach attempts to find solutions to problems based on
one's belief and imagin•ation. · ·
William C. Emory in the book 'Business Research Methods' defines
research is any organised inquiry de.signed and carried out to
11

provide information for solving a problem." .

The Webster's Dictionary states is a careful cri_tica_l inquiry
or examination in seeking facts o.r principles; diligent investigation
in order to ascertain something." ·
V. Clover and H. Balsley define uresearch is the process of ·
systematically obtainitt;g accurate answers to significant and
pertinent questions by the use of scientific method for gathering
and interpreting information."
James Black and Dean Champion· state r,cientific research consists
II

of obtaining information through empirical observation that can

be used for the systematic development of logically related
propositions attempting to establish causal relations among
variables."
 duction to Business Research Methods 3
. Jn trO ,
pattline V. Young defin~s "social research is a scientific undertaking
ivhich, by means of logical and systematic techniques, aims to:
, Discover n~w facts or verify and test old facts,
Analyse their sequences, interrelatio~ships and causal
• . .

• "I?evelop ~e:v
scien_tific tools, concepts and theories, which
would facilitate reliable and valid study of human beh_aviour."

1. Scientific Method:
Research uses scientific method to find facts· or to provide
solutions to specific problems. The researcher needs to follow
. a systematic procedure to conduct.research. There is a set of
procedures that have been tested over a period of time and are .
thus suitable to use in research. This means each step in the
research procedure must follow the other.
Scientific research in any field .of knowledge cannot be
. conducted in a haphazard manner. Scientific research cannot
be merely based on one's beliefs and imagination. To g~t the
best possible research results, the researcher needs to adopt the
scientific method of inquiry or investigation. ·

2. Objective and
. . Logical: ·
The scientific research is objective and logical in nature.
Research is based on valid procedures and principles.
There is a need to collect relevant, accurate and objective data
to investigate into the research problem. Researchers need to
make every possible effort to avoid bias in data collection. After
collection of objective data, the researcher n·e eds to
systematically process the data, analyse and interpret it, and
arrive at logical conclusions. Wherever required, the researcher
needs to systematically verify the findings and conclusions. ·

1,
\
4 Business Research Methods (S. Y.B.M.S.: SEM-IV)
Introduction to Business Res,
3. Applied and Basic Research:
' research findings 1:
The research can be broadly classified into two_ broad groups•· -_whole population.
,• Applied Research _6. Controlled Nature
In real life experi1
• Basic R~search
outcome. A single
· l' d ch is designed to solve practical problems of the When similar eve
Appdte researld rather th.an to acquire knowledge for nature of factors,
mo ern wor , - . . .
knowledge's sa~e. The goal of applied research 1s to rm:rrove - while others are b
the human condition. It is generally used to solve a particular very easy to contr
problem. For instance, a social resea~ch can be conducted to conducted. ·
study the problem of unemploym~nt m rural areas, and based However, it is <
on the research findings; appropriate measurE;S can be taken ·experiments in tb
by Goyernment authorities to reduce the proble~ of · limited extent la
unemployment. sciences as well.
Basic or fundamental research is drivel). by a scientist's curiosity 7. Development ot
or interest in a scientific question. The main goal of basic
A systematic re
research is to expand man's knowledge. There is no obvious
theories. Such p
commerciai · value to the discoveries that result from basic _
organisations to
research. For instance, basic research can be undertaken to study
better way. The
the ~rigin of the universe. ·Basic research lays down the rese.arch may e1
foundation for the applied scien~e that follows. which have not
4. Empirical Nature of Re.search: 8. Multipurpose .
Research can be based on direct experience ·or observation by Research is a n
the researcher. Empidcal research is undertaken to study certain _facts or verify o
situations or events based · on experiments, observation and but also may li
surveys. In _empirical_ research, the researcher develops a relationship b,
hypothesis and then collects data to prove !t _o r to disprove it. scientific tools
5. Generalisation: reliable and va
Research findings can be ap'p lied to larger population. A 9. Manipulation
researcher can conduct a research ·on a sample of respondents The researchE
that represent the universe. The sample selection must be done manipulation
sys_tematically so that it properly represents the whole done with a<
popu-lation _or the universe. The research findings based on generality .
sample population can then be generalized and applied to the Encyclopedi
whole universe. Therefore, generalisation takes place when manipulatioi
generalizing

J_
4
Y.B.M.S.: SEM-lll) Introduction to Business Research Methods 5
' research findings based on sample. responses are applied to
vo broad groups: whole population. · . _

.6. Controlled Nature of Basic Research:

In real life experience there are many factors that affect an ·
outcome. A single event is often the result of several factors.
When similar event-is tested in research, due to the' broader
il problems of the
nature of .factors, some factors are taken as contro_lleg. factors
knowledge for
. while others are tested for possible,effe,ct. In pure sciences it is
rch is to improve
very easy to control certain factors because lab experiments are
solve a particular . conqucted. · · ·
be conducted to ·
. areas, and based However, it is difficult to perform controlled laboratory
ires can be taken experiments in the case of social science research, although to a ·
the' proble~ of .· limited extent laboratory experiments are possible in social
sciences as well. ·
:ientist' s curiosity 7. Development of Principles an4 Theories:
ain goal of basic A systematic research helps to develop new P!iriciples and
ere is no obvious theories. Such principles and theories can be useful to several
result from basic organisations to. manage and deal with people and things in a
ldertaken to study I better way. The general Jaws or theories developed through ,.
h lays down the I·
_rese.arch may enable us to make reliable predictions of events,
ws.
which have not yet occurred.. . .
8. Multipurpose Activity:
·or observation by
Research is a multipurpose activity. It helps to discover new .
;en to study certain
. facts qr verify old facts. It helps not only to predict future events,
, observation and
rcher develops a but also may help to control such events. It establishes causal
-or to disprove it.. relationship between variables. It also helps to develop new
scientific tools, concepts and theories.~which would facilitate
reliable and valid study of human behaviour and other aspects.
·r population~ A 9. Manipulation of Concepts: ·
le of respondents
ion must be done The researcher tries to manipulate things, or concepts. The
manipulation· or purposeful control of things, or concepts is
ents the whole
done with a definite purpose so as to arrive at statements of
idings _based ·on
generality. D. Slesinger and M. Stephenson in 'The
d applied to the
Encyclopedia of Social Sciences' define _research as "the
kes place when
manipulation of things, concepts or symbols for the purpose of
generalizing to extend, correct or verify knowledge, whether

. I ,, ·
_ v
6 Business Researcl1 Methods (S. Y.B.M.S.: SEM-l\1)
Introduction to Business Resea1
that knowledge a ids iu cons tn,ctio n of theonJ ·Or in the practice
ofan art." OBJECTr
For instance, a researcher may manipulate the environment hl Researchers undertake ri
a workplace such as l.igh~g, or layo':1t _or seating arrangement
the important purposes c
to find out its impact on the productivity of the employees.
as follows:
10. Quantitative and Qualitative Research: · 1. To find solutions tc
Research undertaken to measure quantity or amount is called Research can be uni
as quantitative research. For instance, research undertaken to problems. For insta
find out the number of unemployed graduates or the number to find solution tc
of unemployed in general. On the other hand, research, which products in the r
is undertaken to find out the quality of a particular situation or undertake researcr
phenomenon, is called as qualitative research. For instance, a . poor results.
research undertaken to find out the reasons as to why
employees remain absent from'work, or why people behave in A government' or~
certain manner. The motivational research is an important type the problem of wa
of qualitative research. · Qualitative research is especially to ascertain the i!
important in the behavioural sciences where the main aim is to particular city, ani
find out the underlying motives of human behaviour. enables-to find a
which in turn hel
11. Research - a Process: various organisat
Research is a systematic process. It involves a number of steps. _
The main steps include: 2. To verify and-te~
Research may be
• Formulating the research problem or situation theories. Such ve
to improve the k
• Development of a hypothesis
events. This is '
• Preparing the research design sufficient or relE
and therefore, th
• Determining san:iple design· (in case of social research) can be made in
• Collecting of data ,,.,
3. To obtain infm
• Analysis of data Research is unc
· be easily 9btain
• Testing of hypothesis an institution c

• Generalizations and interpretation For instance,

understand th
• Preparation of report or thesis
undertake pre
• Follow-up of report
itroduction to Business Research Methods 7

{esearchers undertake research with definite objectives. Some of

he important purposes or objectives of research are briefly stated
15 follows:_

1. To find solutions to problems:

Research can be undertaken to find ·solutions to solve specific
problems. For instance, an organisation may initiate research
to find solution to the problem of declining sales of'their
products· in the i:narket. . An educational ·ins,titution can
undertake research to find out the causes of low attendance or
· poor results.

A government' organisation may undertake research to solve

the problem of water scarcity in a particular area or district or
to ascertain the impact of slums on t}:l.e quality of life in a
particular city, and such other research activities. The research
enables-to fin~ appropriate solutions to specific problems,
which in turn helps to improve the quality of performance in
various organisations or institutions. ·
2. To verify and-test existing laws or theories: .
·Research may be undert_aken to verify and test existing laws or
theories. Such verification and testing of existing theories helps
to improve the knowledge and ability to handle situations and
events. This is true when the existing theories .may p.ot be
sufficient ,or relevant to handle certain situations and events,
'and therefore, through research, improvements or modifications ·
can be made in the existing laws or theories.
3. To obtain information:
Research is undertaken to obtai,n information, which may not
be easily 9btained during the ordinary course of functioning of
an institution or an organisation.
For instance, marketing research ' may be undertaken to
understand the- chi;mges in consumer behaviour. A firm may
undertake product research to bring about improvement or

I
l J,
 8 Business Research Methods (S. Y.B.M s .SE .
. .. M,1v
modification in the existing product on the basis of feectb )
. d fr O m consumers,
obtame · dealers and others. a~l Introduction to Business Resea1

To extend knowledge: For example, a rese

4. between advertising
hers undertake research to extend the . advertising causes (r
Researc . . ( h h existi ,
1 dge in physical sciences sue as p ysics, ch . ng
know e . 1 . ·. 1 . el1)1st~, 8. To develop new too
h atics etc ) as wel as m socia . sciences (like s . -1,
mat em , ., ) Th k oc1olo
ement psychology, etc .. ·. e nowledg _gy, Research helps to d1
mahnagced by' undertaking research 1n generale can be study of an unkr
en an . . _ · and b exploratory researc]
fundamental research m particular. · Y
such phenomenon.
5. To establish generalizations and general laws: .
9 To develop newp1
Research can be und~rtaken t~ establish generalizations and A systematic rese,
general _la~s in a particular society. In oth~r words, statements theories. Such prir
O'f generality can be stated through research. For inst
. . . ance, organisations to rr
various laws, prmciples· and mode~s ha_ve been developed better way. The g
through research. The AJDA (a~tenhon, mterest, desire, and research may enal
action) model, the law of _demand and supply, the law of which have not yi
gravitation, etc., have been developed through observation,
10. ·To develop innm
j
experimentation, and other methods of research.
Research may be
6. · To predict events: the welfare of m,
Research may be undertaken to predict future course of events: organisation to d
For instance, research may be undertaken to find out the impact
• New and irr
of gro.wing unemployment of educated youth on the social life
of the society in future. The findings of such research would • lmproved o
not only indicate the possible impact, but also would make the
• Better teclu
concerned authorities to take appropriate measures to reduce
unemployment to reduce the growth of population, and to • New sourc
·. overcome the negative consequences; as and when they take
place.

7. To analyse inter-relationships:
Research may be undertaken to analyse inter-relationships
between variables, so as to derive causal explanations, which
in .turn would enable to have a better understanding of our Business organisa
society and the universe in which we live ..Researchers often · research. The resE
devel(?p hypotheses and · test them to analyse the causal products, improvE
relationship between variables~ reduce cos'ts. Resei
'
price, promotion,]
· soon . .
 Introduction to Business Research Methods- 9
For example, a researcher may study causal relationship
between advertising and sales, i.e. to find out whether or not
advertising causes (results) higher sales.
8. To develop new tools and concepts :
Research helps to develop ·new tools and ·concepts for a better
study of an unknown phenomenon.· For this purpose,
exploratory research is undertaken to achieve new insights into
such phenomenon.

?· To develop new principles and theories

A systematic research helps to develop new principles and .
theories. Such principles and theories can be useful to seven~J
organisations to manage and deal with people and things in a
better way. The general laws or theories developed through ·
research may enable us to make reliable p;edictions of events, ·
which have not yet occurred. · ··
10. ·To develop innovative ideas:
Research may be undertaken to generate innovative ideas for
the welfare of .mankind. For instance, research may enable an
organisation to develop innovative ideas in respect of :

• New and improved products

• Improved organisation structure

Better technology

•- New sources of raw materials, etc.

Business organisations can gain significantly with the help of

· research. The research and development helps to develop new
products, improve· the· design and qualit of existing ones,- a_nd to
1
reduce costs. Research helps to solve busmess problems relatmg to
price, promotion, phy~ical distribution, packaging, positioning, and
· soon.
10 Business Research Methods (S. Y .B.M.S.: SEM~I\l)
The need and importance of research in l;:msiness is stated as follows~
1. Product Development:
Through marketing research, a business firm can identify the
customer requirements, and therefore, it is possible to design
new models or to modify existing products to satisfy the
customers.
For instance, through marketing research, a car manufacturing
company may find out that the customers look for features,
after-sale-sale service, re-sale valuer right price, fuel efficiency,
and so on. Accordi.n gly, the product will be designed and
marketed.
Also, with the help of research and development, the car
company would make every possible effort to develop the car
with the. best possible features at the right prices depending
upon the target customers.
2. Reduction in'Costs:
Research in business h~lps a firm to reduce costs. The research
may indicate the areas where high costs are involved. It may
be possible for the firm to cut down \he costs in certain areas
Therefore, cost-reduction will improve the overall efficiency o
the organisation, i.e., the organisation will be in a position t<
achieve higher returns at a lower cost.

3. Marketing Mi~ Decisions:

Research in business enables a firm to arrive at soun
marketing-mix decisions ·w ith r~spect _to product, pric1
promotion, and physical distribution.

(a) Product Decisions: As mentioned earlier, marketir

research enables a firm to design the right product to satis
customer requirements.

(b) Pricing Decisions: Marketing research enables a firm

analyse coippetitors~ pricing, customers' pr i
sensitiveness, etc .., and accordingly, the firm can fix 1
right prices. For instance, if the customers for a particu
product are not price sensitive, the firm m~y charg
Introduction to Business Research Methods 11
higher price, especially when it enjoys a ·good image in the
market. , _

(c) Promotion-mix Decisions: Research in busin:ess may ·

enable·a firm to adopt effective promotion-inix(publicity,
advertising, sales promotion, etc.). If promotion-mix
research is not conducted, the firm may give more
emphasis on certain elements of promotion-mix, thereby,
.ignoring the other elements. ·

(d) Place Decisions: Research in business may als.o enable the

firm to take appropriate place decisions, with reference to
area of distribution, channel seJection, incentives to channel.
intermediaries, etc. For instance, the firm may increase
marketing efforts in those areas where it gets low sales, or
if th~ existing channels are not .effective, it may sel~ct
alternative channels, etc.
4. . Customer Relationships: ·
, ,

Research in business may help the firm to develop good

relations with its customers, especially,,the priority customers.
Research can help the firm to collect valuable inputs about the
priority ·customers. Based on the inputs, the firm can adopt
various customer relationship techniques such _as package of
loyalty incentives, one-to-orte marketing, effective after-sale-
servke, incentives for custo:rner suggestions, etc.
5. - Dealer Relationships:
Nowaday~, dealer relation~hips are vital for the survival and
success of a business organisation. Marketing research enables
a business firm to main good relationship with dealers. Through
research, the firm can get feedback from the dealers regarding
their expectations. Accordingly, the firm may take appropriate
decisions relating to deal~r' s compensation and incentives_so
as to maintain good relationship.
6. · Corporate Image:
Nowadays, firms need to build, maintain, and enhance
corporate image_in the minds of stakeholders - customers,
dealers, employees, and others. For instanc~, consumer research
may enable the firm to obtain feedback on customer
12 . Business Research Methods (S. Y.B.M.S .: SEM~J~
Introduction to Busi
requirements and expectations, and accordingly the compalt
would make the right efforts to saHsfy customer and devela~ Proper rese,
sources of h
good image in.the minds of customers. Also dealer ~esearch
vyorking ca1
enables the company to obtaµ1 feedback on dealer req1;1-rrements
and expectations, and thus the cqmpany would ma~e efforts tG 10. MarketExF
enhance dealer satisfaction, which in turn would develop gooq - With the hE
image of the firm in the minds of dealers. identify the
marketer ec
7. , Competitive Advantage: beunderta1
Through research, a company can take proactive decisions, such such as p
as introduction of new ·models, .introduce price changes, salesmansl
undertake innovative promotional schemes, and so on. The
proactive decisions can confer competitive advantage to the
firm. ··

Also, through research, the firm can find out the strengths of
the competitors, which make certain buyers to prefer The research ac
competitors' brands. Based on competitors' research, the firm
of the main ty1
can improve its market offering vis-a-vis that of the competitors, 1. . Pure Rese
and thereby, enhance its competitive advantage.
. '
· Basic rese,
8. Human Resource Plans and Policies : is undert2
Resear.ch may be underta:ken to frame ~ffective HR plans and mamgoa
policies in respect of : instance,
of the un
• Recruitment and Selection the appli1
\

• .Training to employees In basic

obvious •
• Performance Appraisal
. .
basic res1
carried 1
• Promotion and Transfers-
governrr
• Compensation Plans, etc .
Basic re
9. Financial Management : · world.11
~esearch _may _be u~dertaken _for . efficient manag~ment of observe1
fmance. Fmanc1al management covers two broad areas : scientifi
explora1
• Sources of Funds researd

• Application of Funds

I
I
Introduction to Business Research ·Methods 13
Proper research may enable a company.to determine the best
sourc~s of fu~ds, and ~o ensure proper application of funds for
workmg capital and fixed capital.

10. Market Expansion:

With the help of marketing research inputs, the company can
identify the markets that have good potential. Accordingly, the
!Tiarketer can en~er in new markets. Entry in new markets can
be undertaken with the help of appr9priate marketing activities,
such as pr_omotional activities - publicity, advertising,
salesmanship, sales-promotion, etc. . ·

The research activity can be classified into different categories.- Some

of the main types of research are as follows:

1. Pure Research:
Basic research is also called as pure or fundamental research. It
is undertaken to develop a theory or a body of knowledge. The
main goal of basic research is to expand man's knowledge. For
instance, basic research can be undertaken to study the o~igin
of the universe. Basic research lays down the foundation for
the applied ·science that follows.
\ . .

In basic research, there is no commercial angle. There is no

obvious commercial value to the discoveries that result from
basic research. In almost all countries, pure research is mainly
carried out by universities and institutes financed by the
government.
Basic research ·advances fundamental knowledge about the
world. It focuses _o n refuting or supporting theories that explain
observed phenomena. Pure research is the source of most new
scientific ideas and ways of thinking about the world. It can be
exploratory, descriptive, or explanatory; however, explan~tory
research is the most common.
I

!'

I
14 Business Research Methods (S. Y.B.M.S.: SEA,Hv -
I

I
Basic research generates new ideas, principles, and theor· · l In_tro_duction to Busin1
which m~y not be immediately utilized. How_e ver, the n~s,
3~ Empirical Rei
. theories or ideas form the basis of progress and developrne\iv · Empirical Re
1
in different fields. For instance, tod~y' s computers could n~ experimentat
1
exist without . research
. in pure mathematics conducted . over a of gaining ]
century ago, for which there was no known practical applicatio11 observation
· at __th¢ ti1:1~· B_asic .re~earch rarely' help_s pra~titi~ners directly _ conducted to
with their everyday concerns; nevertheless, 1t ·stimulates new The word en
ways of thinking that have the potential to revolutionize an.d observation,
I I. dramatically improve how practitioner~ deal with a problell\ method is th
\
in the future. is based one
refers to the
2. . Applied Research: - observation
Applied research is a scientific study designed to solve practical
problems, tathei: than merely acquiring knowledge. Applied Empirical e
research is used to find solutions to everyday problems, and experiences:
develop innovative technologies. · - ·' ' ' Through qt
empirical q
Psychologists working in human factors or organisations ·answerable
/ undertake applied research. The goal of applied research is to · by field a1
I
I
- improve the human condition. It is generally used to handle a resear!2hers
particular situation or .to solve a particular problem. · analysis to ,
la:t,oratory 1
For instance, a ,social research scholar may undertake research .
to find out job .satisfacfi~n of college teachers ·affiliated to · 4. · Scientific a
University of Mumbai. The research findings can be applied -_Scientific n
by colleges or the University to improve the level of job · the researc
satisfaction of the teachers. Also, a research can be undertaken the applic,
to study the job satisfaction of nurses in Govt. hospitals vis-a.- informatio
vis private hospitals in the dty of Mumbai. The -research the prope1
findings can b~ used by Govt. hospitals and private hospitals process cc
to improve the jo? satisfaction levels of th~ nurses. follows:

The applied research can be undertaken by business 1. !dent:

organisations as well. For instance, a business organisation may
<

n. - Revie
undertake applied research to find out the causes of poor sales
_of a par_ticular brand in the market. The research findings can 111. Form
be used to design necessary meas tires to overcome the probleJ!l . 1v. Prep,
of poor sales.
v. Desif

·l
I J1
In_tro.duction to Business Research Methods
3; Empirical Research: 15

Empi~ical Re_search can be defined as "research based on

experimentation or observat · · ,, E .. h.
. . . ton . mp1nca1researc 1s a way
of gaim~g knowledge by means of direCt and indirect
observation or experience or experiment. Such research is
conducted to test a hypoth~sis. ·

The wor~ empirical means information gained by experience,

observation, or experiment. The central theme in scientific
method is that all evidence must be empirical \Vhich means it
.is based o.n evidence. In scientific method the word "empirical"
refers to the use of working hypothesis that can be tested using
. observation and experiment. ·

Empirical evidence (based on one's direct observations or ·

' \
experiences) ~an be analyzed quantitatively or qualitatiVely . .
Through quantifying the evidence, a researcher can. answer
empirical questions, which should be clearly defined and
answerable with the evidence collected. Research design varies
by field and b}' the question being investigated. Many
researchers combine qualitative and quantitative forms of
· analysis to answer questions better which cannot be studied in
laboratory settings, particularly in the social sciences. . ·
4. Scientific and Social Science Research:
. ScientifiC research is a Systematic process undertaken to study
the research problem and to arrive at conclusions. It re~es _o_n
the application of the scientific method. ~t _provides sc1entif1c
information and theories for the explanation of t}:le nature and
th rties of the world around us. The scientific research
e prope
process cons1s. ts of a sequence of steps, which .are. stated as .
follows:
1. Identifying and ~election of research problem
11. Re~iew of literature

111. Formulating hypothesis

iv. Preparing research design

V. Designing a questionnaire
 16 .
f !Qili&Qj
Business Research Methods (S. Y.B.J:,A.S,: SEM,Jv
I
:--
Introduction to Business R
v1. Preparing a sample design
The important fe,

I
I
vi_i. , Collecting data
viii. Processing data
ix. Analyzing and Interpreting the data
•
•
It is necessai
. meaning in 1
It helps tot
perspective.
x. Testing of hypothesis
x1. Preparing research report
• It relies exce

• In conductir
manipulate
xii. Follow up of the, report
Social Science research deals with the behaviour of people iii 6. , Explora~ory Res
their different roles, such as consumers, consultants, learners, Exploratory resE
trainers, employees, producers, parents, artists, etc. The social the nature or cai
science research provides workable solutions to economic a~d the causes of t
social problem.5. The importance of social science research can researcher. The
be stated as follows: are as follows:
• Social science research helps to modify social .behaviow
for social welfare
• To define t

• To develo1
• Social science research enables the organizations to develop ·
I
j·
appropriate methodology to study, analyse, and to take • To identif3
To isolate
suitable action to overcome the problems relating to human
resources.
• among th(

• Social science research contributes to improve standards • To establi:

of social development. Descriptive R,
7.
• Social science research is a tool for social planning and Descriptive rei
control. the populatic
describe the " 1
5. Historical Research: It does not <

- Historical research involves interpreting past events to predict Therefore, dei

future ones . It ~ncludes analyzing events that occurred in the provide a sys1
remote or recent past. Ac.c ording to Wilfiam Wiersma, historical ' as possible. It
research can be defined as, ua process of critical inquiry into or frequency o
. past events, in order to produce an accurate description and such as dete1
interpretation of those events". central tender
research metl
m eth od.
 Introduction to Business Research Methods 17

The important features of historical research are as follows:.

it is necessary to define the situations of the ·past and its
• . meaning in the light of the present problem.
It helps to understand th~ issues better and in·the right
data • ' perspective.

• It relies excessively on secondary sources o( data.

• In conducting historical research, the researcher can neither
manipulate nor control any of the variables. ·

the behaviour of people in 6. . Explora~ory Research:

mers, consultants, learners, Exploratory research is conducted to explore information about
rents, artists, etc. The social the nature or causes of research problem. It is conducted when
solutions to economic and the causes of the research problem are not known to ' the
f social science research.can researcher. The purposes of conducting exploratory research
are as follows:
o modify social behaviour e · To.define the problein more clearly
• To develop hypothesis
te organizations to develop
• To identify alternate courses of action
J.dy, analyse, and to- take
·oblems relating to human • To isolate key variables ·and to develop relationships
anion·g the variables for further examination
~s to improve standards • To establish priorities for further research

7. Descriptive Research:
:or social planning and Descriptive res·e arch or statistical research pro_v ide_s data about
the population or universe being studied._But it can only . .
describe the "who, what, when, where and how" of a situation.
It does not describe· what caused a particular situation.
past events to predict Therefore, descriptive research is used when the objective is to
tts that occurred in the provide a systematic description that is as factual and accurate.
:1m Wiersma, historical as possible. It provides the number of times something occurs,
'i critical inquiry into orfrequency of occurrence. It lends itself to statistical calculations
urate description and such as determining the average number of occurrences or
central tendencies. The two most commonly types of descriptive·
research methods include observation method and the survey
method.
r
Business Research Methods (S. Y.B.M.S.: SEM-Iv)
18 Introduction to Busin
One of its major limitations is that it •cannot
• help detenninI
what qmses a specific behaviour, motivatioI_l or occurrence.)
cannot establish a causeaeffect relationship between variable:
Variables:
The descriptive research provides facts o_f a parti~ular e_vent 01 1.
. situation. It gives a description of the state of affanrs, as 1t exists A variable is an er
of a particular event or situation. The researcher has no control can vary can be cc
over the situation or event. He can only r~port what has j:ncome are all ex,
happened or what is happening. For instance, a researcher may take different val1
report 9n the absenteeism in a particular organisation or several different times.
organisations in a particular industry or even in different For the purpose o
industries. Details can be given for a: particular period - in terms value to the varial
of the total number of absentees, the number of days which the as an attribute.
employees have remained absent, the department or _section in
· which there is more absenteeisrp., the class of employees who Depending on thE
remain absent the most, and such other details. The research as:
will not indicate why the employees remained absent.
Continuous
Another example of descriptive research is in respect of the • within a givi
study to understand the general pattern of buying behaviour income, tern
of consumers during a partjcular season or in general
Non-contim
throughout the year - the quantity of purchases, the frequency •
Discrete va1
of purchases, the type of purchases, and so on. This research
-categories a~
will not provide answers as to why (reasons) the customers
on number <
buy certain quantity or quality or the reasons -for low/ high
frequency of purchases, and so on. Non-contini
variables i.E
8. Casual Research: · gender hast
_ Casual research investigates cause-effect relationship between status h?ts t,
two or more variables. The objectives of casual research are:
To understand the cause-effect relationship between two The variables ,
• or more variables. · . - relationship:
Independe
To focus on those variables or elements having greater •
• manipulate
- / positive effect. - ·
To.eliminate certain variables or·elements having negative
w~ys, such

• e~ct. · •
Dependen
e?(pected to
be termed
• To develop action plans . andmeasUJ

1 .
r
,ds (S. Y.B.M.S.: SEM-n;) Introduction to Business Research Methods 19
annot help determine
·ation or occurrence. It
hip between variables.
1. Variables:
of a particular event or
ite of affairs, as it exists A variable is an entity that takes on different values. Anything that
searcher has no control can vary can be considered as a variable. For example, age, weipht, ·
only r~port what has income are all examples of variables. This is because these entities
tance, a researcher may take different values for different people or for the same person at
•organisation or several different times.
y or even in different I

:ticular period - in terms For the purpose of data analysis, the researcher assigns a numerical
Lmber of days which the value to the variable. A specific value assigned to avariable is called
iepartment or _section in as an attribute. · ·
class of employees who Depending on the attributes assigned, the variables can be classified
1er details. The research as:
·emained absent.·
Continuous Variables: These variables take on values either
arch is in respect of the . • within a given range or from an infinite set. For example, age,
~rn of buying behaviour income, temperature, etc.
season or in general
1urchases, the frequency , Non-continuous Variables: These variables are also called as
nd so on. This research Discrete variables. These variables produce data that fit into
reasons) the customers .categories as only certain values are possible. For example, data
, reasons for low/ high on number of children, number of students, etc. ·
. Non-continuous or discrete variables include Dichotomous
variables i.e. variables having only two values. For example,
t relationship between gender has two values: male and female. Similarly; employment
·casual research are: status h~s two values: employed and unemployed.

tionship between two The variables are . also classified on the basis of cause-effect
relationship:

nents having greater Independent Variable (IV): IV is that variable which is

manipulated by the researcher. IV can be termed in _different
w~ys, such as: predictor, antecedent and manipulated variable.
1ents having negative
• Dependent Variable (DV): DV is that variable which is
expected to be affected by the manipulation of the IV. DV can
be termed in different ways, such as: criterion, consequence
and measured outcome·.
 20 Business Researc~ Methods (S. ¥.B.M.S.: SEM~IV)
Introduction to Business ReseQ
For instance, in case of the hypothesis: z~ Qualitative and Qua
"
"The New Education Policy will result in,increased employment (a) Qualitative Researcl
opportunities to students." Qualitative research
disciplines, especi,
the IV is New Education.Policy ·!esearchers aim to gc
the DV is employment opportunities to students. behavior and the r
qualitative method i
Mode;ating V ari~ble (MV): MV is that variable which is known making, not just wr
•• as second IV and is believed to have a significant effect on the focused samples are
originally stated IV-DV relationship. For instance, a resea
to why employees •
For instance,
behave in certain r
"The New Education Policy will res.u lt in increased employment important type of q
opportunities to students, especially the B.Com gra,duates." · especially important
aim is to find out thE
the MV is the B.Com graduates.
• is
Extraneous Variable (EV): EV that variable th~t is directly
(b) · Quantitative Resea
Quantitative resear,
not related to the purpose of the study but may affect the DV.
The effect noticed on DV due to EV is called as 'experimental · numerical data that
error'. A good research is designed in such a way that the effect methods. The objec
on DV is a result of the IV and not due to EV. . ., and employ mathe11
pertaining to pher
For instance, central to quantit
fundamental conrn
".For the students with higher academic grades, the New
mathematical expn
Education . Policy will result in increased employment
opportunities." The res~archer ask:
sample of numeric2
the EV is the students with higher academic grades. analyzes the data v1
·• Intervening Variable (IVV): IVV is that variable that affects may prpvide unbic
DV. t11Fough the cause of IV and MV. 1arger population. <
broad questions ,
For instance, researcher looks fc
themes .and patten
"The Ne~ ~ducation Policy will result in increased employment
i opportu:111hes ~o _students, especially the B.Com graduates by More generally, qt
conducting trammg programmes." · sciences such as ps

I the IVV is the training programmes.

community health
political science, an
1rch.Methods (S.Y.B.M.S.: SEM~IV) 21
Introduction to Business Research Methods
othesis: z. Qualitative and Quantitative Research:
,..
l result in'increased employment (a) Qualitative Research:
Qualitative research is a method of inquiry employed in many
disciplines, especiaIIy in the social sciences. Qualitatiye
Jolicy
·researchers aim to gather an in-depth understanding of human
iportunities to students. behavior and the reasons that govern such behavior. The
qualitative method investigates the why and how of decision
:v is that variable which is known. making, not just what, where, and when. Hence, smaIIer but
0 have a significant effect on the focused samples are mqre often used than large samples.
onship. I
For instance, a research undertaken to find out the reasons as
to why employees remain absent from work or why people
I behave in certain manner. The motivational research is an
ill result in increased employment
peciaily the B.Com gra,duates." · important type of qualitative research. Qualitative research is
especiaIIy important in the-behavioural sciences where the m~in
lduates. aim is to find out the underlying motives of human behaviour.
:vis that variable th?t is directly · (b) · Quantitative Research:
the study but may affect the DV. Quantitative research is explaining phenomena by coIIecting
to EV is called as 'experimental numerical data that are analysed using mathematicaIIy based
gned in such a way that the effect methods. The objective of quantitative research is to develop
:i not due to EV. ' and employ mathematical models, theories and/ or hypotheses
pertaining to phenomena. The process of measurement is
central to quantitative research because it 'provides the
,er academic grades, the New fundamental connection between empirical observation and
It in increased employment mathematical expression of quantitative relationships.
The res~archer asks· a specific, narrow question and collects a
1 higher academic grades. sample of numerical data from the respondents. The researcher
analyzes the data with the help ·of statistics. The analysed data
vV is that variable that affects
'MV. may prpvide unbiased result that can be generalized to some
'larger population. Qualitative research, on the other hand, asks
broad questions and collects data from participants. The
researcher looks for themes arid describes the information -in
~sult in increased employment themes .and patterns exclusive to that set of participants.
ia1Iy the B.Com graduates by
~-"
' More generally, quantitative research is widely used in social
sciences such as psychology, economics, sociology, marketing,
·ammes. community health, health & human development, gender and
political science, and less frequently in anthropology and history.
 Business Research Methods (S. Y .B.~.S.: SEM-IV)
22 Introduction to Business Researcl,
I\
(d) Understandable- v
with main questioi
Each research problem is unique {e) Manageable-fits tl
. and requires af special emphas·lS
\\ 1.
.
and approach. One way to face the uniqueness ·o every problem is
to tailor the research work according to needs of each problem. (f)
· restrictions.
Ethical - does nc

I The research process consists of a sequence of steps that have to be

followed while undertaking a research project: 2.
respondents and t
Review of Literature·
' .
1. Identifying ahd Selection ~f Research Problem: ., The researcher should
relating to the problem
Identification or formulation of research problem ~s the first such as journals~ book
and most important step of research process . The problem matter. Particularly,
formulation is like an identification of a destination before similar research studiE
undertaking a journey. It is often said that a well defined , review of literature v
problem is half solved. · problem.
Before formulating the resean;:h problem, the researcher musl 3. · Formulation of Hypo
· always think - what the problem is, why it is a problem, and The researcher should
for whom the problem is. · is a tentative ~ssump
Precise definition. of the problem would help the researcher to consequences. The hy]
collect the relevant data for solving the research problem. The of insight and knowl1
researcher would be in a better position to arrive at suitable may prove to be eithE
conclusions based on the analysis of data collection, and For instance, the rese,
accordingly solutions can be designed to solve the problem or may hypothesize tha
to deal with a particular situation.
• Product price is
The researcher needs to list the broad area or topic of the
problem; divide the broad area into sub-areas, i.e., narrow down • Middlemen are ·
the broad area into sub areas, and then formulate the problem.
• The product is l:
The follo_wing are some of the essentials of a good problem
If the test of the firs
formulat10n: ·· ·
. conclusion is that pl
(a) Researchable - the problem c~n be investigated through the researcher cone
the collection and analysis of data. · proceeds to test thE
guides the research 1
.., (b) Interesting- the problem keeps the researcher involved in
it.throughout the research process. . . 4. Research Design:
The researcher mus
(c) Purposeful - the findings of which must be useful to solve and systematic plan
the problem or situation. .
½ethods (S. Y.B.M.S .: SEM-Jv) 23
Jntroduction to Business Research Methods
(d) .Understandable - well formulated and logically structured
with main questions and sub-questions.

equires a special emphasis (e) Manageable - fits the researcher's skills, resources, and time
[ti.eness of every problem is · restrictions.
needs of each problem. (f) Ethical - does not embarrass or har~ the researcher,
~nee of step~ that have to be respondents a~d the society.
project: 2. Review of 1:iterature:
:arch Problem: _, The researcher should undertake extensive literature survey
relating to the problem. He may consider various publications,
~search problem is the first
such as journals~ books, research reports, .and other published
arch process . The problem
matter. Particularly, the researcher should go through the
ion of a destination before
similar research studies that were conducted previously. Such
~n said that a well defined
. review of literature would provide a good insight , into the
problem.· ·
roblem, the researcher must'
3 · Formulation of Hypothesis:
is, why it is a problem, and ·
The researcher should formulate the hypothesis. The hypothesis
is a tentative ~ssumption made to test its logical or empirical
rould help the researcher to consequences. The hypothesis should be formulated on the basis
the research problem. The of insight and knowledge about the problem. ·The hypothesis
Jsition to arrive at suitable may prove to be either wrong or right. .
is of data collection, and
ed to solve the problem or For instance, the researcher studying the declining sales trend,
may hypothesize that sales are declining because:

"Dad area or topic of the • Product price is too high, and/or

b-areas, i.e., narrow down • Middlemen are not getting adequate margin, and/ or
n formulate the problem.
• The product is becomi~g obsolete.
tials of a good problem
If the test of the first hypothesis le~ds to its acceptance, one
. conclusion is that product's price is too high. If it is rejected,
>e investigated through the researcher concludes that the price is not too high and
proceeds to test the other hypothesis. Thus, the hypothesis
guides the research effort and suggests what data are required.
'researcher involved in
4. Research Design:
nust be useful to solve The researcher m ust prepare a research design .· It is a logical
and systematic plan prepared for conducting a research stu dy.
 . , R nrc) Methods (S. Y.B.M.S.: SEM-1\!J /utroduction to Business Research Methods 25
Business ese 1
24
. t for collection, measurement and 6_ SampHng Design:
lt can be called as a blue pnn h desian provides guidelines to Generally, it is not possible to collect data from each ·member
analysis of data. The rese_arc_ eriod within which research is .of the universe or population und er study due to limitations of
th
researcher regarding e tJI~~pta to be collected, the area and time, effort, and money. Therefore, the researcher needs to select
to be conducted, th e ty:h~ch :he data is to be collected, the a sample of respondents that represent the universe.
the respondents fromll t' and data analysis, and so on.
techniques of data co ec wn . Sampling,design is a plan to select the appropriate sample to
The research design must include the following aspects: collect the right data so as to achieve the research objectives. A
. sample represents those individuals chosen from the population
• A clear statement of the research problem. of interest as subjects in an experiment or to be respondents to
a survey.
The sources of data collection.
There are certain essentials of a good sampling design:
·The ti Ill e period of research study•
• The sample must be representative of the universe.
• The area or place where research is to be conducted. •
• The sample, must be decided depending on the research
The resources required to conduct the research.
• objectives .
The tech11iques of data collection.
/
• • The researcher must use proper method of sampling to
/
• The techniques of data analysis. select a sample.

The 1111iverse of research. • The sample size must be economical or cost-effective.

•
The sample size of respondents, if any.
• • There must be proper sampling plan .

• The aims or objectives of the research.

• ,The sample size .s hould be suitable to collect relevant data.
5. Designing the Questionnaire: • The san;zple unit must be selected properly.
lf the researcher cannot solve the problem with the help of
secondary data, observa tion and experimentation, then he • The sampling frame must be up-to-date to. c;ollect
should make efforts to collect the primary data from the field information of members of the universe.
for wh ich he requires a questionnaire. While designing a • The sample size must be flexible and not rigid.
qu es tionnaire, the following points must be kept in mind:
7. Collection of Data:
Type of information requ ired.
Problem solving is essentially a process of collecting
• Type of technique to be used for conducting the research, information, The data can be collected from various sources -
1.e.'. whe th er telephon·e interview, personal interview or primary and secondary. While collecting data care should be
mad. · taken of:

• Proper wording and proper sequence of questions . • - Information is.up-to date and free from bias.

• It is objective an d relevant to the needs of problem.

 . R ch Met/wds (S. Y.B.M.S.: SEM.J"
Business esear v) Introduction to Business Research Methods 27
26
It is complete in all respects. such as ~hi-square test, F-test, etc., have been developed for
• such testing. The testing of hypothesis will result in either
s. Processing of Data: .
. most! available m a raw orm and
f accepting it or rejecting it.
The collected data 15 Ysed Processing of data involves, 11. Preparation of Research Report:
therefore, it needs to be proces . . .
The research findings and conclusions are presented with the
. . I h 1 to weed out unwanted and irrelevan1
(a) ~::!'.n1f ~ls~ h:;: to check errors and omissions in data
help of research report. The research report is divided into three
parts:
collection.
(a) Preliminary Contents - whichinclµdes title of the report,
(b) Coding - It involves assigning codes (num~rica], letter of authorization, letter of transmittal, and table of
alphabetical, alpha-numeri~al) to the categories _or contents (index).
responses. It is required especially when the sall)ple size
is large and when there is large number of responses. (b) Main Body - which includes introduction, methodology,
findings, limitations if any, conclusioris and
(c) Classification. It refers to grouping of data unde~ differen1 recommendations.
categories or classes such as age, gender, educatwn, area,
/ etc. It facilitates tabulation of data. (c) Concluding Part - which includes appendix and
bibliography.
(d) Tabulation - It i;volves transferring of classified data in a
tabular form. Tabulation of data facilitates analysis and 12. FoJlow-up of Report:
interpretation of data. The researcher should s~bmit the report to concerned
9, Data Analysis and Data Interpretation : aµthorities. For instance, a doctorate thesis is to be submitted
to g_uide for approval and then to the concerned university. A
Organisation of data is generally followed by its analysis a~d
business research report is to be submitted to the concerned
interpretation. The purpose of analyzing data is to establish a
management of business organisation. The researcher should
relation between the information and problem. Analysis
fi~d out whether his report is accepted. If accepted, whether
reviews the data in the light of hypothesis or research problem.
It is difficult to demarcate a line between analysis and ~Is recommendations are accepted and implemented. If
interpretation. They quite often overlap and so it is difficult to implemented, whether the recommendations are successful in
find out the end of analysis and the beginning of interpretation. solving the problem.

Interpretation refers to analysis and generalization of results.

It enables t~ cle~r -the meaning and implication of study. Ji c;HARACTERISTICS'. OF GOOD RESEARCH
enables -to lmk fmdmgs with research problem and arrive al
logical-conclusions.
. The qualities of good research involve the following characteristics:
10. Hyp.othesisTesting:.
1. Scientific Method:
bAfter
. analysis
. . and. interpre ta t·!On o f d ata, the researcher must The researcher needs to .follow a systematic procedure to
em ahpos1t10n to test the hypothesis (if framed earlier) The conduct good research. There is a set of procedures that have
researc er should ·find out w h eth er or notthe
· research findings
·
been tested over a period of time and are thus suitable to use in
suppoFt the hypothesis or prove to be contrary. Various tests, research. This means a good -research follows a series of steps
1n · Introduction to Busi~ess Research Methods 27
,. a :ra\\r
)Ing of dat f ~tlt1. . such as ~hi:-square test, F-test, etc., have been developed for
. . a ll)v i\~! such t~sh~g. The testing of hypothesis will result in either
:ln.tect a . ()h_,f>. accepting 1t or rejecting it.
nct . -~
ld Otn.i . ltt~l 11. Preparation of Research Report:
ss101'\.s . evah,
. Q\ rl ~,
. 'lq~ The research findings and conclusions are presented with the ,
codes ( help of research report. The research report is divided into three
parts: ·
) th n lllll
e cat ~tic
\Then th egories ii\ (a) Prel~minary Contents - which inclµdes title of the report,
b e.salll l o, letter of authorization, letter of transmittal, and table of
er of res ,p e si~.
. Ponses '<{ contents (index).
1
f data unct ·
er
l d er, ed ucat·
diffe
re~
(b) Main Body - which includes introduction, methodology,
. l():n_ findings, limitations if any, conclusions and
. , are~ recommendations.

of classified d t . . (c) Concluding Part - which includes appendix and

·1· a a 111
c1 1tates analy . 1 bibliography. · · ,
s1s_ano
12. Follow-up of Report:
The researcher should s~bmit the report to concerned
aµthorities. For instance, a doctorate thesis is to be submitted '
~d by its analysis ano to guide for approval and then to the concerned university. A
; data is to. establishi business research report is to be submitted to the concerned
l problem. Analysii management of business organis'ation. ·The researc~ershould
:; or research problem find out whether his report is accepted. If accepted, whether
tween analysis ano his recommendations are accepted and implemented. If
and so it is difficultto implemented, whether the recommendations ·are successful in
n ing of interpretation solving the problem.

1eraliz~tion of re 5ul~
tplication of study.~
)roblem and arrive al
The qualities of good research involve the following characteristics:
1. Scientific Method:
. rnusl The ~esearcher needs to follow a systematic procedure to
a, the res~arc~er '[ht
conduct good research. There is _a set of proced~res that ha:7e
1.{ framed ear her) .. m.
f dl11o· been tested over a period of time and are thus smtable to use m
)t th e research l1'l ~esll1 research. This means a good -research follows a series of steps
:on trary . Various ·
 Business Research Methods (5. Y.B.M.S.: SEM-Iv) 29
28 Introduction to Bu siness Research Methods
. ch procedure. To get the best possible rese,, h
m the resear d h . ifi' "'C · the problem. If a pilot study indicates a f<)vourabl~ response,
esearcher needs to a opt t e sCient c methoct
results, the r . . of the researcher may go ahead with the research design .
inquiry or investigat10n.
6. Good Research Design:
2. Objective and Logica~: . dI . I. tu I . · ·tY· A go'od
A research design is a blueprint of the research act1v1
A good research is ob1ective an og1ca 111 na re. t 1s based
research must have a good research design. A good research
'd rocedures and principles. Researchers need to lllak
on va IJ p ible effort to av·o1'd b'ias 111. da ta co IIection.
. Aftee
design includes:
every pas S h r
·co II ec•t·10 n of obi·ective
· data, the researc
. er needs 1o • Focus on research objectives
systematically process the data, analyse a~d 111terpret it, and
. t logt'cal conclusions. Wherever reqmred, the researcher • Clear statement of the problem
arnvea if th f' d' d
needs to systematically ver y e m mgs an cone us10ns.
I ·
• Time period of the research study, etc.

3. Empirical: 7. Selection of Right Techniques:

A good research is empirical. Research cahnEbe b~~ed on direct The researcher must select th e right techniques for data
expen·ence or observation by .the .researc
. . mpmca 1research collection and data analysis for conducting good res earch
• dertaken to study certam situations or events based on activity.
1s un . In .. al
experiments, observatioi:i and surveys. empmc research,.
the researcher develops a hypothesis and then collects data to Some of the methods of data collection are:
prove it or to disprove it. • Survey/ Interview
4. Generalisation: • Observation
A good research should be a~le :o apply researc~ findings to
larger population. Research f111dmgs can be applied to larger • Experimentation
population. A researcher can conduct a research on a sample Some of the methods of data analysis are:
of respondents that represent the universe. The sample selection
must be done systematically so that it properly represents the • Methods of central tendency
whole population or.the universe. The research findings based
• Time series
on sample population can then be generalized and applied to
the whole universe. Therefore, generalisation takes place when • Correlation, etc.
research findings based on sample responses are applied to
whole population. 8. Ease in Implementation:
A good research has ease in implementation of its research
5. Pilot Study: · find ings . As far as possible, the resea rcher shou ld avo id
A good research involves conducting a pilot study. Pilot study complicated procedures and techniques which are difficult to
is a research activity undertaken on a small segment of the adopt/ follow. The research work should not overburden the
research universe. The pilot study is conducted to · find oat research staff, as the quality of research work may get adversely
whether or not the researcher would be in a position to collect affected . Also, sufficient time must be given to the staff for
the relevant data from the sources or areas so selected,' and collecting an d analyzing the data.
whether or not the collected data would be relevant to solve
 Business Research Methods (S. Y.B.M.S.: SEM-Jv 31
30 . - ~- .·- ... ,,,, • • ~ -. ) Introduction to Business Research Methods
_MEANING AND NATURE OF HYPOTHESIS) 3, Specific:
A hypothesis should be specific and explain the expected
relationship between the variables and the conditions under
The formulation of hypothesis or hypotheses is an important ste
which these relationships would hold valid. For example, the
in the formulation of research problem. The hypothesis is a tentat' P
. . 1"d. ive concept of socio-economic class must be specific in terms of
roposition formu 1ate d to d e termine its va 1 1ty. The hypothe .
P . I indexes such as income, occupation, education, etc. Such specific
may prove to be correct or incorrect. n any event, it leads to
clarity ensures the validity of the results because more specific ·
empirical test. Whatever the outcome, the hypothesis is a questi:
the statement, less the possibility that the outcome is of mere
put in such a way that an answer of some _k ind can be forthcomin n
accident or chance. ·
It is an example of the organized skepticism of science, the refus!i
to accept any statement without empirical verification. 4. . Related to Available Techniques:
A hypothesis should be related to available techniques. The
Webster's Dictionary d_efines hypo!hesis as "an unproved theory,
researcher who does not know what techniques are available
proposition, supposition, etc., tentatively accepted to explain certain to test the hypothesis would find difficult to frame useful
facts or to provide a basis for further investigation, argument, etc." questions. If hypothesis is not related to available ~echniques
William Goode and Paul Hatt define hypothesis as "a proposition then it would not be possible to research. Therefore; the
which can be put to a test to determine its validity." ' researcher must make sure that relevant techniques are available
for testing the hypothesis.
G.A: Lundberg defines hypothesis as "a tentative generalization,
the validity of which remains to be tested." 5. Theoretical Relevance:
A hypothesis should be related to a body of theory or some

/
The following are the characteristics of a good hypothesis : theoretical orientation. This condition relatei;-to the theoretic
rationale of the hypothesis, i.e., what will be the theoretical gains
1. Empirically Testable:
of testing the hypothesis? If the hypothesis is related to some
A hypothesis should be empirically testable. It should be so
theory, research will help to correct, refute or validate the theory.
stated that it is possible .to deduce logically certain inferences
from it, which in turn can be tested by field observation. In 6 Consistency: 0

other words, it should be possible to collect empirical evidences ' Hypotheses should be l~gically consistent. Two or more
to test the hypothesis. Statements like "Intelligent parents hypotheses logically derived from the same theory must not
produce smart children" "Intelligent students score high in be mutuaUy contradictory.
exams" are commonplace generalizations and cannot be tested,
as they merely express sentiments and their concepts are vague. 7. Objectivity: A hypothesis should be free from bias of the
researcher. The .value-judgement, feelings and emotions of the
2. Conceptual Clarity:
researcher have no place in framing the hypothesis. Th e
The hypothesis must be conceptually clear. The concepts used researcher should take every possible care to see that the
the hypothe~is should be clearly defined. An ambiguous hypothesis is free from any bias or personal prejudices ..
J'P 0thesis, which is poor~y defined, cannot be tested, as there
ould be no _stan~ard basis for knowing what observable facts 8. Consider all Pertinent Areas of Problem:
would constitute its test. . The researcher while framing a good hypothesis must consider
all the pertinent areas or aspects of the problem.

I _
L
_
- 32
Business Research Methods (5. Y.B.M.S.: SEfv!.J
. .• - . VJ Introduction to Business Research Methods
33

SIGNIFICANCE OF HYPOTHESIS 4, Suggests the Type of Research: ·

The hypothesis suggests which type of research is likely to be
Hypothesis plays an important role in _the r_esearch process. It guides most appropriate in solving the research problem.
the research activity in the proper dlfectwn. In all analytical nd Technique of Analysis:
experimental studies, hypothesis should
.. be framed in order to liIVe~ · 5. Toe hypothesis indicates the most appropriate technique of
a proper direction to the research acllv1ty. analysis of data, which in turn helps the researcher to draw
The role or significance of hypothesis in a research activity is proper conclusions from the research findings.
ex plained as follows: Development of Theory:
6.
1. . Provides Definite Focus: . Hypothesis contributes to the development of theory. It makes
A hypothesis provides~ definite foc~s on the r:search prob!elli. an attempt to link theory and investigation. A hypothesis can
be deduced from theory. When it is tested with the help of
R. E. Chaddock states, _A hypothe~1s gives point to the _inquiry,
and if founded on suff1c1ent prev10us knowledge, guides the research studies, and if found valid, it forms part of theory.
lines of i11vestigatio11." Cohen and Nagel statement that "w
cannot take a single step forward in any inquiry . without e
hypothesis, may well be an a_ccm'ate statement of the value 0;
' TYPES OF . HYPOTHESIS
hypothesis in scientific investigation.'' G.A. Lundberg remarks
"The only difference between gathering data without a Hypotheses are classified in several ways. Some of the important
hypothesis and gathering them with one is that in the latter classifications of hypotheses are as follows:
case, we deliberately recognize the limitations of our senses
On the Basis of Level of Abstraction:
and attempt to reduce their fallibility by limiting our field of I.
investigating so as to prevent greater concentration of attention Goode arid Hatt have identified three broad levels of abstraction
011 particular aspects, which past experience leads us to believe
typified by hypothesis:
are insignificant for our purpose." · (a) Common Sense Hypothesis: At the lowest level of
abstraction, are some hypotheses that state the existence
2. Specifies Sources of Data:
of empirical uniformities. These hypotheses frequently,
A hypothesis specifies the sources _of data. Therefore, the
though not always, represent the scientific examination of
researcher would consider only the relevant sources of dat~ '
common-sense propositions. For instance, "Well-paid
which in turn would speed up the research activity.
:m_rloy~es are more motivated than Jess-paid employees."
3. Determines Data Needs: D1soplined workers are more loyal to their organisation."
R. E. Chaddock states, that without hypothesis much useless The hypothesis of this type invites scientific verification
data may be collected in the hope that nothing essential will bi of "common-sense propositions." The common-sense
omitted or important data may be omitted which could hai,e hypotheses play an important role in the growth of a
been easily included if the purpose of inquiry had been more particular discipline or science. .
clearly defined.
(b) Complex Hypotheses: At a relatively higher level of
abstraction are some hypotheses that are concerned with ·
complex ideal types, These hypotheses aim at testing the
34 Business Research Met'1ods (S. Y.B.M.S.:,SEM-J½ 35
J,itroduction to Business Research Methods . .
existence of logically derived relationships betwee 111
to that of urban areas." "The quality of education
empirical uniformities. For exam pl~, human ec~logy eari: privately managed schools is far better than that of
described a large number of empmcal uruform1ties. Lani government schools."
values, industrial concentratwns, types of businesses
mental disorders, and m~y othe_r phenom~na appeared t; (b) Relational Hypotheses: These hypotheses ~escribe"the
show unquestionable unif~rm1t1es m d1stnbut10n. Further relationship between two variables. The relat10ns~1p may
~tudy and logical analysis of these and other related be positive or negative correlation or causal relat10nshtp.
findings led to the formulat10n of various hypotheses For example: "Educated people spend more on clothing
concerning the way m which these were related. as compared to uneducated people." "Literate couples have
fewer children as compared to illiterate couples."
(c) Analytical Hypotheses: At the highest level of abstraction
are some hypotheses that are concerned with the relation (c) Causal Hypotheses: These hypotheses state that the
existence of, or a change in, one variable leads to an effect
of analytic variables. These hypotheses occur at a !eve) of
on another variable. The first variable is called the
abstraction beyond that of ideal types. Whereas the
independent variable, and the second one is called the
hypotheses of empirical uniformities lead_ to the
dependent variable. For example "Increase in the female
observation of simple difference, and those dealmg with
literacy results in lower birth-rate."
.ideal types lead to specific coincidences of observations,
the study of analytic variables requires the formulation of 111. On the Basis of Nature of Hypotheses:
a r~lationship between .changes in one property and c
The hypotheses can also be classified as:
changes in another. For example, the study of human
fertility might show empirical regularities by wealth, (a) Working Hypotheses: The working hypotheses provide the
education, region, size of community, and religion. If this basis for further investigation. While plarming a research
were then raised to the level of ideal type formulation, one s_tudy, hypotheses are formed, which may not be very
result might be the hypothesis that there are two high- specific initially. They are subject to modification as the
fertility population segments in India, i.e., "Low-income investigation proceeds.
illiterate couples in rural India" and Low-income illiterate
couples in urban India." At a still higher level of (b) Null Hypotheses: These are hypothetical st~tements
abstraction, the effects of region and income might be held denying what is explicitly indicated in working
constant. This would allow a better measurement of the , hypotheses. For example, even though there is a
relation between the variables : education and fertility. relationship between literacy and population growth, a null
hypothesis may state "There is no relationship between
II. On the Basis of Functions: literacy and population growth." The null hypotheses are
There can be two types of hypotheses on the.basis of functions: formulated for testing statistical significance.

(a) Descriptive Hypotheses: These hypotheses describe the (c) Statistical Hypotheses: These hypotheses are statements
characteristics such as rate, size, form, or distribution of a relating to statistical population. These are derived from a
~adable. The vari_able may be an individual, organisation, sample. These are quantitatively measurable. For example,
mshtuhon, s1tuahon,.event or an object. For example "The "Community A is more literate than Community B."
rate of poverty is more in rural areas of India as compared
 36

___ __________
, SOUl{CES
_:,_
Business Research Methods (S. Y.B.M.S.: SEM.J
.~

OF HYPOTHESIS
------~ VJ Introduction to Business Research Meth ods .
market, and develop a hypothesis such as " Ed11cated cr'.stomers
prefer branded items than illiterate customers._" The history of
37

science is full of instances of discoveries mad e iust beca use the

Hypothesis can be developed from a number of sources. Sorne right person happened -to.make the right observation du e to
the important sources are as follows: characteristic life-history and exposure to the world of events.
1. Intuition: Analogies: Analogies facilitate the development of hypothesi_s.
A person may get ideas t? de~elop hypothes~ due to one's OIVJi 6. Jn the study of sociology, one comes across analogies wherein
intuition. Ideas can str'.ke like a flash. It 1s often said that the society is compared to a biological organism, the natural
reflective mind is the spring of knowledge. The story of laws law to the social law, thermodynamics to social dy namics, etc.
· gravitation propounded by Newton at the sight of falling app~; Such analogies provide valuable insights, which facilitate the
is the case of intuition. development of hypothesis.

Research Studies: Culture: Culture provides the basis for developing hypotheses.
2. 7.
Research on one problem can help to develop hypothesis 1 Culture involves knowledge, beliefs, morals, laws, customs and
solve another problem. Also, the research studies conducte~ traditions, etc. The cultural norms and values are passed from
earlier can be a source of probl_ems, as the resea~her can develop one generation to another through institutions like famil y,
hypothesis based on the earher research studies. educational institutions, religion, etc. Western cultures are
affecting the Indian society, especially in urban areas. The
3. Consultations: younger generation is more influenced by western culture in
The researcher can hold discussions with experts to develop urban areas, and therefore, a-hypothesis can be developed as
hypothesis. For instance, in academic research, the research "Urban yo11th in India is more influenced by western c11lt11re as
student can take the help of a guide to develop the hypothesis, compared .to rural youth in India."
which can be subsequently tested through research findings. Continuity of Research: The continuity of research in a
8.
4. Theory: particular field can facilitate the development of hypotheses.
As the research progresses, the researcher may reject some of
. A hypothesis can be developed from a body of theory, which
the hypotheses, which may lead to formulation of new ones
may lead by way of logical deduction, to the prediction thatil
capable of explaining dependent variables in subsequent
certain conditions are present, certain results will follow. A
researches on the same subject.
theory represents what is known. Logical deductions from the
theory lead to new hypothesis. The hypothesis must be valid,
1f the the_ory holds, true. For instance, the theory of human
MEANING AND DEFINITIONS
relat10ns m management states that effective human relations
help to improve productivity. On the basis of this theory a OF RESEARCH DESIGN
hypothesis ~an be ?~veloped that "Effective manageme:it.
labour relattons facilitates higher prod11ctivity."
A research design is defined as "a logical and systematic plan
5. Observation: prepared for directing a research study. It specifies the objectives of
Hypothesis can be developed through observation. For instance, the _study; the methodologies and techniques to be ad.opted for
one can observe genera_] pattern of buying behaviour in the achieving the objectives." - Philips Bemad
"R esearch d • Busines~Research Methods (S. Y.B M .
· ·5.. S£1,1
st
colle t· e gn actually constitutes the blue p . '½
c wn measu . rint ~o 1
' rement and analysis of the data." '.' r th, Introduction to Business Research Methods·
B. The area where research is to be conducted. 39
"R · • David & Nach 1111-s .
. esea~ch ~esign is the plan, structure, strategi and 9. The methods or techniques of data processing.
tnvestigat10n conceived so as to obtain answers 1 of 111 eans 01
questions. It is a plan or an overall scheme or a progr:;setlrc\ 10. The resources required to con,luct the research.
research." . John. W11111 e 01
• Best
Thus it is obligatory for a researcher to undertake research d . NEED AND IMPORTANCE
Th . .h d h d I f es,!sl\
. e researc esign is t e groun pan or conducting the rese OF RESEARCH DESIGN
investigation. It provides guidelines to researcher to help hi:ch
keep a track of his actions and to know that he is moving in t~o Research design is vital to the researchers, as it enables to collect the
. right direction in collecting the data: Whatever may b~ the n,'lture ; right data to achieve the research objectives. The importance of
research problems, all research designs should con tam answers t0 research design is stated as follows:
. the following questions: 0
1. Guidelines to the Researcher: Research design provides
• What the study is about? guidelines to the researcher in respect of:

• Why the study is undertaken? • When to start and when to complete the research work?
• What data to be collected?
• What are the types of data required?
• Where data can be found? • From where the data to be collected?
• How the data to be collected?, etc.
• What techniques of gathering data will be adopted?
The above questions provide necessary guidelines to the
• In which areas, the study will be undertaken? . researcher. Therefore, the researcher will be able to collect the
right data from the right source at the right time. ·
. The research design niust include the following aspects:
2. Organizing Resources: Research design facilitates organizing .
1. A clear statement of the research problem. of resources for collecting the data. The resources include:
2. The purpose or obj~ctives of the research. • Funds required for collecting the data .
3. The time period of research study. • The equipments/instruments and materials required to
conduct the research (especially in the case of physical
4. The sources of collecting data. sciences like physics, chemistry, etc.)
5. The procedures and techniques of collecting data. • The manpower to collect the data.
6. The universe of research design, d . enables the researcher to make proper
The research esign h. h • turn will help the
t f the resources w ic m
?. The sample size of respondents, if any. arrangehmen o duct the research effectively, so as to achieve
researc er to con
the research objectives.
40
Business Research Methods (S. Y.B.M.S.: S£lvj
3. . .
D1rections ·Iv)
to the Research Staff: The research design p .
rov.1des
· This is becaus Introduction to Business Research Methods
necessary d irections to the researc h staff. 41
research design provides necessary guidelines in respecte, th, 5. Collection of Relevant Data: Research design helps to collect
. - Of: the relevant data and that too within a certain time frame. For
• Sources of data. instance, the research design indicates:

• Techniques for collecting data. • The area of research

• Area/place where the research is to be conducted. • Universe of research

• Resources to be utilised. • Sample size, etc.

• Time frame of research work. Therefore, the researcher or his staff will be able to collect
relevant data from the right area, from the proper universe and
Due to the above factors, the researcher can provide proper from the appropriate number of respondents.
directions to the research staff, so that they collect relevant data
6. Objectives of Research: Research design helps to achieve
to achieve research objectives.
research objectives. This is because; the researcher will collect
4. Selection of Techniques: Research design helps to select the right data at the right time and from the right source. Also,
due to the use of proper techniques of analysis the researcher
appropriate techniques both for data collec.tion and data
will be able to analyse the data properly, and then take
analysis. There are various methods of data collection such as:
. appropriate measures or action, which in turn will help to attain
• Survey or Interview the research objectives.

• Observation 7. Monitoring of Expenditure: Research design helps to monitor

research expenditure. Research design provides guidelines
• Experimentation. regarding the amount of resources or funds to be utilised for
research activity. In the case of commercial research, the
For instance, obs~rvation method may' be used for studying . researcher would get the funds approved from the sanctioning
buyers' behaviour. To get feedback from the customers, the firm authority. Accordingly, the researcher will utilise the funds as
may use survey or interview meth9d. per the research budget. Therefore, there will be proper control
over the funds in respect of research activity.
There are also various methods of data analysis such as:
8. Execution of Research Work: Research design helps in .timely
• Measures of Central Tendency (Mean; Median, Mode) execution of the research work. This is because; research design
• Time Series (Simple Moving Average, Weighted Moving indicates the start time and the completion time of research
Average) activity. If research design does not indicate the time frame,
there is every possibility of the research work gettmg delayed
• Correlation Techniques, etc. and the researcher may i:iot be able to achieve the obiectives.

For instance, to forecas t sales, a researcher may use the time 9. Motivation to the Staff: A systematic research design motivates
series techniques. To judge the effectiveness of advertising, the the staff to collect the right data from the righ_t source. Also due
advertiser or researcher may correl ate the increase in advertising to the timely completion of research activity, the rese\lfch staff
expenditure with that of the increase in sales . may be rewarded with monetary and non-monetary mcenl!ves.
 42 Business Research Methods (S. Y.B.M.S .. SEM-1 11)
Introduction to Business Research Methods . 43
For instance, research design will enable proper condu _ Techniques of Data Collection: The researcher_ must decide
research activity. Proper analysis of research work will enct O/ 3 about the techniques of data collection. The techniques depend
the organisation to take suitable action, which in turn w able
upon certain factors:
bring higher returns. As a result of higher returns, the res/Ulq
staff may be adequately rewarded. arch • Nature of problem

10. Improvement in Decision-Making: Systematic research desi Respurces available

facilitates proper collection of d~ta. Also, the selection of suita:
•
Time frame, etc.
data analysis techniques fac1htates proper analysis of d e
. . d . . ¾ •
Therefore, the researcher can arn~e at proper ec1s10n-makin · There are various methods of data collection such as interview
especially in the case of commercial research. g, method, observation method, etc.
Decision on Universe: The researcher must state the ·u niverse
4. for conducting the research work. Universe refers to the type
STEPS IN RESEARCH DESIGN ,
of population from which the research information can be
collected. For example, if the research is conducted on baby
The main steps in research design are as follows : foods,. then the universe will be mothers with little kids.
1. Define the Problem: The researcher must clearly define th Sample Size: The researcher must state the sample size. The
5.
problem. Clarity of the problem wi ll help the researcher/ sample size depends upon certain factors such as time frame,
decide on the research objectives. For instance, in the case 0 funds availability, nature of problem, etc. For instance, if the
commercial research, the research problem may be stated Government conducts a census of population, all households
'Decline in Sales'. Therefore, the objective of the research w~ in the country will be covered under the sample size. But if it is
be 'To Increase the Sales'. an academic research, the sample size may be very small, say
about 500 respondents.
/ In case of academic research, proper definition of the proble
will ena ble the re_searcher to develop hypothesis, which neeX 6. Areas of Data Collection: The researcher must state the areas
to be tested with the help of research. from which the data to be collected. The selection of area
depends upon:
2. Sources of Data: The researcher must decide about the sources
of data. The sources of data depend upon the research problem. • Location of respondents.
The sources of data are broadly divided into two groups: .
• Nature of problem,
• Primary sources
• Time fa ctor, etc.
• Secondary sources.
For instance, a company selling only in one local market w ill
~ormally, the_sources of data can be secondary as they are less conduct resea rch only in that local area. But if the company
ime consummg
consider · and less e xpensive.
. . Th e resea rch er may sells its products throughout the country, then it may select
th respo~dents from all over the country.·.
the help i:::c::;:r;r~:s; t e problem cannot be solved with
7. Methods of Data Analysis: Depend in g u p on the type of
research, the researcher should select the most appropriate
 Busi11ess Research Methods (S y ,
method for d · ·8-M.s.: St
analysis ata analysis. There are various metho Ai.,f
meth d such as measures of central tendency /s of d,1. /11trori11ctio11 lo Business Research Methods
o , etc. , 1111e s 45
8. eri, • Resources required ·
D ecis'
ion on Resources: The most important d .
research d · · eci · • Techniqu es to collect and analyse data.
es1gn 1s in respect of resources required 510nI
research work. Resources include: to cond~, 11. Approval of Research Design: The researcher may get the
research design approved from concerned au thor ities. For
• Physical resources instance:

• Manpower • In the case of commercial research, the research design may

be approved by th e top managem ent.
• Money or Funds
• In case of academic research, the research design wou Id
The researcher should make proper arrangement of r be approved by the research guide.
sot h at t11e research 1s
· conducted effectively,
• which in esourc
turn
help to achieve the research objectives. lVij • For social research, the approval of research design may
be required from the sponsoring agency.
9.
Period/Time frame of Research: The researcher has to rn kt After approval, the researcher may make modifications in the
an important decision regarding the time frame ohhe rese a, research design, if so required.
. . r· f d
activity. mie rame epends upon certain factors sucn11
arc
availability of funds, availability of manpower, type of researci\
etc. · ESSENTIALS OF A GOOD RESEARCH DESIGN
For instance, in the case of highly complicated research, the
To make research design more effective, the researcher must consider .
researcher may require a longer time frame to conduct the certain guidelines or essentials:
research activity, and vice versa.
1. Focus on Objectives: The research design must focus on the
10. Preparation of Research Design: The researcher must prepare research objectives. The research objectives must be very s]ear
the research design after following the above steps. The research to the researcher as well as to the research staff. The research
design should clearly include the following aspects: objectives differ depending upon the type of research: For
• Statement of the research problem
instance, the research objective of commemal research .will be
based on commercial angle, and that of social research will have
• The objectiv_es of research the objective based on social welfare.

• Sources of data . Example of Commercial Research Obj~ctive: 'To increase 11'.~rket

sir are from current 10% to 20% wit/1111 a penod of 5 yea1s
• Time frame of research activity
Example of Social Research Objective: 'To reduce poverty of
• Universe of research . the weaker sections.'
"bTt . The research des ign sho uld not be ri gid. The
• Same size of respondents . 2.
Flex1 I I y. . I be flexible dependi ng upon the situation.
research design mus I · f
• Areas of data collection . .
For mstance, a resea1·ch desiITT1
o indicates a particu ar IITTle rame
 Business Research Methods (S. y BM
to comp! t h · · .S.: SJ:J,,i
be in a e _e_ t e research work. ·But the research staff · ·IJ;
In s posi~10n ~o collect the relevant data within the fun111<1y 1½ Introduction to Business Research Methods
uch a situation, the time frame can be increased. e &"ll\ _47
3. • The availability of time.
~ilo~ ~tudy: It is always advisable to conduct a pilot •
. The availability of manpower .
final12mg the research design. Pilot study is a research8 ~dy Ii
U~dert~ken on a small segment of the research univers Ctivi\ •The methods of data collection and analysis, etc .
pilot study is conducted to find out whether or e. 1l;
researcher would be in a position to collect the relev not t¾ Depending upon the funds and other resources, the researcher
may take suitable action. For instance, a researcher may not
from the sources or areas so selected, and whether orant da1
plan for an exhaustive research work, if there are constraints
·collected data would be relevant to solve the problernn~t t¾ relating to money, manpower or time.
pilot study indicates a favourable response, the research~ ft¾_ 6.
go ahead with the research design. Otherwise, he rnay ~Illa) Simplicity: The research design should be simple and easy to
necessary changes in the research design with respect to: a¾ understand. The language used in the research design must be
clear and simple. Wherever required, research design must be
• Sources of data. supported by footnotes, Technical jargons must be avoided. The
use of simple language would enable the approving authorities
• Area of data collecti0n.
to approve the research design and accordingly sanction the
•Techniques of data collection, etc. funds, especially in the case of commercial research. Simplicity
of research design would also enable the research staff to collect
4. Accept;mce: The research design must be acceptable to th the relevant information, and also to analyse th_e data properly.
persons concerned. For instance: 1 7.
Cost-Effective: The research design 'should be cost-effective.
In a commercial research, the research work based on research
• In the case of commercial research, the research design
design must bring benefits to the organisation. The research
must be acceptable to the higher authorities. This is
design should enable ptoper collection and analysis of data,
because; the higher authorities are required to approve ii
which in tum should facilitate proper decision-making. The
and accordingly sanction the resources to conduct the benefits on account of sound decision-making must be more
research activity.
than the cost incurred on research activity:
• In the case of academic research, the research design must In case of social research, cost effectiveness must be seen from
be acceptable to the research guide, as he needs to approve the long term point of view and not just short term point .of
the research design before the research activity starts. view. The money spent on social research must.generate social
development benefits from long-term perspechve._
• In case of social research, the ~esearch design must
acceptable to the social or other organisations that are going 8. Ease in Implementation: The research design should facilitate
to finance or sponsor the research activity. proper implementation of the research act!vity. As .far as
possible, the research design should. avoid complicated
5. Suitability: The research design must be suitable to achieve procedures and techniques which are diff1cultto adopt/ follow.
research objectives. Certain factors to be considered while Th
e researc h w ork should·not overburden the_research
ff t staff,.as
d AI
finalizing the research design are as follows: the quality of research work may get adverse Iy a ec ': . s0
sufficient time must be given to the staff for collecting and
• The_availability of funds.
analysing the data.

j'
 9, T . . Business Research Methods (S. Y.B M .
ra'.1ning to th R . .5.. 5£1\,J
Proper tr . . e esearch Staff: To conduct effective -/i,I
helps t ammg
. must be given
. tothe research staff. resec1
T . tc~ Introduction to Business Research Methods
behavioo improve: knowledge, attitude: skills and ra1l\in! 7. Resources Required for Research 49
skills ur. The research staff not only requrres knowJed 80 c1i/ B.
t to conduct proper research but also the· right ge &nj Time Frame for Research
ohwards the research work, and good social behavi attitlld
t ey interact with team mates and with the respond our Wheii1 . (Refer to expl~nation of the above points under the topic - Steps in
Research Design.)
10. ents
Selection of Right Techniques/Methods: There are v ..
methods of collecting data such as:
• Survey/Interview
ariolls
LTYPES OF RESEARCH DESIGN
• Observation The types of research design can be broadly divided into two groups:

• Experimentation TYPES OF RESEARCH DESIGN

There are also various methods ;f data analysis such as:
Exploratory Research
• Measures of Central Tendency (Mean, 'Median and Mode) Conclusive Research

• Time series, sw;:h as moving averages f { t

Descriptive Research Diagnostic Research
• Correlation Techniques, etc.
I. EXPLORATORY RESEARCH
Depending upon the type of research, the researcher should
· select the appropriate meth9ds for data col1ection and analysis. It is conducted to explore irlformation about the nature or ca~ses of
research problem. It is conducted when the causes of the research
problem are not known to the researcher. For instance, management
AREAS/ SCOPE OF nESEARCH DESIGN ' may conduct exploratory research to find out the causes of declining
sales in the past few months. The sales may have declined due to
number of factors such as:
The following are the main areas or scope of research design:
• Problem in quality of the product_.
1. Sources of Data
• Increase in competition.
2. Techniques of Data Collection
• · Ineffective promotion-mix.
3. Universe for Research
• Poor management of channels of distribution.
I 4. Sample Size
• Incompetence of the sales force, etc.

I
5, Areas of Data Collection
There may be a possibility of some cause of the problem and the
6. Methods or Techniques of Data Analysis exact cause is not known to the researcher, then in such a situation
an exploratory research may be conducted to find out the most likely
cause(s) that lead to reduction in sales.
 50 Business Research Methods (S. Y.B .M.S.: SE/\,j,[v
Exploratory research can be conducted forthe f oIIowmg · purpo I Jnt•oduction
, to B.usiness Research Methods 51
ses: . _ Development of Hy~othesis: Exploratory _research may be
2
• To define the problem more clearly. required to develop suitable hypotheses relatmg to the research
problem. A hypothesis is a proposition, which can be put to
• To develop hypothesis. test to determine its validity.

• To identify alternate courses of action.

3. Research Priorities: Exploratory research may be conducted to
establish priorities for research work. For .example, an
• To isolate key variables and to develop relationships a 1110
. .
the variables for further exarrunat10n.
ng exploratory review of customer complaints may indicate the
product(s) that require urgent attention on the part of
To establish priorities for further research. management to take appropriate i;\Ction:
•
FEATURES OF EXPLORATORY RESEARCH: METHODS OF EXPLORATORY RESEARCH:

1. Unstructured: Generally, exploratory research do_es not have a 1. Secondary Data Analysis: The researcher may analyse the
formalized set of objectives, sample plan or questionnaire. It is relevant secondary data for gaining information to solve a
conducted when a researcher does not know much about the particular problem.
problem and needs additional information. The researcher.may · _ Experience Surveys: The surveys may be conduded to get
2
collect the data from any sources - primary or secondary responses fro!ll those who experienced problems or difficulties.
sources. For instance, experience surveys may be conducted on the
students who have faced difficulties or problems during the
2. Flexible: Exploratory research is systematic but it is very
examinations.
flexible. It allows researcher to investigate the problem frorn

/
whatever sources he/ she desires. The extent of information also 3. Case Analysis: The researcher may analyse former situations
depends upon the researcher. similar to the present one, so as to arrive at suitable solution to
solve the present problem.
3. Informal: Often exploratory research is conducted at the outset
of a research project. Exploratory research is aimed at gaining 4. Focus Groups: Small groups of people discµss a problem - its
additional information about a topic and generating possible causes and its effects. The focus group discu's sion is monitored
hypotheses. Therefore, it is described as informal. by a moderator.

USES OF EXPLORATORY RESEARCH: 5. P~ojective Techniques: The projective techniques help to gain
insights into problems or situations. The researcher m.a y explore
1. Background Information: When little information is known information from a group of participants. They may be given
from the existing data, the.researcher may conduct research to incomplete sentences to complete, which may help to gain
gain background information. For instance,.a doctor may obtain insights into the problem. For instance, the sentence may be:
background information about the patient before treatment. In Mr. A does not attend the lectures because ..... ... ...
the case of commercial research, the researcher may obtain
I background information of the past performance of the firm,
the competitors, the practices of the firm, and so on. This will
help the researcher to probe the problem properly.

I.
 II D l3 usiness Research Methods (S. Y.B.M.s .
. ESCRIPTIVE RESEARCH: .. 5€A1,1~
It is conduct d .
Introduction to Business Research Methods
aspect f e to obtain descriptive information about
and Period II (April to June) Th I · · 53
detaile~ a prob)em. For instance, a resea:cher may like tocer1•i1 th
. • e sa es m Penod II have increased.
D 1.
e~
t _rnformation about students appearing for M.Com p kiio¼
versity of Mumbai in respect of age, income, gender, occuart 10/ !
In Period II, e marketer had also increased advertising. Therefore,
the manufactu:er (through causal research) may like to know
whether advertiSmg has caused the increased sales in Period JI.

A descriptive research may be undertaken for commercial PUr OBJECTIVES OF CAUSAL RESEARCH:
A marketer may try to find answers for questions like: Pose. 1.
To understand the cause-effect relationship between two or
more variables. .
• Who the customers are? (Firm's/Competitors') .
,2.
• What they buy? (Quantity, quality, size) To focus on those variables or elements having greater positive
effect.
• Where they buy? (Place - departmental stores, malls, etc.) 3.
To eliminate certain variables or elements having negative
effect.
• When they buy? (Time/ season)
4. To develop _action plans .
• How they buy? (Cash/credit/installments)
• How they use the product?
,_ MEANING . OF SAMPLE
!..,_ . . AND SAMPLING
, Through descriptive research, it would be difficult to answer the
question "Why". In order to answer the question "Why" - people

/
A sample is a part of the universe that can be used as respondents
buy certain items - causafresearch is required. · to a survey or for the purpose of experimentation, in order to collect
relevant information to solve a particular problem.
TYPES OF DESCRIPTIVE RESEARCH
Donald Tull and Dell Hawkins define sample as "those individuals
. 1. Cross-Sectional Studies - measure popu!ation aspects at only chosen from the population of interest as subjects in an experiment
one point of time. For example - sample survey by a magazine or to be the respondents to a survey."
publisher of magazine readers.
"A sample, as the name applies, is a smaller representative of a
2. Longitudinal Studies - measure repeatedly the same larger whole." - Goode and Hatt ·
population over a period of time. For example, a marketer may Sampling design is a plan designed to select the appropriate_ sample
undertake panel surveys.
in order to collect the right data so as to achieve research ob1ectrves.
III.. CAU~AL (DIAGNOSTIC) RESEARCH "Sampling is the selectio~ of certa'.~ percentage of a group of items
according to a predetermined plan • Bogrdus
Causal Research investigates cause/ effect relations_hip between two
or more variables. For instance, a research may be conducted to fmd
out the relationship between advertising and sales. For example, a
marketer sold the product in two periods say Period I Gan to March)
 Business Research Methods (S. Y.B,M.s .S
L METHODS OF SAMPLING
' .,
1 .. Ef11,J~
Introduction to Business Research M~thods
The method f . . . . nd 55
so samp1mg can be broadly dlV!ded mto tw 6. U er probability me th od, each element of the population
tkhnows m1advance about the possibility of being included in
esampe.
. r-----------, o grollPs:
METHODS OF SAMPLING
Types/Methods:

Probability Methods The various probability methods are explained as follows:

Non-Probability Meth 1.
• Simple RandomS111npli11g '---:---.::-:..:.::_•Ods Sh~ple_Random Sampling: This is the most popular method
• Stratified Sampling • Cm1ve11/e11r Sarnpfiug which 1s normally followed to · collect research data. This
• Cluster Sampling • Judgem,111 Samp/i,,g technique provides every element or unit (such as household)
• Multl-stageSamp/ittg • Q11ota Sanv1li11g
an equal chance of being selected in the sample. It is suita,ble
·• Snowball Sampling
for selecting a sample from a small homogeneous group like
M.Co'm students in the University of Mumbai or Upper Income
I. Probability Methods: Households in the city of Mumbai.
There are two sub-methods:
Probability sampling is also known random sampling. Probability
means possible chance. Therefore, each element of the population (a) Lottery Method: Where each element is given a number
has known chance or opportunity of being selected or included in and then the numbers are mixed or placed in a box or urn.
the sample. For example, if a sample of 100,students is to be selected By draw of lots, the ·sample is selected. ·
' from a population of 1000 students, then each student has a 10%
chance of being selected. (b) Random Tables: The elements or units are given numbers,
and the numbers are placed in rows. ·The sample is selected
Features: from the rows at random. For example, the researcher may
decide that all ~lements in 2nd and Sth rows to be the sample.
1.
It •is a systematic and objective method of sampling that 2· Stratified Sampling: The population is divided into various
provides equal chance to every element of the population in strata or segments based on income, occupation, age, religion,
· getting selected in the.sample. gender, etc. After dividing the population into different
2. segments, certain number of sample members is selected from
The results of probability sampling are more accurate a11d
reliable. each stratum or segment. There are two types of stratified
sampling:
3. It helps in the formulation of a true representative sample by • Proportionate Stratified Sampling
eliminating human biases.
• Disproportionate Stratified Sampling
4. Probability sampling is commonly, used in research because of
its sound theoretical and mathematical base. In the case of proportionate sampling, certain fixed percentage
is taken of each stratum to represent the sample. I~ case of
5. It is a mechanical method rather than mental process of disproportionate, there is no such percentage considered to
sampling, which eliminates personal bias in the sample represent the sample.
selection.

J
 56
Business Research Methods (S. Y.B .M.s.: SE
Examplei ' ·
A-r-11
I Introduction to Business Research Methods 57

Income • Stage 4 - Randomly selecting and interviewing private

No. of Proportio- Disprop
Households ' nate (1%)· f Ot,
sector banks from the selected towns. This is called as four-
•onate stage sampling design.
Upto 90,000 20,000 200
90,000 to 10 lakh 180 If the samples at each stage are randomly selected, then it is called
5,000 50
10 lakli to 100 lakh 2,000 60 as multi-stage random sampling.
20
- 30
Sample Size , 270
270 Multi-stage sampling method is suitable in case of large geographical
area. This method is cost-effective and is easy to administer.
3. Cluster Sampling: Cluster sampl mg is also called as 'A
Sampling'. Under this method, instead of selecting indivi/1 11. Non-Probability Methods:
units, fhe researcher di vides the popula tion into clusters~
groups and ·accordingly sample is selected. Non-probability sampling is also called as deliberate sampling. The
researcher deliberately chooses the specific items for the sample as
: 0x~:~~~~~;:2,e;~ea~et~~o5~~o:!~;~t :;~~c~i;s~ ~ : %:~si~'.:' a part of the research study. This type of sampling method does not
t , provide any basis of estimation of the probability that each item in
certain groups· based on certain characteriS ics or on the bas\ the population has being included in the sample. For example, for
th
of area. If the researcher wants to study e reading habits~ the study of literacy level of people living in a state, a few towns
the students of the University of Mumbai, th en the studen1 and villages may be deliberately selected for intensive study to act
may be divided into various grou.ps on the bas is _of area, sa1 as a representative sample of the entire state.
Western Mumba i, Eas tern Mumbai, Sou th Mumbai, and Nort!
Mumbai. If the total sample is 200 students - boys and girl\ Features:
then 50 students can be selected from each zone.
1. The choice of the researcher is supreme with regards to selection
4. Multistage Sampling: Mu lti-stage sampling involves multiple of sample from the population for the purpose of the study.
sta ges in sampling design. Generally, th ere are maximum ol
four stages in m ulti-stage sampling. For example, in order to 2· In case of non-probability sampling method, there is an element
study the level of efficiency of priva te sector banks in lhdi, of biasness on the part of the researcher.
the researcher may resort to multi-stage sampling design ana 3. The results of non-probability sampling may be inaccurate as
the stages could be: , the sample may not truly represent the universe.
• Stage 1 - Selecti ng large primary sampling unit such as 4. In this method each and every element of the universe may not
states in India. get an equal chance of being selected as a sample for the research
study.
• Stage 2 - Selecting d istricts from the states and interviell'
all private sector banks in the selected d istricts. This i! 5. Generally, this method invo lv es less time and cost in
called as two-stage sampling design. conducting the research activity.
• Stage 3 - Selecting towns from th e districts and interview
aU private sector banks in the selected towns .This is called
as tlu-ee-stage sampli ng design.

I
 Business Research Methods{S. Y.B.M.S.: S/;
Types/Methods:
i\.i.lf,
The various no b b' . Introduction to Business Research Methods
n-pro a 1hty methods are explained as foll 59
ows, 4. Snowball Sampling· s b II .
I. h ·
samp mg tee nique that · now a sampling
d b is a non-probability
l. Conve~ient Sampling: The sample is selected as e ·
otential b' . is use Y researchers to identify
convenience of the researcher. Under this metho/ rt~ p su iects m studies where subjects are hard to locate.
element of the population does not get a chance of :vel Researchers use this s r
sele~ted. Fo_r instance, the researcher re_searchi~g intoe1n1 tud . amp mg method when the sample for the
s y is_ very rare or is limited to a very small subgroup of the
reading habits of M.Com. students of University of M 1~
pofpula hon. This type of sampling technique works like chain
may sel~ct the sa1:1ple respondents by visiting certain re erra1. After observing th 1· · •a1 b'
as per his convenience. The researcher may consider fact g~ f .t . e m11 su ject, the researcher asks
~r ~ssiS ~ce ~om the subject to help identify people with a
select the sample of respondents, such as his contacts Wi;s1 s1m11ar trait of interest.
college authori'ties, convenient location of the colleges, andt~
on. He may select the students from those colleges with Wh The_ process of_snowball sampling is much like asking your
he has good contacts, or which are located at a convenient pl 01 subjects t? nommate another person with the same trait as your
· ace . nex~ subject. The researcher then observes the nominated
2. Judgment Sampling: The sample is selected based on th• subjects and continues in the same way until the sufficient
judgement of the researcher or some other person assisting 1 number of subjects 1s obtained.
advising the researcher. For example, if the researcher ~
For example, if a researcher is studying a rare disease; the
studying the newspaper reading habits of college students, the' r~searcher may_ opt to use snowball sampling since it will be
he has to use his judgement to select a proper sample ~ difficult to obtain subjects. It is also possible that the patients
respondents. The researcher may get assistance from teachen with the same disease have a support group; and therefore,
of the college to locat.e those students who actually reai one sample respondent may lead to another having the same
newspapers. This is because; the teachers may have a bett~ problem.
knowledge of certain students who actually read thi
newspapers. -.
. -PROBABILITY MET~ODS vjs
3. Quota Sampling: Under this method, the researcher allocate
certain quota (relating to respondents) to certain groups unde ic: NON-PROBABILITY METHODS
study. The quotas may differ from each area depending upon
certain factors ,Jike-age, occupation, income, etc. For example,1 Probability Methods Non-Probability Methods
researcher studying the newspaper reading habits of colleg;
1.Meaning:
students may select 10 colleges for data collection. He may fa
Probability sampling methods The non-probability methods of
quota for each college based.on certain criteria. He may selet1 enable true representation of
100 students from one college, may be because the number d sample enable the researcher to
universe. Each and every element collect data as per the convenience
students is ·more ,in that college; and he may select only 21 of the universe gets an opportunity of the researcher, and these
students .from another college because of less n~mber ol or chance of being included in the methods do not enable each and
students·in that college. sample. every element of the universe to be
included in the sample.
'
60 Business Research Methods (5. Y.B.M.S.: SEM-IV) 61
/Jl/rodu ction to Business Research Methods
2.Methods: ----- 7 Bias in Sa mpling:
The probability sampling methods The non-probability methods
There is no bias in selection of There is a possibility of bias in
include: include:
sample·respondents. This is because; sa mpling, as the researcher may
_Judgement Sampling
- Simple Random Sampling sa mple members are chosen by select the sample members as per
_Snowball Sampling
- Multistage Sampling adopting a mecharuca l method his convenience or accidenta lly
_Convenience Sampling
- Cluster Sampling rather than using judgement and an d not by adopting a sys tematic
_Quota Sampling
- Stratified Sampling convenience of the researcher. mathematical base.
3. Time Factor:
Non probability methods may
8. Representative Sample:
The probability sampling methods The sample members represent the The sample selec ted may no t
require comparatively lesser time
require relatively more time in universe beca use each and every represent the un iverse. Each and
to collect and analyse data as the
collecting and analysis of data, element of the universe gets a every elemen t of universe may not
iample size is generally less as
especially when the sample size chance or opportunity of being get an equal chance of being
compared to probability methods. selected or included-in the sample.
is large, and covers a wider area. selected as sam ple member.
4. Cost Factor: - 9. Convenience to Researcher:
The probability.sampling methods The cost involved in collection of Probability methods may not The non-probability methods offer
are relatively more expensive as data and analysis of data is provide convenience to researcher convenience to the researcher to
comparatively less because the to collect da ta. The sample collect data. Also there is
they involve collection and
analysis of d~ta from larger · sample members are generally less members are selec ted by adopting con venience in ana lysis of d ata
sample respondents. in number as compared to mathematically base, and therefore because the sample size is less and
probability methods. the sample members may be few responses are collected from
spread over a larger area which the sample members. For exam ple,
5. Area Coverage:
Generally, the area coverage is may ca use a little bit of the convenient method of
The probability sampling methods
may involve comparatively larger restricted to local area and as per inconvenience to the researcher sampling offe rs convenience in
area coverage to collect data. The the convenience of the researcher. to collect da ta. data collection.
probability methods are used to Therefore, the scope of area 10. Suitabilitt;:
collect accurate and exhaustive coverage is limited. The proba bility methods are The n? n-probability methods are
data from a large number of suitable, where the·da ta collected more su itable in the case of
sample respondents. needs to be accurate and objective academic research or at times
6. Accuracy ofData: in nature. These methods can be even in the case of social research.
The probability sampling methods The responses given by the sample used in commercia l research However, such methods m ay not be
may provide accurate and reliable members may be inaccurate as the wherein the respondents need to used fo r commercial research.
data because the sample members sample may not truly ~epresent the be true representative of the
are chosen systematically. Since universe. Secondly, the researcher wliverse under considera tion .
the sample represents the uni·verse, who comes across the respondents
the data collected is expected to be accidentally, may not get adequate
accurate and reliable, information from the respondents.
62 Business Research Methods (s.'y ..
BM .5.:SEM
-- · -
REVIEW QUESTJONS.l . Introduction to Business Research Methods 63

·1
·Iv;
OBJ - QUESTIONS
. ECTIVE .

1. What·and
State is research?
e . · Ex PIam
. its
. nature and characteristics
Fill in the blanks with th
2. . xpamtheobjectivesofresearch
1 · J. - - is a systematic . e co~ect option:
1. ranc o knowledge. mvesttga tio n to search for new facts in any
3. iscuss the
0Explain th need
. an d importance
. . research.
of business b h f
4. D . e different types of research. (a) Research design
(c) Research (b) Sampling
5. escnbe the stages in research process.
Research uses (d) Survey
What are the characteristics of good research? specific proble-;;;;- me th od to find facts or to provide solutions to
6. 2.
P a'.11 t e meaning and nature of hypothesis.
7. Ex I · h (a)' scientific
(b) social
Exp lam the significance of hypothesis. . (d) subjective
8. (c) natural
9.
What are the various types of hypothesis? ;;;;:ict."
research is designed to solve practical problems of the modem
10. State and explain the sources of hypothesis.
11. Define research design. State the need and importance of r (a) Basic (b) Applied
d~ign =•"o ' (c) Quantitative (d) Qualitative
. - - - researc 15 to expand man's knowledge.
12. What are the steps involved in research design? 4. The main goal of h·
13. Explain the essential features of a good research design. (a) Basic (b) Applied
(c) Quantitative (d) Historical
14. What are the areas or scope of research design? 5. - research involves interpreting past events to predict future
15. What are the types of research design?
ones.
16. Explain the methods or techniques of sampling. (a) Historical (b) Exploratory
(c) Descriptive (d) Basic
17. Write short notes on: - research is conducted when I.he causes of I.he research problem
(a) Variables 6 are not known to the researcher.
(b) Qualitative and Quantitative Research
18. Distinguish between probability sampling methods and non- · _ (a) Historical (b) Exploratory
(c) Descriptive ·(d) Qualitative
probability - variable is the ore which is martlpulated by the researcher.
19. A marketingsampling
manager,methods.
on seeing the sales performance of Brand A,
observes that its sales have declined in past three years. He asks his 7 (a) Independent (b) Dependent
research wing to study the problem to determfae why the sales have _ · (c) Intervening (d) Moderating
- variable is directly not related to the purpose of the study but
declined. What type of research will this be? Give
1
your
01 reasons.
,0. A phMmaau<ical cump~Y likes w,I IW' '° YJ''
. med'c="' 8 (a) Moderating
controlling blood pressure in order to determ~e which 1s the best. _
rnay affect the dependent variable. .
(b) Extraneous
(c) Intervening (d) Depend_ent
What is the set of hypotheses that can be tested • _...c- can be called as a blueprint for coIIect10n, measurement and
. A company wants to launch a neWP droduct in ~e ~arket. What type
9. analysis of data. (b) Sampling
21 of research design would the R&D epartmen use. (d) Hypot.hesis
(a) Research design
(c) Research
 65

I Y.B.M.5
Bu,i ess Rescnrch Methods (5. t ·,,troJucfion to Business Research Methods
JI .: SEM-IvJ
1
64 -d•" w as
10. is ,a plan
achieo• ;he ,.,.,nh
to select
11
oh1•"'"·
the appropria_te
O
sani P .e co ec t the rig1il o. ---;,wole•• eliti,,g o,d<ng d=""""""'
Jabol•tion
(•I Hypodio is M , ,.,.,,hd,sign 2 (•I D•" ooJJ•ti= (bl D•Ja P""""""'
(•I Sam pl•0 desi g>' (di p , ;acoU"'<ion (,I D•<• w•lysi• (d) DaJa <n""1'reJation
H Tl• - ;s, " "' " pmposi<iOO fo,mola <ed ;o d•Jeno;, , iS 2'· Wh<" • ,.,..,,he< is able Jo ,pply ,,.,.,ch findings Jo J,ege<
,,Jidiry. papol•<i°' i<;, ,,lied " -
(•I qoo <iom•i• (bl ¢ , a<chd• 'W' (•I fogic,l (b) obj«<i"
(•I ~ mp<• do ig,, (di hypo<h•is (•I =P'''"l (d) gen<"J;,,,ti=
n. _ hy po<h•' ' , , , , ,h, 1ow•' Jeoel of , 1,s;,adioo 2'· - • • ,.,..,h •J<,i<y w,deeW<~ , ,m,Il ,egm~I of ;J,e

(•I
(,I Comp<
C • -
• - "" ·(bl Ac,Jy;ical
(dI Rs Ja;iooa< ,.,.,..,,,uuee=
( •I Ptl o;s<od Y (b) DaJa ooUection
n ._ ="",.,~ w ""''"""" "" " " "'""' o la<ioosh1 '3.
,.,_~"°'mo• ,., ;, hi••
(•I s,,nple '"''"
-
(di HYPo""" ro,mo1'tion
hyvo<h<= ,re qowtiJatieely m=m•d
(•I
(, I
o,scrip<i••
""""°'' 0) Di•<'ffi"'
(d) ,a,;o,iotl
(•I o,s<rlpti«
(,) Arutly&.i
(b) 5;,;;,tical
(d) N o1J
H. •••""" w•- ""'P""' m.> d whe" <h• sa mple fo,,
0
U ,-C- 'C' hypo<h•ti"I s<•«-" d=yiag wtia< is ~plicitly
,mO, ;s,•Y " " o, <;,,,;; oJ ' ,od<'<ol m w°'king hypotl,••
(•I Com•''"' (hI j odgm'"' (•I Noll (b) C.Mi
(cl '"""''" (di CJ;o<ec (,) R0•tio"'1 (d) Comp!<'
rn • ~•"'"""'"
JS. _ research is undertaken to solve theproblem of wa ter s""''. 25- info-"=''" be ooflected popu I"'"'"
_referstothetypeof . from wtiieti<tie ,.,..,.,
c es,gn (b) Universe
(a) Applied (b) Pure (a) Resear h d ·
c sEmpirical (d) Historical . ~<ypeof==hd,sign<s
(c) Sampling d h .
16. In
( )case of_ - . research, no commercial angle is u· , v alved . 26 (d) Data collection
' ""''"=
/ (bl Applied J""hl•m"'°°'kno=lolh . - w '""""-of""'a"ti
(d
() Oore (01 Cmpick• I (al D=ripti•• '=""'"·
l ?. - - - " ' " " ' • P'''"' pheoo=~ b JI ·
;~'t ~::;;:~;i:e:
(cl Core using mathematically
('I Qo,oHOHoe :a;:d (•I Espio,•""Y
~~~';!;:.merical data 27. . P"P°'"""
In casle of descriptive
"'" a P"iod
.
research,
(bl Qmdosi,e
(d) D,:gnostic
ofHm_e____ asure repeatedly the same

18. The . research er u n d erta kes (d) Descri

b ptive . (c)
(a) Scientific
Research stud, ·es
studies ·(b) Cross-sectional studies

="'"''"'"''"
publrcationslikejournals,
(a) hypothesis
Jlw ~ohlbooks
em. , -re- - repru•
'""'' y co ns idemhe
,od ring· pobldi..
va rious 28. Clos•u=pling is .iw otlled " (d) Lon~rudiMI sJudies
(a) area - - - samph!jg.
(•I ,m,pledoig,, OI qdmm,rr, (•I mo!H-,rag, 1 M sttatifi<d
19. In order to collect . (d) literature survey 29. Non-probability sa . . (d)· quota
a -- primary data from the field ti (a) random
(c) area mp mg is also (b)
called as
deliberate · . ·
(a) book . ' ,e researcher requires - - - - sampling
(d) · - - sampling meth d m II-stage
(c) quest ionna ire (b) research report
hypothesis 30 not
In get
(a) a chance of being seolected.
snowball 'every
(d) element
u] · of the popu Ia 1·10n does

(c) quota (b) judgement

(d) convenient
 ·~-,v1. :,.: s12
- \ aJ, (IO) - ( . ••, , PJ - (b), (4) .- (a), (5)- (a), (6)- (b), (7) - "1-Iv!
(c) (17) c), (11} - (d), (12) - (c), (13) - (b) (14) (a), (8). (b)
(24) - (a) (~}sj18~;?>, (19) - (c), (20) - (b), (21)- (d;, <tl2?:> )- (a), <i6\~ Introduction to Business Research Methods
Match the following:
, (26) - (c), (27) - (d), (28). (a), (29) _ (b~' <23) •.(bj
II.
II. ' . •
I. 1.
t .
S ate Whether the following statements are True or F
, (30). (<1 ,
J
A. Group A
67

1.
Research · ft . a se: Pitre research Group B
1s o en referred to as 'scientific enquiry' i It 2.
(a)
2. problem or situation. n °a srec;~, 3.
Applied research
(b)
Behaviour of people
I 4.
Empirical research
Cause-effect relationship
Observation is undertaken to study certain situatio~s or Social science research between two or more

I
. 3. on experiments. events bas,d 5. Casual research variables
4. Researcl) findings should not be applied to larger population. (c)
Expands man·~ knowledge
A research undertaken to find out the reasons as to why (d)
Commercial angle
5. · a b sent from work 1s
remain . an
· exa,:nple of quantitative research
ernploy •es (e)
Experimentation or
Resear_c h may be undertaken to verify and test existing la,,,' observation
theories. s o, -
6. Ans: (1) - (c), (2) - -(d), (3) - (e), (4) - (a), (5) _(b)
7. Marl<eting research helps to develop good customer relationsh_; s
8. Applied research is also called as fundamental research. _P · B. Group A
9. There is no commercial angle in basic research. 1. GroupB
Continuous variable
10. Descriptive research is also called as statistical research. (a) Second independent
A variable is an entity that takes on different values. 2. variable
11. Non-continuous variable
A specific value assigned to a variable is called as an attribute. 3. (b) Measured outcome
12. Independent variable
Age is an example of non-continuous variable. 4. (c) Number of children
13. Dependent variable
5. (d) Income
Pilot study is a research activity undertaken on a small segment of Moderating variable
the research universe. (e)
14. ,.,. Manipulating variable
A hypothesis should be empirically tested.
15. Ans: (1) - (d), ·(2) - (c), (3) - (e), (4) - (b), (5) - (a)
"There is no relationship between literacy and population growth"
an example of statistical hypothesis.
16. N. Explain the concepts:
. Observation is an important source of hypothesis. 1. Research
17.
Research design helps to monitor research expenditure. 2. Variables
18. A research design should not be rigid . .
19. 3. Quantitative a~d qualitative research
Probability sampling is also called as deliberate sampling. 4. Research design
20.
Cluster sampling is also called a s area sampling. 5. Sampling
21.
In case of non-probability sa mpling method, the universe g e ts an equal 6. Hypothesis
chance of being selected as a s a mple for the research study .
22.
Stratified sampling is a type of non-probability sampling m e thod.

Ans: True: 1, 5, 6, 8, 9, 10, 11, 13, 14, 16, 17, 18, 20

False: 2, .3, 4, 7, 12, 15, 19, 21, 22
 ,......-
68
Data Collection _and Processing
69

TYPES OF DATA SOURCES

o ·ATA co'LLECTION pata play an imp?rtant role in research. Facts, information or

premises sy 5tema~ical~y collected and formally presented for the
AND PROCESSING purpose of drawmg mferences may be called data. Statistical
inforrn'at~on colle~ted, ~ompiled and presented for the purpose of
establishing relationships between variables can be included in the
'data. ·

Data can be collected from primary or secondary ·sources. Primary

I data refers to the information obtained first hand by the researcher
• .. Types of Data Sources . 011 the variables of interest for the specific purpose of the study.

Primary Sources of Data Secondary data refers to second-hand information gathered from
/ • the existing sources .
Methods of Collection of Prima,y Data
• There are two types of data, namely Primary and Secondary. The
Questionnaire-_Meaning And Importance prim~ry data are tho~e, which are collected afresh and for the first
• time, and therefore 1t happens to be original in character. The
Limitations of Questionnaire - For Reference
• secondary data are those which have already been co1lected by
Stages in Designing Questionnaire someone else and which have already been passed through the .
• statistical process. ~e researche·r has to decide which type of data
Essenti~ls of Good Questionnaire
• is appropriate for his study and accordingly he should select one or
more method of data collection.
• Types of Questions
Types/Classification of Q·uestionnaire- For Reference A researcher begins the task of data collection after the research
• problem has been defined and the research design is prepared. The
/ • Secondary Sources of Data researcher has to decide which type of data he would be using for
his study and accordingly has to select the method of data.
• Methods of Collection of Secondary Data - For Reference
For that he should. first explore the secondary data available from
• Primary Data Vs. Secondary Data various sources and examine the possibility of using it for his study .
Scaling- Concept And Techniques This is called survey of literature. In several cases, he will find the
• secondary data inadequate, outdated and unusable and hence
realises the need for co~lecting primary data.

PRIMARY SOURCES OF DATA

Primary sources of data refer to first hand information wh~c~ is

collected to solve a specific problem. The researcher takes up ongmal
 Business Research Methods (S. Y.B.M.S.: SEfv1
ro ·½
1 Data Collection and Processing
investigation with the target group to gather accurate inforrnati ~
0 71
• Observation method
and data.
• Experimental method
Features of Primary Data:
6. Availability: The primary d t .
' gathered by the researcher a -~~s ~ot available but needs to be
1. First Hand Information: It is collected directly_ frorn the
The researcher has to spend':: lotto; ~elp of his research staff.
respondents. For example, the college may take up primary data collect data from primary time, effort and money to
collection source from their students to -study the facilities sources The p · d
supplement the second~ry data'. · nmary ata sources
provided by them. The College may collect the data ~hrough
survey and/ or int~rvie~. The data collected through this source 7. Reliability: The primary data is I' .
is original and more specific. · check the reliability of the data ~e table. Th~ researcher can
1
data is collected from the · tdis collected first hand. Latest
· respon ents.
2. Ti~e-consuming: This type of data collection is a tirne.
consuming and a lengthy process. I~ involves: 8. Accuracy: The data collected from . .
primary sources 1s more
accurate. The researcher can check th f
• Deciding the types and size of respondents • f · . e accuracy o data at the
time o mterv1ews. He can check th b
• • h e accuracy y cross-
• Scheduling the survey/ interview questiorung t e respondents, whenever required.

• Gathering the responses Advantages of Primary Data:

/ . 3.
• Data analysis and interpretation, etc.

Expensive: Primary data collection is an expensive affair; as it

1. Provides First Hand l~ormation: It is collected directly from
the .respondents. The information can be collected through
involves: various methods - survey, observation and experimentation.
But the secondary data is second hand data collected from
• Training the research staff published or readily available sources.
• Preparation of questionnaire 2. In-Depth Information: The researcher can collect in-depth
• Printing of questionnaire, etc. information relating to the research problem. The researcher
can get detailed information with the help of projective
4. Paper Work: Primary data collection involves a lot of paper techniques. The researcher can cross-question the respondents
work in respect of: and obtain in depth information. ·

• Preparation of questionnaire 3. Reliable Information: The primary data is reliable. The

researcher can check the reliability of the data. It is collected
• Coding and tabulation of data first hand. Latest.data is collected from the respondents. But in
• Analysis of data, etc. the case of secondary data, it may ·not be reliable, as it cah be
outdated.
5. Types/ Methods: There are various . methods of collecting
primary data, such as: 4. Accurate Data: The data collected from primary sources is more
accurate. The researcher tan check the accuracy of the data at
• Survey method

\
r.
72 Business Resenrclz Metl1ods (S. Y.B.M.s .
- .. SE;lvf _
the time of interviews. He can check .the a~curacy by Iv,
. . tl1e respondents, whereve1 required. cr08i Data Collection and Processing
questionmg
. 0 ta· The researcher can collect specific data Limita~ ons of Primary Data: 73
5. Spee if1c a • · d d re] .
to the problem. But in case of secon ary ata, the resea c'lt1~t
may get gene_ral data, and he may have to p~t extra effo;ch~
The primary method of dat . -
limitations: a co 1lection suffers from certain
ifc data relati.no- to the problem- Quite often ts~
get spec I o h h d , spe .
data may not be available throug t e secon ary sources. Ctfii 1. Paper Work: Primary methOd f
0
paper work. For in t . data collection involves lot of
6. Supplements Secondary Data: The researcher may try to s collection with the h 1s a;ce, s~rvey method requires data
the pro~lem with the h~lp of seco_ndary data. Howeve~lv., is required in prckes:·p O {~eSttoru:iarr~. Also lot of paper work
econdary data is not sufficient, he will collect the requir d , tl classification and tabmlgt~ ata which mvolves editing, coding,
s . . d
from the prima1y sources. Thus, pn_mary ata supplernentst
e da,. u a 10n of data.
I secondary data. 2. Expensive: Primary data II t· ,
I
d co ec Ion methods are expensive as
c~mpare to secondary data collection methods For instance
1• 7. Enhances the Quality of Research Work: The primary d
t_ e survey method is expensive as there is a ne.ed to a oin;
enhances the quality of research work, as it is more accur a~ field staff to collect data Salary and k d be P~d
reliable, and specific to the res~arch problem_ ate, h aff · per s nee s to paI to
researc st . However in the case of s d th d h
• genera1ly no need of field
IS ' staff. econ ary me o , t ere
8. Helps in Formulation of Hypothesis: The use of document . '
sources helps in formulation of research hyp~thesis. When: 3. Sampling Errors: It is difficult to select a proper sample of
-investigator has more than one hypotheses m hand, primari . respondents, especiall! whe_n the universe is large and covers
data supports the selection of correct hypothesis. · large area. '.f~e samplmg ~rrors are greater when small sampie
1s selected, as It may not represent the universe or population
9. Flexibility: The primary method allows flexibility in data under study. Even when the sample size is large, sampling
collection. For instance, the survey method permits flexibiliry•. errors take place.
The interviewer can restructure or modify the questions
depending upon the situation. The interviewer may add 4. Time Consuming: The primary methods of data collection
01
delete questions depending upon the situation - the availabilin· involve a lot of time and effort on the part of the researcher to
of time for interviewer, the nature of respondents, and so 0~ collect data as compared to secondary method. For instance,
the survey method requires lot of time in collecting the data as
The interviewer can also ask questions in the language familiai
to the respondents. there is a need to interview the respondents to get the
respondents. A lot of time goes irt planning, getting
10. Helps to Overcome Resistance: Primary method of dati appointments, and in conducting the interview.
collection such as the survey method helps to overcome th1 5. Respondent Bias : The respondents may n9t give proper
!esistance of the respondents to respond to the interview. The responses. They may withhold certain data or may provide fake
researcher can convince the respondent to respond favourably responses in respect of certain sensitive aspects such as salary/
At times, the researcher may offer gift to the respondent to solicil income, investments, payment of taxes.
cooperation in getting responses. The researcher may use
persuasive skills to obtain the relevant information from the 6. Interviewer Bias: There is a possibility of interviewer bias. The
respondents. inte~iewer may fill the questionnaire on his own, or he may
edit the questionnaire as he likes.
 . Research Methods (S. Y.B.M S
Busine55 · .: St
74 ertain problems · ~-Iv
. g of Data: There are c bl . di . in Proce D ta Collection and Processing
J 75
7. Process1n there are pro ems m e ting of d ss~. a
in data. For instance, d. there are chances th ata. ¼i.'1 How much they buy?
more e itors, at dif;:'~ •
there are two ~~it differently, thus affecting the qu :er~ How they react to sales promotion schemes offered by the shop
editors ma_y d. Also there can be errors in ali~ 0 • owners?
h fin ings. , h. h coq• ,
researc . d tabulation of data w 1c also aff
classification an ects ,~ Whether or not they hag~le for the price?
uality of research work. 'I •
q . p. Whether or not they complain on their previous purchases?
8.
f Q · k Decision Making: nmary data Ill
Pr~ble: .. UI~aking. This is especially possible w!
aff~
,

qmc\ e~1s1:large and therefore large volume of respen t~ ;\~vantages:

samp e s~e problem in processing of data. The defns~ , There ~is no respondent bias. The respondents are not
may cr~a e f data may affect quick decision making t ay
phrocessbmlegmos especially in the case of commercial
t e pro , d .
soJv,
sear,t
r: interviewed. They are only observed, and they may not be
aware that they are observed.
where the problems need to ~e solve · at a quick pace. ·• First hand information about people's/ customer's reactions.
The information collected is reliable and accurate, because the
METHODS OF .COLLECTION • information is collected at the point of action or reaction.

OF .PRIMARY DATA . Disadvantages:

Time consuming, as lot of time is required for observing the
The primary data collection methods are as follows:
• actions/reactions of the people under study.
It is expensive method, as competent and train_e d staff is
I PRIMARY DATA COLLECTION METHODS l • required to re~ord the observations of the subjects under study.
! In depth interviewing is not possible, and therefore, the
! ! J .J • researcher may not get complete information.
ObJervation Experime11tal Interview Survey -
Method Method Method There may be bias of the researcher, as he may record certain
Method • observations as per his o~ judgement or feelings.

I. Observation Method:- Types:

Structured and Unstructured Observation:
The resea:cher obtains information of the subjects (persons) under •
~tudy_ w~th the help of observation rather than by way ol Structured Observation is based on the assumption that for the data
mterv~ewmg. For instance, a researcher studying customer buyin~ collection to be most effective every pertinent detail should be
behaviour at shopping ~alls; then he w_ould visit the shopping malli worked out in advance. The researcher identifies the elements to be
and observe the behaviour of the customers in terms of: observed well in advance. The observation is done as planned with
• What they buy/ prefer? very little deviation. The time allotted for observation to each

1 l
r
 p
Busi•ness Research Methods
. .
(S. Y.B. A,, S
ivJ. .: St

76 ~-' mation sought 1s pre-decided

d latent uuor h . h . on-_:t _a_
~i Collection
___ and_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _77
Processing l
respondent an. lled by the researc er. ~,lls~
of observation is contro Television Rating Points
ti n is relatively non-planned by the
Unstructured obse~~e~ts as and when the events take p[;s~ The TRP system was introduced in India in July 1986 to provide
and observes thee ch for the researcher but needs to be c ce,111 uick feedback on the viewership of television programmes on
more flexible app~~ection of the research topic.
0
lll1' 6oordarshan and Metro Channel of Doordarshan: At present,
to not to lose the TRPs are conducted for 300-plus TV stations in India.
. . d Undisguised·Observation: ,
• D1sg111sed an TRP is the criterion that indicates the popularity of a chan~el?r
. ised observation, the researcher acts as an ohs rogramme. This data is useful for advertisers or ad agenaes m
In case of disguti_ . ant or a member of the sample group u e~ ~edia planning. The TAM Media Research calculates the TRPs
well as a par cip .. h h . 11ow11i.
as . his identity and purpose so t at t e samp} ,, as follows:
he will concea1 d d. . e 8ro
rally· The researcher may a opt is guised obse ,
behaves na tu · rvaij First is frequency monitoring in which people meters are installed
techniques, when he believes that:
in sample homes and the people meters record data about the
1. The sample group may not give consent for observation channel watched by the family members.11).e people meter reads
the channels frequencies, which are later decoded into the name
ii. The sample group may manipulate their behaviour due tor of cha.n nels and the agency prepares the national data on the
presence of the observer/ researcher. basis of its sample home readings.
In case of undisguised observation, the researcher acts as an obse
and is physically present as a part of the phenomenon or saI!ln JI. Experimental Method:
group with formal consent. The researcher does not conceal [
identity and may not disclose the research purpose. The researc~
The experimentation method is mostly us~d in the case of scientific
obse~er is free to record the behaviour and reactions of the sam
group. 1 research study. With the help of experimentation, the researcher
may like to study the cause-effect relationship between two or more
variables.
• Mechanical Observation:
,
The experimentat~on method can be also used in developme~t ?fa
When the researcher uses mechanical devices like cameras,
new product. For instance, a firm may like to introduce fuel efficient
recorders, v_ideo tape, light meter, stop watch and other electroo
bike or some other product. Therefore, R & D experiments may be
devices, then it is called as mechanical observation. Generally, wh
required so as to develop the fuel efficient prod_u~t. The new~y
the ~hysical ~resence of the observer/ researcher may not! developed product may be subject to trials before It IS launched m
possible, then, m such a situation, mechanical observation techni~
the market.
is adopted. Also, at times, even wit~ the presence of the obsem
· researcher, such mechanical devices may be used to suppleme An experiment should have two types of validity:
the observer's/ researcher's records and perception.
(i) Internal validihf When the effects of extraneous variables are
eliminated and when the observed effects of dependent variable
can be solely attributed to the exper~ented depe_n~ent variable,
then -the experiment is said to have mternal validity.

i
 Business Research Methods (S. Y.B .A,f
.s ·s
78 . . . When the results of the experilll ·· t\
(ii) External valid~ty. be extended to other settings an;nts · Data Collection and Processing 79
generailsed an .ca;nt is known to have external Val· ~is~
then, the experun . Id1ty_ Illustration of Field Experiments: Hawthorne Experiments
The Hawthorne experiments were conducted in four parts as
Advantages:
follows:
. 'd fi st hand information.
• It prov1 es r part I - Ill~mination Experiments (1924-27): From 1924-27, the
.d liable and relevant information.
• It prov1 es re .. National Research Council made a study in collaboration with
r may be in a pos~tion to develop new tAn},_. Western Electric to determine the effect of different levels of
• The researc he "'"•uli
methods. q' illumination on productivity of labour. The lighting conditions
for the test group of workers were varied, while it was kept_
constant for the control group' of workers.
Disadvantages:
• It may be ver'( expensive as lot of money may be required! The researchers found that the productivity of the test group
experimentation. improved with an improvement in light.i ng conditions, and
surprisingly it also .i mproved, when the lighting conditions were
• At times, lot of time and effort is required on the Part of worsened. To complicate further, the productivity of the control
researcher. 1 group also rose, even when there were no changes~ the lighting
conditions. The researchers concluded that something besides
• The delay in results may generate frustration on the Part lighting w~s influencing the workers' performance.
011
researcher.
Part II - Relay Assembly Test (Room Experiment) (1927-29): In
Types: this experiment, Elton Mayo and his associates were invol v~d.
In this experiment, two groups of six female telephone relay
• Field Experiments: assemblers each were placed in separate rooms and a number of
variables were altered- wages were increased; rest periods of
Field experiments are conducted in a real life situation in Whichi varying length were introduced; the workday and work week
researcher manipulates the independent variable to test were shortened. The researchers,·who now acted as supervisors,
1
hypothesis. These experiments are quite expensive and ~ also allmyed the groups to choose their own rest pe:riods and to
prolonged over a long pe!iod of time. They provide a high de~ have a say in other suggested changes.
of external validity i.e. the findings of field. experiments can l
extended to other settings. Field experiments are used to fu Productivity and morale of the group incre~sed consistently
solutions to practical social problems like: . during the period of experiment. The researchers also observed
that even when the improvements in physical working
Changing organization culture conditions were withdrawn, productiv:ity and IJlqrale of the
group were maintained. The researchers concluded that the
Improving the effectiveness of mass education socio-psychological factors such as small informal groups, non-
Testing new products in markets, etc. directive supervision, partic,ipation in work mat,ters, etc., were
responsible for improvement in productivity. ·
Part III - Mass Interviewing Program (1928-31): A large number
of workers were interviewed to find out the reasons for increased
 B1isi11ess Research Methods (S. Y.B,M.s. :St
M \
'I I vata Collection and Processing 81
I
!
. . This rogram indicated that P!oductivity co '
produchVI~- . 1 are allowed to talk freely about Olatt I.lid ~
Laboratory Experiments:
•
I increased, if peop
are of importance to them.
ers th.
·~ r,aboratory experiments are highly structured and conducted in a
setting specifically created for the research purpose. The resear~her
V _ Bank Wiring Observation (Room Experiment} (l h s a complete c~ntrol on extraneous variables and hence provides
I
I
Part~
..
. ent a group of 14 male employees in th 93~ h~ h degree of internal validity. However, these experiments may
In thzs expenm , b t· f
was placed under O serva 10n or 6
eb
Ula
h
1
~ be generalized to .real lif~ situations. L~boratory expe'riments
wmng room h f nt s
I ,
worker s P
ay was
group as a whole.
made dependent on t e per ormance •
of ~ : useful in physical sciences as well as social sciences.

ff ·
The researchers expected that the morf! e ic1ent workers Wou ,
. JIJ. Interview Method:
. t pressure on less efficient workers to complete thew ¼
it was found that the gro.u p established its tl Interview method of data collection involves face-to-face interaction
0
puwever
H0
standards of output and soe1al pressure_was used to achievet~ of the interviewer/ researcher with the respondents. The purpose
I • 0\v

of interview is to collect as much information as possible from the

standards of output.
respondents.
CONCLUSIONS OF THE HAWTHORNE EXPERIMENTS: "Interview is an effective informal non-verbal conservation initiated
for specific pu,y,ose and focused on certain planned content areas."
1. The social and psyc~10logical(actors ~-d not just monetary
-P.V. Young
incentives and physical work~g con~1hon~ are responsible
for workers' productivity and JOb sahsfact10n. "Interviewing is fundamentally a process of social interaction i.e.
exchange of thoughts, ideas, views, opinion~, and experiences. It is
2. The informal and interpersonal relations among workeri not the words spoken r,vhich matter but also the gestures, glances,
influence the workers' behaviour and performance more facial expressions, pauses, modulation of voice, ·rate of speech, etc.
than the formal relations in the organisation. which also matter a lot." - Goode and Hatt
3. Employees would perform better when they are given
Advantages: _
freedom to participate in decision-making affecting their
interests. The advantages of Interview Method are as follows :
4. Employees would also work more efficiently, when they 1. Reliability: The primary data collected through, interview
believe that the management is concerned of employees. method provides first hand information to the · researcher.
welfare. ' Therefore, this type of data is more accurate and reliable.
5. However, the accuracy and reliability of primary data largely
When employees are treated with respect and dignity, their depends on the quality of interview method. To get accurate
performance tends to improve significantly. ·
and reliable data, the field interviewers must be properly
6. Effecti~e communication between the superiors and trained.
subordmates can improve the r elations and overall 2. Detailed Information: Intewiew method can provide detailed
productivity of the subordinates.
information. The researcher can obtain in-depth information
by aski~g relevant questions. The i~terviewer may ask

;( \
 . ess Research Methods (S. Y.B. A..- S
Busm ,vi. .: St
82
. s depending upon the respo ~-11
dd
a 11·t·onal questton
. d l'lses ¾ Data Collection and Processing
83
information reqlllfe · . 4_ Respondent Bias: The respondents may not give proper
. othesis Formulation: The use of docu responses. They may withhold certain data or rriay provide fake
3. Helps
10
f>';
in formulation of research hypothesis. ~
sources he Ph more than one hypotheses in hand 'v rte~!
4 responses in respect of certain sensitive aspects such as salary/
income, investments, payment of taxes.
investigatort athse selection of correct hypothesis. ' Pti¾
data suppor s S. Interviewer Bias: There is a possibility of interviewer bias. The
4. Flexi~ilittyh:e1J.:'.~::::rv::;o:a:::s:~;~~:~;ili~I :;i:7i::;::s:~~~)r~=~te~~~;"'aire on his own, or he may
collecting th ' t' T lllorh
ti. ns depending upon e s1tua 10n. he inte .
the ques o . . d d. l'V1e¾ 6. Sampling Problems: It is difficult to select a proper sample,
ma delete or add certam questions epe~ ing Upon l
especially when the universe is large, and covers a large area .
. Yt.10n. The interviewer can also
s1tua . questions in a cer1,_
. ask
language with which the respondent 1s more familiar Wi~~ Types:

Personal Touch: Personal interview provides' persona} to , • Personal Interview:

S. in collecting the data. The interviewer can develop a g It is a face to face interaction between the interviewer and the
rapport with the resp~ndent. The p~rsonal touch can en: respondent. The interviewer may ask the questions, and the
the interviewer to obtam frank responses. respondent accordingly responds. There are various types of
6. Helps to Overc~me Resistance: Survey /Interview Illeth personal interviews:
helps to overcome res~stance from the respondents to respo~ Formal and Informal - In formal interview certain formalities
favourably. The interviewer can assure the respondent that his are followed such as prior appointment of the interviewee. In
her identity may not be revealed, if so desired. The interview, informal interview, the interviewer may not follow formalities
can use persuasion skills to collect.relevant information. in data collection.
Disadvantages:
Structured and Unstructured - In structured interview, the
The disadvantages of interview method are as follows: interviewer asks certain questions designed well in advance.
In unstructured interview, no specific questionnaire is followed.
1. Time Consuming: The interview method. is time consumini The interviewer is given freedom to ask certain questions
as lot of time is required to interview the respondents and colle: depending on the sftuation. Structured questionnaire is
the data. normally not followed.

2. Expensive: The interview method is expensive as there is an~ Individual and Group - In_individual interview, only one person
to appoint interviewer. Salary and perks to be paid to the staf is interviewed at a time, and in a group interview, a group of
However, in the case of secondary data, there is no need fo respondents may be asked questions at the same time and at
interview. the same place.

3. Paper Work: There is lot of paper work involved, because th Ge;neral or Specific - In general interview, the questions asked
interview requires questionnaires to be filled in. Also therei may be general in nature. For example, gener~ ~?~ent~ on
good deal of paper work in respect of processing of data. advertising media. flowever, in the case of sp~cific interview,
specific questions may be asked such as specific comments on
T.V. advertising.
 Business Research Methods (S. Y.B,M.s . ·
84 . . . ··S£~
. . and Non-directive - In_ direc_tive intervie ·It D ta Collection and Processing 85
Directive . directions to the mterv1ewee to r \\>, tl a . .
interviewer gives non-directive interview' the t espo~d , of the universe is selected to collect the data. But the sample sh au Id
certain
. questions.
f edomInto respon d to certam. top1c
. Witho
. esPori~~, I be 50 selected that 1t represents the universe.
are given re . . Ut ln h• .
q O
·•,~~ ''Survey researc ts
. .
uestiorung n the part of the mterv1ewer. t ant organized
,f ·
attempt
• •
to analyse' interpret and
• . •
Focused Group Interview. freport the presen
.L. Whitney
5
atus o, social tnstztution group or areas." -

hi t pe of interview, the researcher focuses on the exper· Note: For the advantages and disadvantages ofsurvey method, refer
Intht s
of e grouy p ofrespondents · and its possible . d
effects. Therese ien~
arch to tlte explanation of advantages and disadvantages of 'I Interview
• wer structures the interview well 111. a Vance but has 1er1
.mterv1e t Method.
flexibility in asking the sequen~~ of questions and ~ecidin ,, .
r Of ask ing the same A definite framework of topics is d g Types. •
manne . · . ecid ,
in advance. This type of mterv1ew helps to develop the hypoth Telephone Survey;
1 ting to specific experiences of the respondents. The effecttv •
re a . d d enes; h d f d . .
of the conducted focused group inteTV1ew epen s upon: · It is a met a a con uctmg mterview by telephoning the
Line of thought of the interviewer respondents. A series of questions are asked on phone and the
answers are recorded. It is a very popular method extensively used
Direction of interview resorted by the interviewer in western countries. ·
• In-depth Interview: • Mail Survey:

· h h · d t ·1 I . It is another method of data collection. A questionnaire is prepared

It is a semi-structured approac. w erem
·tud e ai s . ref10ating to tht con ta;..,;..,g a list of questions to solicit information from selected
Ulll•

respondents' opinions, emotions, ath es, conv,c ns, etc. '1! ondents This questionnaire is sent through post or advertised
sought. It is designed intensively to examine the pei:sona! _feelin~ ;~s~ newsp~pOr or magazine, explaining the purpose of the
and emotions of the respondents. The pmpose of the.mterv1ewisa estionnaire and a request to complete and return it to the
get a true picture of the respondent through deep probing into hi ;:Searcher. A rep Iy-paid envelope may also be given to the
mind. This type of interview requir~s a ~ature understanding « respondent wherever possible to encourage the respondents to
human behaviour on the part of the interviewer/ researcher. promptly send the answers.

IV. Survey Method:

• E-mail Survey:

The survey can be census survey or sample survey. In case of censw When the surveys are e-mailed to the potential respondent.s to get
survey, the entire universe is contacted to collect the data. responses, then, it i~ called as e-mail survey. Not inan_y researchers
use e-mail survey as a method of data collection. Email su~ey may
For example, a survey on job satisfaction of doctors in Governme,1 useanthis
be easy methoddue
technique to cqllect data. However,
to the following it is not advisable to
reasons:
hospitals in the city of Mumbai, then each and every doctor in Govt
hospitals in Mumbai needs to be contacted to collect the data. Bui - Every potential respondent may no t be USing e-mai1 or a
because of limitations of time, and money, it may not be possible to computer.
contact each and every doctor, and therefore, the researcher may
adopt sample survey method. In sample survey method, only a part
 Business Research Methods (S. Y.B M
86 . .s 'St-
/i I d t ay not be the true representaf10 ·· c~
The res~on en m . · n of the 1I vata Collection and Processing 87
PopulatiOn.
, . · 1 val ues are link ed to structured data in a database so the
nurnenca
The responses may not be reliable. data can be analysed with traditional data mining techniques.

• Internet Survey: 'fhe top five free social media listening tools are:

. temet survey
An ffi
is a questionnaire that the target • Hootsuite
respond over the internet. Internet surveys a~e usually er 'lll • TweetReach
web forms with a database to store and retrieve the eated 1
statistical software tools are provided for data an:~~ers.1• J<lout
researcher can gather the responses as soon as the re Ysis. Social Mention
. spond •
submit the web forms. Internet surveys can be incorporated er,
website of the company. A pop-up survey can also be i-.... 0n~• Twazzup
b. ....eatect
the company we site. Social media·listening can benefit the organization in the following _
The advantages of internet surveys are: ways:
Effective brand management
It can involve larger sample at a lower cost. • )

Evaluation of marketing campaigns accurately

The information can be easily and quickly collected thr •
internet surveys. · ou~ • Discover market gaps

It is a convenient method of data collection. Respond to positive and negative posts

•
Generally, the response rate of internet surveys is not favour bl· • Analysis of frequent conservation topics
The researcher may send reminder mails to the potential respondae":
,~, 1 · Assessment of competitors
· , strategies
t?
to respond to the web f~rms. ~so, increase the response rate, 1/r
researcher may offer mcent1ves like cash prize or some oth • Interaction with customers through social media platform.
incentives. · e

• ~ ~QuE_
s~~oNN:AI~E,-- . ·
l ~ANING WO IMPORTANCE
Social Media and Media Listening:

So~ial media_ li~te~ing ~s al;<? ~a)~e.d as soc~~l ~<:dia ~onitoring an,

,J
active_ we~ hste~mg. It is a process of identifying and evaluatin1
~hat is bemg said of a person, company, product and/ or brand or A questionnaire is a set of questions, which act as an instrument to
1..f!.ternet. · • collect data from the respondents to a survey or interview.

In l~~e organizations, social media liste$g tools can mine textf~ •

specific key words on s_ocial ne~orking websites, blogs, ? .iscus'~ior. Importance.
forums and other social media. Monitoring software transpos~
specific keywords in unstructured data into numerical values. These The importance of questionnaire is stated as follows:
,. 1. Rele~ant Data: The researcher can collect relevant data with
 . sResenrch Methods (S. Y.8.AA
. Busmes . ,v,. 5.:St .
88 . ire. The questions can be d ~-/1 Collection and Processing 89
el of quest10nna . esi vata ,
:;:tfic~lly to collect the r1ght data. LIMITATIONS OF QUESTIQNNAIRE -
. to the Respondents: The re~pondents c .
2. Convenience . e as per therr • own convenience, i.e., Wh
cl.nfil!A FOR REFERENCE
questionnair t convenient time. Normally, there is ha }n ij
ondents ge · h r~h,
resp the respondents to give t e response , Suitability : This method is useful only among literate
time pressure on s. 1.
respondents who have adequate educational background, who
d Unstructured Responses: The resear h
3· Structure d an ·
t ed and unstructured questions. In the
c er can read and understand the questions properly.
ask struc ur d h case
d r esponses the respon en t as to se} 1 Rate of response : Generally, the rate of response in mail
structure ' . h . ect 2•
• te answer given along with t e question. In I questionnaire is very low. Unless incentives are offered, people
appropna d . fr case
unstructured respons~~, _the respon ent 1s ee to express ~ may not co-operate with the researcher to answer the
her own feelings or op1ruons. questionnaire.
Quality of data: The data collected through questionnaire ma_y
4 Qualitative and Quantitative_ Data: The re~earcher can coj~3.
· the data both in terms of quality and quantity. The qualita~ not be reliable and it needs critical evah.,\ation before use. It 1s
data can be obtained with the help of unstructured questio because the respondent might have answered the questions
The quantity of.data is possible as large number of responde~ without properly understanding it and sometimes is filled by
can be surveyed. some one else on behalf of the actual respondent.
Rigid and restricted response : In questionnaires, the
s. Large Coverage: The questionnaire e~ables the researcheq4· respondent is instructed to give only those information which
cover a large area. With the help of mail survey, the researc~
is asked for. Due to this rigidity, the respondent can not share
can collect the data from respondents located at distant plac~ or supply additional information.
6. Proper Processing: The data can be easily processed. Evens. Attitude of respondents : Usually, the respondents have a
question can be given a code number, and the respons~ casual attitude and lack the required sense of responsibility in
especially the structured ones, can be easily tabulated aru filling up the questionnaires.
apalyzed.
6. Ability of respondents: This method is largely depending upon
7. Easy to Alter: The questionnaire can be easily altered, as !>I the ability of the respondent in expressing his views or
the needs of the researcher. If certain questions are arnbiguo~ experiences. However, very few people who can e~p~ess
or irrelevant, then such questions can be modified or drop~ themselves in a written for~ can sqccessfully offer the reqmred
from the questionnaire. data.
8. Sensitive Information: With the help of questionnaire, 1~7. Personal touch: Questionnaire is inappropriate tool where the
researcher can. collect certain sensitive information from th objectives of the survey requires a good deal ofhplanation to
~espon~ents. The respondents may provide sens itiv1 the respondents.
information without divulging their identity.
- 8. Verification of facts: Answers to the questionnaire cannot be
checked again and hence be accepted as final.
 Business Research Methods (S. Y.B_,..,
,vi.s.,·st
90 • ·~ ,..-•.. . . - ,\l .
/1 I - ,.. . . DESIGNING QUESTIONN . Data Collectio11 and Processz,;ig 91
I

I ··STAGES IN . 6_ Decision on sequence of questions: Sequence of questions is

. e the steps 1·n the preparat·ion Of ' the order m which
. the respondent
. receives. the. questions.
. The
The following ar a. first or openmg question should be obiective, Simple and
uestionnaire: t, interesting. It should build con/id ence and i,n duce the
q st ding of the research problem and , respondents to answer easily. The d ifficu It and em b'arrassin g
0
t. Properunder
fr . an . . the resea re her sho hJectj
the questionnaire, I 1 • questions
d I or personal
d f th and .controversial
. questions should. be

r;o
I

Before ammg h bl nd U d~ aske at t 1e en o e questionnaire. The order of the questions

good knowledge of the researcd ;m a the obi•:/ should be logical to the respondent. Sudden changes in the
This will help the researcher to ra t re evant questions, \ subject or topic may be ~voided. ,
K the target respondents: The researcher shouJct h
2. now d H al' •
good knowledge of the target resfpfon il~nt~. e should ~ 7 Decision on physical characteristics of questionnaire: The
the target respondents i~ terms o am 1anty of the langu physical form of the questionnaire should be easy to handle
familiarity with techn~cal ~ords~ stat~~'. et~. This \Vo: and control. It should be convenient to the researcher as well
facilitate to conduct the mterv1ews ma facilitating znanner.' as the respondents. Before finalising the la yout of the
questionnaire the following matters are to be considered.
3 Decision on number of questions: The researcher shou)ct ,
· a proper decision on the number of questions. As far as Pas ~ (a) Size : The size of the questionnaire depends upon the
number and type of questions, information asked for etc.
the number of questions must be restricted, so tha~,
If the size is small, it is easy to carry and handle. It also
respondent is motivated to give proper responses. In casit helps in sorting out and analysis.
lengthy questionnaire, the_respondent may not reply propt
for tiine and other constraints. (b) Quality and colour of paper: Questions should be printed
on good quality white or light c·otoured paper. The name
4. Decision on type of questions: The researcher has to select &a of the project and researcher should appear on the front
the three types of questions for soliciting information dependr side of the questionnaire.
upon the objectives of the research and the type of respondem 8.
The type of questions are (a) open ended questions (b) multi~ Approval of questionnaire: At times, there maybe a need for
choice questions•and (c) dichotomous questions. Open endr approval of questionnaire from higher authorities. For instance,
questions allow the-respondents to answer the questionsintt in the case of academic research; the questio1mai.re needs to be
way they like. Multiple choice questions provide a lisl1 approved by the research guide.
pro?able ans~ers and the-responqent has to tick the appropru 9· Pilot study to pretest questionnaire: It is an essential and useful
ch~1ce. The d1Qh0tomo~s -question is a two way question, step in preparing and finalising the questionnaire. Pre-testing
which answering is very easy.
involves rewriting questions, changing their order, replacing
5. open ended questions to close ended questions, adding or
Decision ·on the wording: The wording of questions shoul
en~ure ~asy understanding and responding. It can be done~ replacing certain questions, etc. Pilot study involves
usmg simple and common words which are familiar to , administering the rough draft of the questionnaires on a small
respondents avo1·d · b' d L' group of responde,nts who would be interviewed subsequently
m . . ' .. using iase •words, avoid using dou1, in sample survey. The purpo_se of pilot study is to find out the
eanmg questions and ambiguous questions.
clerical and technical errors from the questionnaire before its
 . Research Methods (S. Y.B.M.s .
Busr 11e:,s
>
•. St,\ .
re removed before appr 1
1. Co!lectio11 and Processing 93
92 The defects a . h d t II oval r Dflfa
finaJ~electii~~othatfuturelapsesm t e a aco ectionpr~ TYPES OF QUESTIONS
uest10nna ·
q be avoided. .
can . -testing (pilot study), the q
F . al Draft: After the pdreby word and topic by topu_est~ 1. Open Ended Questions ·· It is also known as free answer or ·
10. in • ed wor . d. . tc 1 free responses questions in which the respondent _can a~sw~r
re re-exam in h the coding an mstructions p . I
a ill go throug . rov~ . his own words and the way which he likes. If the quest10n is
researcherw ce ao-ain to ensure Its clarit}' -~
ondents on o . . . an !;1What is the reason for underdevelopment of India?" The
to the resp the 9.uestionnaire IS given for p . a
. is accurate,
every th m.o0
l'tn~
I respondent
. can write
r
any reason relating to it.
Merits:

ESSENT IA
LS OF GOOD QUESTIONNAIRE
· a) Open ended questions are suitable at the start of quest~on,
.- which are used at the early stages to obtain general reaction .
. • · are commonly used for collection
. . of prirnan,
--, d·
Questionnair~s . f ood questionnaire are given below. b) It provides a variety of responses and useful in exploratory
The charactenstICs o a g . research in identifying new ideas.
Q tions· It should cont~in those questions w~
1. Relevla nt tute; the t~pic and avoid unwanted and irretel'> c) The respondent is not influenced by multipl_e choice
are re evan ti , questions and therefore less possibility of respondent being
questions. Avoid too many ques ons. biased.
2 Clarity: The questions should be easy and sim~le to undersl¾ d) It is popular in unstructured interviews in which an
· Long and co mplex questions should
. be. av01ded.
h f The
. . w01,. informal method of enquiry is necessary.
used should be simple and common d wh1c are. b' am1har to f
respondents. The questions shoul not conta~ 1ased ~011 • Demerits:
and leading questions. Words of double meanmg be avoid~
a) It is not suitable for self administered tljpe of questionnaire.
3. Type of Question:_Questionnaire may contain open-ended1
well as multiple choice questions. b) It does not provide standard responses. Hence it is difficult
for processing and interpretation.
4. Logical Sequence: There should be proper sequence in askr
questions and it should be arranged logically. c) There is a possibility of inaccurate recording as the
researcher may record only a part of the answer which is
5. Physical Appearance: The physical layout of the questionnai; convenient and suitable to him.
the quality, size and colour of paper and printing etc. be prof.
d) It is difficult to summarise or code the free answer given
6. Pretesting of Questionnaire: The questionnaire should ! by the respondent.
pretested before it is finalised.
e) As compared to structured questions, open ended
questions are more time consuming and costly.
2. Close Ended Questions : In close ended questions, both
questions and answers are structured. The respondent is not
allowed to answer in the way he likes. He has to select one
 Business Research Methods (S .Y.B. A., S
94
'VJ, • s
f lternative answers given t ·· €~.,
wer from the set o a d d esti·ons o hitn h oataCollection and Processing 95
an S es of close en e qu . . ·,~
are many ~IP Choice questions: These are accon,p . fYPES/ CLASSIFICATION OF QUESTiONNAIRE
) Multtp e d th ¾1eq
a number of specific answ:rs an e respon~ent is ~, - FOR REFERENCE
emostappropnateanswerasperhisp ast ,_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
1 tth
to se ec erce~ . . . . . .
E Your.house is owned by (1) Self (2) Parents (3) I{ Aquesho~aire is an md1:~ct and unpersonal method of conducting
(4).g.Ot hers (SPecify) e]alj~ survey. It is a set of questions arranged in an orderly form and sent
d t f bt · · ·
. . to the res po~ en s or O am~g information on a particular matter.
Your occupation IS (1) D?ctor (2) ~ngmeer (3) lawye Questionnaires ~e used as a duect method in conducting interview
T cher (5) Others (Specify) r1 for data collection. In case the respondent is at a distance the·
ea uestions are smta . bl e for so 1·ic1.ti.ng inforrnati
. questionnaires
. . can be mailed to him. Questionnaires are clas;ified
Such q . mW -
it saves time in processmg. , . . .
. st . . . 1. Structured questionnaue: It 1s the type of questionnaire which
b) Dichotomous or Two way Que wns · These are tw0 is segmented and designed to elicit detailed and accurate
questions which offer only two alternative answers ½l information under the given titles and subtitles. It helps to
as male/ female~ cash/ creilit, married/ single etc, \ ' obtain accurate res~onse and _apt information required by the
questions are qmck and easy t~ answ.er. It he!ps fastedi~ researcher. It helps m processing the data as the questions are
arid tabulation and free from mterv1ewer biasness. well structured. .

3. Semi-Structured Questions: A guestionnaire can be structUJi

semi-structured or unstructured. A structured questionnaire~
Accord:ghto P. i. Yo~~~'. "~ structured questionnaire is the
one w tc con ams e, tntte, concrete and preordinated
predetermined and definite questions and the wordings ~ . questions, with additional questions limited to those necessary
order of questions are same to all respondents. It can have do~ to classifJ! inadequate answers or to elicit a more detailed
(the respondent gives yes or no) or open-ended (given freed~ response.
to answer in his own way) questions. Such type of questionn~ Answers to such questionnaires are usually given with high
is known as semi-structured questionnaire. precision without any vagueness or ambiguity. For example,
4 Filtered Questions : Such questions are induded in U1 questions on age, marital status, number of children etc. are
· questionnaire which have many sub- questions with one m~ structured. The structured questionnaire is again classified into:
question. If the answer to one question is yes, then tt1 a) Close form questionnaire: It is the one in which questions
respondent has t9 answer the subsequent questions coverir, are set in a manner that leaves only a few alternative
the specific aspect of the answer.
answers. For example, Yes or No" . The respondent has to
11

E.g.: Do you smoke? Yes/No. choose one from the given options.

Those who reply in affirmative would be asked to specifylt 1

b) Open ended questionnaire: Here the respondent has full
sub questions Cigarettes/Pipe/Bi.dis. · liberty to choose his own style, diction of lang_uage
expression, length and perception. The respondent 1s not
How much per day, etc. If the answer is No, the responde: restricted in his replies.

y
can filter out the sub-questions to go to the next questions.
 BUSII•·zess Research Methods (S. y. B•M.S ·s
(1

I
96
2 Unstructured Ques:~~~~:,
• is
ire : It a simpie .. tA~ .
subdivision. It is us~:Slio~ oata Collection and Processing
· without any s~gn:enwhich simple questions are a for sit _ features of Secondary Data
. 97

Of studies, m . h. rran :1
types tains a set of questions w 1ch are not 8e~,
I sequence. It co~ h nee gives sufficient scope for s~, Recorded Data: Secondary data is collected from recorded and
1
.in advance _an dernainly forconductingintervie: Vati~ · published/unpublished paper source. The data has already
answers. It_is use ore information may be cones. 8~ce1 been colle~ted by somebody for some purpose other than the
more flexible, m ctect frt project on hand.
respondents.
.. d Q tionnaire: These are the questionnaire . 2. Easy to Collect: The data is readily .~vailable and hence easy to
3· Codif1e t dues . . d sin \l'l
answers are given m co e numbers. Aga· 1 1 collect in the most econ<?mical way. The researcher needs to
the expec e d 1 . nst put less efforts to collect secondary data as compared to primary
ti 0 four or five expecte answers are given a d
ques ~ ' ·ven a code number. The respondent has tn .~ data.
answer 1s g1 . . o ticl
circle the selected number only. It is convdement for PtOces~ 3. Quantitative Data: Secondary data is called quantitative data.
an d eas il y Und erstandable to the respon ent. ·· It can be used -for substantiating the primary data.

4 Uncodified Questio~nair~ : These ar~ simple ~uestiollnat 4.. Comprehensive: Secondary data is available on every subject
· 'thout any codificat10n. Simple and drrect questions are and even in those subjects where collection of primary data is
Wl '1 difficult. It can supply volumes of large data.
in this.
Sources: Secondary data may be collected from primary source
5_ Pictorial Questionnaire: In such fqu~stionnaires, the seleti 5.
alternative answers in the form o pictures are given andi or secondary source. For example, the government conducts
respondent is required to ti~k. the picture concerned. ~ primary census surv~y and its report is called census report.
questionnaires are useful for illiterate and less knowledg~ Newspapers and magazines may publish a part of this report
which is called secondary source. The researcher may go
respondents.
through the census report which is the primary source.

6 Relevance: Secondary data is a readymade information which

SECONDARY SOURCES OF DATA j . is meant for general purpdse. The researcher must"evaluate the
secondary data with reference to its accuracy, relevancy and
The secondary data is readily available data from publisheo . reliability.
printed sources. The secondary data is generally used in the cax 7. Availability: Secondary data is available internally (from the
academic research and to a certain extent in the case of soo company's own records, accounts, etc.) or externally ,(books,
research. However, commercial research requires more of prillll magazines, p~blished reports, etc.).
data as compared to secondary d_ata.
-8. Less Time Consuming and Less Expensive: Secondary data
Generally the researcher first makes an attempt to obtain informan requires less time to collect the data as compared to·primary as
t_ from secondary sources to solve the problem. However, when t
secondary data is insufficient and outdated I the researcher res~
it is readily.available. Secondly, it is less expensiv~ be~au~e there
is rto need to conduct surveys or other methods of pnmary data
•
to prnnary data.
collection.

z
 Bl/ S1·,1ec;, Research Methods (S. Y.B_A .. S
•• 'VJ , ·s

98 .
O
fS condary Data ·· f:1½I oata Collcctio11 and Processing 99

Significance e Limitations of Secondary Data Collection Method:

- Of condary data sources are as f -- . . . .
01
The main ad vantages se · . lo;,,, There are certam lnmtations of secondary data collection methods
. . The cost of collecting
Expensive. Th secondary
. data is. su ch as follows:
1. Less th that of primary data. . ere IS no neect to Ill~
lower an t· naire and there Is n_o need of fielct Pre~ 1. Problem of Accuracy: The quality of secondary data is affected
an d print ques 10n ' . staj/.
due to the problem of accuracy. One cannot be certain of the
. C nsuming: The secondary data can be co])
2. Less Time o ompared to primary data. This is becau eci~ genuineness of data. For instance, the statistical data provided
a faster rate as c . se;tt on the internet from v~rious sources is vastly different. Also,
.1s no nee d to conduct interviews/ surveys. •~ the quality of internal secondary data may be exaggerated or
biased. The records may not be well maintained or organised.
Less Processing of Data: The secondary data m~y require
Therefore, one must be cautious in using secondary data.
3. processmg . of data·' as a good amount of data 1s availahi~ e
·..
processed form. 2 Problem of Reliability: The secondary data lacks reliability.
The published data may be outdated, and therefore, it may not
4 Quick Decisions: Second~ry -da~a may permit qui~k dec~io serve the purpose of the current research work undertaken by
· making; as the data is readlly available and that too m pr°"'i the researcher. Also, the reliability of the source that provided
form . the published data may not be genuine. At times, some sources
provide unreli1;1ble and biased data.
I· 5. Supplements Primary Data: Tbhe s:con~aryh data 1111 3
supplement primary data. If ~eed e, t e rese_arc er may~ · Problem of Adequacy: At times, the secondary data may be
secondary data along with pnmary data to arnve at decisiOil accurate and reliable, but the data may be insufficient to solve
the current research problem. The secondary data may not
6. Less Paper Work: Secondary data involves relatively less pap provide complete data to solve research problem. A researcher
work, as the data are readily available from pu blishe~ sour~ cannot completely depend on secondary data to solve the
7. Large Volume of Data: The researcher can obtain large volun research problem. Therefore, the researcher needs to collect the
data from primary sources as well to solve the research proble~.
of data from various publication sources across the world.n
volume of data is so large, that even a large -individual resear: 4· Lack of In-depth Information: The secondary data not o~y
firm may find it impossible to obtain such large volume of da: may lack adequacy, but it may not provide in-depth ~ormation ,
The large volume of data may enable the researcher to ~, to solve the research problem. For instance; commercial research
solutions to the research problem especially when sut requires in-depth knowledge of consumer ~eh~viour which can
problems or situations have taken place earlier in some olti be obtained more through in-depth queshorung of the target
organisations or places. respondents.
8.
No Sampling Errors: There is no question of sampling errcr 5. Problem in Quality Decision Makin? : The_ s~condary ~ata
in secondary data. This is because; there is no need for selec~ may affect the quality of decis~o~-makmg. Th1_s is because, the
sample of respondents. In the case of primary data collecti: data may be inaccurate, insuH1c1ent and unreliable. Therefore,
there is a need for sampling and therefore, sampling errors am the decisions made purely on the basis of secondary ~a.ta would
However, it is to be noted that the published secondary dt bring poor outcomes. To improve the quality of dec1s1ons, one
may be subjected to sampling errors.
P'i 100
Business Research Methods (S. Y.B,M.s.
II . rimary data as it is first hand info ·· 8t~J
may requrre P ·
eliable and accurate.
r111aij
o~. oata Collection and Process_ing
more r I fhe internal secondary data is collected fr . t 101
6 Problem of Spee ic
if . Data: The secondary
.. data lllay k as: om m ema 1records such
- g ··1 . nature rather than specific. The resea h~ l'I .
enera m 'f' bl Thererefer ~p,·\~ • Purchases and sales records.
spec1·t·1c data to solve spec1 1c pro ems. d
1 secon ary data b 0re,H
researche r cannot depend mere Yon h 'f• · , lW • Debtors and creditors records.
collect specific data to resolve t e spec1 1c problelll.
• Division/Unit wise production or operations records.
7 Unsuitability: The secondary data may not be Suitable il1 c
· cases. Secondary data may be of less use in th • Financial statements including balance sheets, profit and loss
1 b . I
commercial research. To so ve usmess re ated pr hi e, e cas statements, and fund/ cash flow statements.
. f . o ellts
researcher may require more o pnmary data rathe , •
secondary data. For example: to so~ve the problem of 1~ gr: Employees'. records including performance, compensation,
claimsr service records, etc.
customer complaints, one need to find out from the cu
.
with the help of a survey me thod or interview . methods1olllt USES OF INTERNAL SOURCES:

than solving the problem with the help of secondary da::.~

The data available from the company's internal source can be

B. :::~~:;~:::~~:1.::::t:n~as Th
i~:~nr:s!~i!~;h:f:: 1.
of the secondary ~ata. The secondary data may be ba~
effectively used for the following purposes :

Product Analysis: An analysis of sales record according to size,

varieties, or specific, individual products or groups can indicate:
influenced by t~e bias of the respondent-s and also that of ~
researcher. At times, the researcher may have casually coll!!ct[ i) the percentage of sales which each product contributes to
the secondary data and got it published. Therefore, one sho~ the company's total sales and profit.
not blindly d~pend on secqndary data.
ii) increase or decrease in sale of each product.
, .- 2. Customer Analysis : It includes list of wholesalers, retailers,
MET~ODS OF .COLLECTION .OF distributors and dealers. Such analysis is useful to (a) disclose
·,
the volume of sale of each product, (b) measure the profitability,
SECONDARY DATA -· FOR REFERENCE (c) measure the changes in the trend which may occur in retail
trade.
The sources of secondary data collection can be internal sourceso 3.
external sources. Analysis of Sales Territories : A study of sales according to
sales area can furnish the changes in the volume of sales
according to each product and between two or more ·~erri_t~ri~s.
I. Internal Sources: This helps to evaluate salesmen's performance, profitabiltt~ m
each area, i_dentify strong and weak areas for marketing
An organisation may use certain data from its own records to t~ purpose.
certain decisions. For example, promotion of employees can be ]Jas(>,
4. Sales Frequency Analysis : It is useful to establish _seasonal
on the perfo:m~ce appraisal reports. Also, the past perform~ and cyclical fluctuations in sales (For example, sale of ice cream
of the organisation can be used for planning th~ operations.
and soft drinks is seasonal).

it
 Business Research Methods (S. Y.B .•~.s.s
A.,

'• t\
102 1 . . Analysing the balance-sheet
Financial Ana r•;•
·an provide information "about <lliQ , Data Collection and Processing 103
5. financial statemen / casteful expenditure. Pron~, (iv) Statistics on Foreign Trade - Brought out by The Director
or identify areas o w , '
General of Commercial Intelligence - gives r:nonthly
A nalysis• Maintaining a minimum statistical data on imports and exports.
k Turnover · d d . st 1.
6. Sto~
r~
1 t d to both sales an pro uction. A.n 0cfi ~
is drrectly ale to find out capital requirerne'n ts ari~1h~
stock Ie~e ·te be purchased at one time. ' ll1111i~
(v) Price Statistics - The wholesale price index for
commodities like food grains, fuel, power etc., published
order quanh y
by the Ministry of Commerce and Industry, Government
of India.
II. External Sources:
(vi) (a) Other Publications : It includes official publications
. f
can collect data from external secondary of Planning Commission, Rese_rve Bank of India,
The orgamslase1~:ndary data can be collected from .various Pusbo~rQ
The externa Ministry of Finance, Ministry of Food and Agriculture,
Indian Labour Bureau etc.
sources sue has books' magazines,dnewspapers,b mternet
d and o~
repor ts an d records. The external .ata .can e ~se
. . to take var1
.~ (b) The publications of National Sample Suryey (NSS)
decisions such as prod_u ct designing, pricing, proiiloti(, on issues relating to social, economic, industrial and
distribution, etc. other statistics.

The following are the various external sources of seeonctary da\ (c) Publications of Commodity Boards, Expo~t
Promotion Councils, etc.
1. Govern.ment Publications : Central, State and loc _
government bodies produce and publish a vast amount of di 2 · Syndicated Sources : Some private organisations _like
that can be used for marketing research purpose. Some ofi advertising agencies, research firms, etc. as a part of their
important government public~tions are : business regularly collect data useful to marketing re?earchers.
Such collected, analysed ~d tabulated data are sent periodically
. (i) Census Report : The Census of India's population to the subscribers. Sometimes the data are made available
conducted every 10 years at national level by the Regis1n against payment of fees to non-members. Further, such
General of India. The data provides basic demogra~ organisations do undertake data collection projects on contract
information about the population by sex, age, educati~ basis. Syndicated services usually collect data :- on consumer
etc. goods market, TV viewing, magazine readership, me~ia
advertising, etc. In India, organisations like - The Operation
(ii) Statistics of National Income : Published every yead ·Research Group (ORG), The India Market Research _Bure~u
The Central Statistical Organisation. It supplies vit (IMRB), Pathfinders etc., provide inf?rmation on retail ~u~it,
statistics on agriculture, industry, trade, savin( . crop surveys, media and readership research, telev1s10n
consumption, capital formation and so on. . viewership, Rural and Urban Market Index, etc.

(iii) Industrial Statistics: Publish~d by The Central Statistia 3. General Publications: It includes publications like new;p;pe~s;
Organisation - provides industry wise information I~ magazines, trade and professional jourpa 1s , r~ e
workers employed, total production of.selected industri~ Associations and Chamber of Commerce. L1k_ew1se,
etc.
professional directories published by pr?fessional bodies e.g.,
The Industrial Directory of India, Directory ~f ~orld Cha~ber
of Commerce and Trade Associations, pubhcat10ns of Umted

,.(_ i
 104 Business Research Methods (S. Y.B !vi
. . . . -S.: s
.
Nat10ns, specialised hbranes, research .pr 01•-=c t s t~)
0

. . t can be tapped for data collection Pu ' tefe

hbranes e c. 'Pos~. \ 1
Data Collection and Processing
7. Paper Work: 105
PRIMARY DATA VS. SECONDAn_-y b, Primary data involve a lot of paper
work in respect of questionnaire, Secondary data involves relatively
' ~
tabulation and analysis of data. less paper work, as the data are
.----P-ru~M~A~R~Y~D~A~TA~-,-~S~EC~O~N~DARvn ' , readily available from published
1. Meaning: A:ti\ 8. Specific/General: sources.
Data collected first hand is Primary data are more specific
The secondary data is
called primary data. . available ColJecteq in nature. Such data are collected !he secondary data may be general
from readily to solve a specific problem. m nature. The researcher has to
such as publications. sources
analyse the secondary data, and
2. Methods/Sources:
select only that data which are
The primary data are collected
throuoh surveys/ interviews,
The secondary data is c II 9. In-Depth Infonnation: relevant to his research work.
0 . from internal sources suo hected Through personal interviews,
experimentation and observation. , c as
company s records, and frolll the researcher can collect In secondary data collection, it
external sources such as ne in-depth data. The researcher may be difficult for the researcher
.
papers, magazines, govern Ws. can ask for additional questions to get in-depth data. The researcher
rnent
. d ustr re orts, etc to get detailed information. has to solely depend upon the
recor d s, m readily available data.
3. Accuracy: '10. Need for Sampling:
The primary data are more There is a need for sampling in
accurate as such data are collected The secondary data may not be the case of primary data. It is not ln secondary data collection, there
first hand by the researcher or accurate; as such data are is no question of sampling, as the
possible to collect data from the
field interviewers. · collected from readily available entire universe. researcher does not collect data
published sources. by way of surveys or interviews.
4. Reliability: 11. Respondent Bias:
The primary data are more reliable In primary data, there is a possibi!~ty In the case of secondary data,
as they are collected first hand The secondary data may not be
of respondent bias. The respondents there is no question of respondent
by the researcher. reliable, as such data are collect~ may provide biased responses bias, as no interviews or surveys
from readily available sources, depending Uj>_On the situation. are conducted.
5. Time Factor: which may or ma not be accura:, 12. Quality of Decision-Making:
To collect primary data, the With the help of primary data, the
quality of decision making may be The quality of decision making
researcher has to spend lot of The secondary data can be collect~: may be affected in the case of
time and effort. from internal and external sourc~ enhanced as the data collected is secondary sources, especially if the
6. Cost Factor: at a comparatively quicker pace. up-to-date and first hand. data are outdated and inaccurate.
To collect primary data, the
researcher has to incur more The cost of collecting secondary
expenses such as preparation data is comparatively lower as .-SCALING -_ CONCEPT AND TECHNIQUES
qnd printing of questionnaire, there is no need to prepare and
training to the interviewers, etc. print questionnaire, and there A scale is a continuum containing a set of statements l?gically r~lat~d
.is no need of field staff.
to the attitude under study. It extends from the highheS t po~tt~
terms of characteristic and the lowest porn
· t in terms of c aractens 1c.
 Business Research Methods (S. Y.B_,.,.
'st
1v1.s ..

I 106 d' te points between the two ext i\1.1

interme ia . . rem
. f the charactenstICs of: homogeneit ePo· Data Collection and Processing
I
and there are 107
The scale must sat1~ _Y Y, line'\
validity and reliability. . ~tr Jie intermediate P~int~. Each point on the scale is assigned a
urnerical value, which is generally not printed in the questionnaire.
. 111 •n-ni
. valves ass11,1 • ·ng numerical values to variou s de ~he numerical value with each possible answer - i.e. 5 indicates
Scaling • d ali' ty and other abstract things. TI-- 8re ~
· · att1tu e, qu ' ,ere ·1 trong agreement with the attitude and 1 indicates strong
opuuon, ti opinions of respondents as Well as I'll searcl ~isagreement with the attitude.
tries to assess ,e easlJ ,
f .
intensity o it. reH
Ukert scale is a popular scaling technique in social science research
The two main techniques of scaling are: due to the following reasons:

It is simple and easy to construct.

1. LIKERT SCALE: •
. type of summated scale developed by Rensis L·k
It is a . f t en
• It is more reliable as it provides precise information about the
respondents' opinions on the issue.
ated scale consists of a senes o statements to wh· ·1

sump:dents are asked to express either their favourabI Ich • It is more specific in terms of degree of expression (usually five)
res h d h eness on the cpntinuum of agreement-disagreement.
unfavourableness towards it. In ot er wor s, t ey are asked t i
or disagree with the statements and tlie degree thereof Witha~1
. . I va. I ue corresponding toe~. 2. Semantic Differential Scale:
item. Each response is given a numenca
I favourableness or unfavourableness. The scores may be sum t This scale refers to a collection of rating scales using bipolar
,. .to measure the participants
·· ' s overa II att·t
1 u d e. tna~
adjectives i.e. strong-weak, fast-slow, etc. This type of scaling
technique is.used to conduct research study relating to brand image
Likert scale is most frequently used in_social scien~e research.,
and corporate image. A semantic differential scale is based on a seven
this type of scale, the respondents are asked to choose one of fi
levels of agreement/ disagreement. point rating scale for each of the several attributes relating to the
research problem. The bipolar adjectives are the extreme points with
For example: the central points being neutral. For ~xamp1e, the extreme points
could be good-bad, modern-traditional, old-new, etc. The in between
Statement: Male employees have negative attitude towards th~ categories have blank spaces ..The assigned numerical values help
female colleagues. to arrive at the scores ·for comparison of different adjectives. The
scores can be assigned either 7 to 1 or 1 to 7 d_epending upon the
Levell Level II Level III order in which they run. Generally, the scoring starts with 1 which
Level IV LevelV
Strongly Agree is unfavourable and ends with 7 which is favourable. A respondent's
Neither Disagree Stroi1gly total score is the measure of his attitude.
Agree
Agree Nor Disagree
Disagree Example: The attitude of customers towards the stores may have
Numerical Numerical Numerical Numerical Numeric~
the following adapted semantic differential scale:
Value== 5 Value = 4 Value= 3 Value= 2 Value= I
th
Inb e Likert scale, the five points form a part of the scale and in I~
a ove example at one e d f h .
th e otI1er end 1s °
. ' strono dn· t e scale 1s strong agreement and i
• b rsagreement. In between the two extremt
 Business Research Methods (S. Y.B. A..
108 ivi .s . s
...--.--- 2 3 4 5 6
1
2--~ oa taCollectio11 and Processing 109

1 High 15 _ 'What do you mean by secondary sow-ces of data? Explain its features.
quality 16 Explain ~he advantages of secondary data in research study.

---
products 9tia!i~, · What are the limitations of secondary data?
17. . hb p.
2 Numerous 18 _ Distingms etween nmary ~ata and Secondary data.
19 _ Explain the concept and techniques of scaling.
V
brands
available br¾ds Write short notes on:
20. . h d
(a) Observation met o
3 Well known (b) Experimental method
brands lJnk.no½
available
br¾ds (c) Interview method
available (d) Survey method
4 Customer
friendly Custoin
un&· OBJECTIVE QUESTIONS
staff iend1
staff .'
5 Beautiful Dgiy
store I. Fill in the blanks with the correct option :
store
layout
layout
-
1. --- observation is relatively non-planned by the researcher and
observes the elements as and when the events take place.
(a) Structured (b) Uns~ructured
REVIEW QUESTIONS. (c) Disguised (d) Senu-structured . . .
2. In case of ___ observation, the researcher will conceaJ his identity
and purpose so that the sample behaves naturally.
1. What do you mean by primary sources of data? Explain its featUr!S (a) structured (b) unstructured
2. Explain the advantages of primary data in research study. (c) disguised (d) semi-structur~d
3. What are the limitations of primary data? 3. When the researcher uses devices like cameras, .video tapes, stop
4. Discuss the different methods of collection of primary data. wa tches, etc ., then, it is called as ___ observ ahon.
5. What are the types of observation method in primary data collectio~ (a) mechanical (b) structured
) disguised (d) disguised
6. State and explai~ the different types .of experimental method u
primary data collection. 4. f he un method of data collection is mostly used in the case of
scientific study research.
7. Explain1 in brief, the different types of interview method in priman
data collection. (a) interview (b) survey
8. (d) experimentation
What are the various types of survey methods? (c) obsef rvation the entire universe is contacted to collect the data.
9. Explain the meaning ar~d importance of questionnaire. 5. In case o - - - 1
10. What are the limitations of questionnaire? ( ) Personal interview (b) e-mai survey .
a (d) telephonic interview
11. Describe the stages in designing questionnaire. (c) census survey . . f t tements Joo-ically related
12. 6. A ___ is continuum contauung a set o s a o
Explain the essentials of good questioru1aire.

i 13. to the attitude w1der study.

What types of questions can be designed while framing a (b) variable
questionnaire? (a) scale
(c) ques tionnaire (d) data
14. State and explain the class ification of questionnaire.
 Business Research Methods (S.YB.lvf
.S.:s
110 ·gru·ng numerical values to Vari
• volves asst hi Otts r1
t~J1
7. -. .- lil . de, uality and other abstract t ngs. ~egr 1 Data Collection and Processing .
op1ruon, attitu q (b) Scaling \ 111
(a) Resear_ch (d) Data analysis 17. "Strongly agree, agree, neutral, disagree, strongly disagree" is an
example of - - - scale.
( ) Sampling af
c . h begins the ~ask of - - - ter the resear (a) Semantic (b) Differential
8. A researcd ef~ ed and the research design is prepared. ch Prob, (c) Qualitative (d) Likert
has been e m d
a) data • te rpretation . , (b) ata processing 8 In order to collect primary data from the field th h · ·
( lil
1· , e researc reqUires a
(c) data ana lysl·s (d). data collection
.
f data refer to first hand information which is (a) questionnaire (b) hypothe,sis .
9. -
solve- o fi
a sped ·c pro bl em. colJec111.1
'~I (c) book {d) research report
(a) Primary sources (b) ~condary SOUrces 19. - - - are usually created as web forms with a database to store
and retrieve the answers.
(c) Multiple sources (d) Single so~ce
(a) E-mail surveys (b) Mail surveys
10. In case of _ _ , the. researcher . can cross-question the respond
and obtain in-depth information. (c) Internet surveys (d) Telephone surveys
(a) secondary data collection (b) primary data collection zo. In case of - - - interview type, the researcher focuses on the
experiences of the group of respondents and its possible effects.
(c) data analysis (d) data processing
(a) personal (b) focused group
11 There is a need to appoint field staff to collect data in case of (c) in-depth {d) telephone
data collection method. -...._ r.. . ~. .
(a) primary (b) secondary ~ns: (1) Unstructured (2) disg,iised (3) mechanical (4) experi_mentajion (5)
(c) ancillary (d) experimentation census survey (6) scale (7) Scaling (8) data collection (9) Primary
12. _ _ is also called as social media monitoring and active we sources (10) primary data collection (11) primary (12) ·social media
liste!ling. listening (13) Pre-testing (14) secondary (15) Likert (16) Semantic (17)
Likert (18) questionnaire (19) Internet surveys (20)focused'group .
(a) Internet survey (b) Social media listening
(c) E-mail survey (d) Web survey
II. State whethe~ the following statements are True or False:
13. _ _ involves rewriting questions, changing their order, replacui
open-ended questions to dose-ended questions. 1. Secondary sources of data refer to first hand information which is
collected to solve a specific problem.
(a) Survey (b) Pilot study
2. Prim!3ry sources of data collection are expensive and time consuming
(c) Pre-testing (d) Data•collection as compared to the secondary sources.
14. The _ _ data is readily available data from published or print~ 3. There is a possibility of interviewer biasness in case of primary data
sources. collection.
(a) primary (b) basic 4. It is ·essential to conduct pilot study for pre-testing questionnaire.
(c) secondary (d) qualitative 5. Dichotomous questions are a part of dose-ended questionnaire.
15. - - scale is a type of summated scale developed by Rensis Likert 6. Secondary data is easy to collect as compared to primary data.
(a) Likert (b) · Semantic 7. Secondary data is more reliable as compared to primary. data.
(c) Differential (d) Qualitative 8. Likert scale involves bipolar adjectives which are the ex~eme points
16. . . sc_ale refers to a collection of rating scales using bipola: with the central points being neutral.
adJechves 1.e. strong-weak, fast-slow, etc. ·
(a) Likert (b) Semantic A11s: True : 2, 3, '1-, 5, 6 False : 1, 7, 8
(c) Differential (d) Qualitative
 112
Business Research.Methods (S. Y.B M
. .
l
.s.: s
JII. Match the following:
GroupB
Group A
(a) Five-point sca1 ·
1. Primary data L . e
Secondary data (b) ' ogical sequence
2. (c) Seven-pointsCale .
3. Questionnaire
Likert scale (d) Accurate
4.
Semantic differential scale (e) No samplingerrors
5.

. IV. Explain the concepts:

1. Primary data ·r
2. .Secondary data
3. Interview
4. Questionnaire
5. tikert scale
6. Semantic differential scale
 113

• Meaning of Data Processing

• Data Processing Stages
• Meaning of Data Analysis
• Characteristics of Data Analysis
• Purposes of Data Analysis
• Types of Data Anaysis
• Data Interpretation
• Essentials of Data Interpretation
• Methods of Interpretation
• Significance of Data Interpretation
• Importance And Significance·of Data Processing
• Testing of Hypothesis
• Steps/Procedure In Hypothesis Testing
• Types of Hypothesis Testing

'
 114 Business Research Methods (S. Y'. B.-M.S.: SEA,f :
[] ~1'.~~~ 'n~!.,v-,:$\~ · •·· · w ..... ..... . -IVJ

raw data .
Aft er coll ecti ng dat a, ·the met hod of con ver ting
a ana lysi s into
mea nin gfu l stat eme nt; incl ude s dat a pro cess ing, dat
invo lve·s ;;~d
dat a inte rpre tati on and pres enta tion: Dat a pro ces sing
ion and gra ~ll)
stag es suc h as edit ing, cod ing, clas s¥ic atio n, tal? ulat p c
·
pre sen tati on of data .
for th
Dat a processing is a process of skil full y organising of data
ing c~
pur pos e of dat a ana lysi s and interpretation_- D_at~ pro cess
ted or it can be
be don e man uall y whe n the dat a coll ecte d 1s. hm1
e quantities.
don e mec ~an ical ly whe n the coll ecte d data inv olv e hug
dat a ·collection
Dat a pro ces sing is the inte rme diar y stag e betw een
dat a collection,
and dat a ·a naly sis. The com plet ed inst rum ents of
d note s contain
suc h as inte rvie w que stio nna ires , data she ets, and fiel
f,

are like raw

a vas t mas s of data . The coll ecte d data inst rum ents
vid e answers .
mat eria ls and ther efor e, they can not stra igh taw ay pro
d for skilful
to rese arc h que.s tion s. The reJo re, the re is a nee
·
· man i'pu lati on of data , i.e., data pro cess ing:
arch design.
Dat a pro cess ing nee d~ to be plan ned at the stag e of rese
onl y afte r the
l Mo st of the rese arc_h ers thin k of dat a pro ces sing
be confronted.
com plet ion of data coll ecti on. The rese arch ers may
ided if careful
wit h cert ain une xpe cted pro blem s whi ch can be ·avo
s earc h design
plan nin g for data pro cess ing is und erta ken at the re_
r

like ·c odin g of
stag e. The adv anc e plan nin g may cov er suc h asp ects
Thi s should
data , cate gor izat ion of vari able s, tabu lati on of data , etc.
Vi be don e wit h refe renc e to the req uire men ts of.t ~sti ng
hyp othe ses or

I inve stig ativ e que stio ns.

classifica~io~,
Pro cess ing of dat a is a pro cess of edit ing, ~odi:ng,
pro ces sing ~s
tabu lati on and gra phi c pre sen tati on of data . Dat a
on.. Analysis
req uire d for the pur pos e of ana lysi s and inte rpre tati
· · . · .
and inte rpre tati on of data facilitates:
Data Analysis and Interpretation 115
• Research Findi ngs .

• Draw ing Conclusions

• Recom mend ations

The variou s stages in data pro~e ssing are as .follows :

ISTAGES IN DATA PJ,WCESSINGI

J,
Editing
J,
Codin g
J,
Classification
J,
Tabulation
J,
Graphic Presentation

Editing is the proce ss of checki:pg errors and omiss ions in data

collection, and makin g corrections, if required. Editing is ·requi red
when : · ·
.• There -is inconsistency in respo nses given by the respo ndent s.
• Respo ndent s may provid e incorrect or false responses.
• Some va,g ue/ incomplete answe rs given by the respo ndent s.
• No responses are provi ded by the ·respo ndent s for certai n
- I

questions.

Examples:
1. Certa in respo nses are incon sisten t with other answe rs. For
instance, in resp.onse to a questi on on qualit y ·of products,.the
 ◄
116 Business Research Methods (S. Y.B.M.S.: SEM-I\1)
respondent may state that he prefe.rs to buy high quali
products. In another response relating to price of the produc~
the respondent may state that he norm~lly purchases lowe;
priced items. There is inconsistency in the two. responses as
· high quality products are generally priced higher. Therefore
th~ researcher must.use judgepient to bring consistency. '
l

2. A respondent .may mark_two options in the case of multiple.

choice responses, instead of only one for a particular question ·
For .instance, responclent may tick mark twice for a partkuI~
question' - ·'how often you read a newspaper'- the respondent
may reply every day ~swell ·a s once in a week:
' .
3. At times, the data collected may -be inc_o rrect due to unethical
interviewers.· Also in the case of mail s·u rvey, the r.espondents
may provide c;asual responses which may be incorrect.

Separate Ink for Edit~ng_:

The changes made by the editor must be done in separate ink,
preferably green ink or in pencil. The·ink should nofmatch with the·
ink used by'the respondent or interviewer. ·

Types of Editing
(a) Field Editing: Editing u~dert'aken at the time of field survey is
called as field editing. At the ti.me of_interview, the interviewer
may use _several abbreviations due to time constraint. These ·
abbrevi~tions need to be spelt out' fully, at the time of processing
of data. The interviewer must also go through the completed
questionnaire to find out whether or not there is a _ne~d for
completing partial or correcting vague answers.
. .

(b) Central Editing: Editing done at. the central office is called
central editing. A single ~ditor should carry out this task so
that consistency in editing can be ensured. But in the case of
'I large studies, two_or more editors can handle .the task. The
multipl'r editors must be given.proper guidelines for consist~ncy·
in editing. Som~times, the entire questionnaire may be divided
in two parts, and ea~h part c~ be edited by separate editor. .
Data Analysis and Interpretation 117
Essentials of Editing ·
1. Accuracy : Editor must look for accurate responses. Sometimes,
interviewee may give wrong responses. Therefore, interviewer
must be given proper training to obtain accurate responses. The
editor must use judgement to correct incorrect respoI1.pes.
2. Avoid Bias : The editor must avoid bias. For instance, if the
editor is positively inclined towa,rds advertising as a tool of
promotion, he may edit the answers in favour of advertising,
even thou'g h the responses given by respondents · may be
otherwise. · ·

3. Consistency : There must be consistency in editing. If there is

more than one editor, they must follow the same patter:n of
editing. For instance, if there are incomplete responses, all the
editors may follow the same norm for considering the
incomplete responses as nil response. or. all the editors may
complete the incomplete responses by using their judgement.
4. Completeness : The questionnaire must be checked to find out
whether or not the respondents have answered all the questions.
If certain questions are unanswered, the editor may use ·
judgement to .deduce proper responses for the unanswered
responses b~sed on other completed responses.
5. Training to Editors: Training must be provided toJhe editors
to edit the data properly. Training not only helps to improve
the skills of the editors for effective editing, but it also develops
a positive attitude of the editor towards editing and the
organisation.
6. Reliability : Data collection must be reliable. Editor must .'
consider only the up-to-date data and discard the outdated data
at the time of editing. This is because; outdated data may lead
to wrong research findings. ·
7. Uniformity: Editor needs to check whether or not each and
every respondent has interpreted the questions and instructions
uniformly. For instance, the question may have a scale of 1 to 5,.
where 1 = excellent and 5 = poor. But the respondent may
 <
118 B·usiness R esear~h Meth~ds (S. Y.B.M.S .: SEM-IV)
interpr et the scale the other way. In such situatio ns th
respon ses would be defecti ve. • · ' e
8. Econom ical : Editing must be econom ical. The time, money-
effort' involv ed in editi~g must bring good ·r eturns to~~
researc her. As far as possibl e, the researc h~r must keep a ch k
.
on t h e costs mcurre d oned"1tmg.
. ec

It is a proces s of assign ing codes to the variou s stateme nts or

ques~o ns in the questio nnaire. <;::oding is special ly require d when
the sample size is large and there is large amoun t of data collection
from respon< :Ients. Coding faci~ita tes proper tabulat ion and analysis
·of data. · ' _ ·

Types of Codes
• Numer ical Codes such as Code 1, Code 2, Code 3 and so on.
• Alphab etical Codes· such as.Cod e A, Code B;_Code C and so on.
• Alpha- Numer ical Codes such as Code Av A , A ,. B , B , B , etc.
2 3 1 2 3
Cod~n g can be consid ered as an elemen t of classifi ~ation. For
exampl e:· The researc her may conduc t a study on TV Viewer ship.
The categor ies may be Males.( Code M) and Female s (Code J➔) .
. ~

The main categor ies may be further divided into four a~e gro~ps :

• U pto 12 years • ·13 to 19 years

• · 20 to 39 yecl!s • 40 ye~ns and above
Thus, there will be 8 ·~odes = M ~, M 2, M 3 , M 4 and ·pv F 2, F3, F4·
Steps in Coding_ .
· · · ust
1. Specif y the·Catego ries : The researc her or data process or :ed.
specify the categor ies into which the respons es can be class d
For instanc.e, the catego ries. m~y inc. 1u d e:_ A
ge, Gen er,
Educat ion, Income , etc . .
 119
pata Analysis and Inte_rpretation
Allo cate Individ1:1ai Cod es : The rese arc her mu
st allo c~:~
Fo~ exa mp . ;
z. indi vidu al cod es to eac h cate gor y of resp ons es. es foll ows ·
Mal es in the fou r age gro ups may be allo cate d cod as
M , M 2, M 3, and M 4 .
1

t ca~ e?o ~es .

It is the proc ess of gro upi ng of coll ecte d data into diff eren
.classif1cat~on
Therefore, cod ing is an elem ent of clas sific atio n. The
der wis e,
can be acco rdin g to diff eren t categories: age gro up wis e, gen
wis e, etc. ·
educatio~al leve l wis e, inco me·gro up wis e, occ upa ti~n
-gr oup s-· For
Each of .t he cate gori es can be furt her div ided into sub
diff ere nt .
example: The age gro up can be furt her div ide d into
mid dle age d,
categories suc;:h as: .chi ldre n,. teen age rs, you ng adu lts,
and senior. citiz ens.

Principle~ of Cod ing / Cl~ ssif icat ion

I

1. Mu tua lly Exclusive:

c cas e or
The c~te gori es mu st be mut uall y exc lusi ve. A spe cifi
onl y. For
resp ons e mu st be clas sifie d only onc e in one cate gor y
resp ons e
instance, on the basi s of occ upa tion , one may plac e the
cate gor y.
of a ·par ticu lar resp ond ent in a defi nite pre -de term ined
to two
But the pro blem may aris e, if th~ resp ond ent belo ngs
to the
cate gori es. For inst anc e, the resp ond ent may bel ong
sult ant
man age rial occ upa tio~ (full ti.me) and pro fess iona l or con
i~al s.t aff
occ upa tion (pa rt time ) or a -p erso n m·a y wor k as cler
be solv ed
{full time ) and stud ent _(par t time). Thi s pro blem can
nci pal
by clas sify ing the info rma tion or resp ons es und er 'Pri
Occ upa tion ' stat us. · ·

2. Appropriateness:
rese arc h
The clas sific atio n/ cod ing n:iust be app rop riat e to the
loy alty of
· wor k. For inst anc e, a researche_r stud yin g bra nd
cert ain
read yma de garm ents may clas sify the pop ulat ion in
s and the
~ou ps app rop riat e to the surv ey. The sen ior citi zen
bra nds
kids may be igno red as they are not mu ch loya l to the
as far as read yma de gar men ts are con cern ed.
120 Business Research_Methods (5. Y.B .M.S. : SEM-IvJ
3. Exhaustive:·
. The re mu st b
The classification mu st be exh aus tive in nat ure
fitted or placecte
a sep ara te cate gor y wh ere the resp ons es can be
y. For instance·
The res pon den t mu st belong to a certain cate gor
re mu st be '
if clas sifi cati on· is bas ed on stu den ts the n the
cati on is don :
cat ego ry for eve ry class of stud ent s. If the d~s sifi
, the n certain
onl y on the bas is of arts, com mer ce and sqe nce
g stud ent s
stu den ts ma y be .exc lud ed suc h as eng ine erin
ing to oth e;
ma nag em ent ~tu den ts, and stu den ts bel ong
the re mu st be
pro fess ion al or technical cou rses . The refo re,
ny gro ups , the
sev era l classifications . But if the re are too ma
und er single
res ~ar che r ma y inc lud e the iso late d gro ups
cate gor y called as "General Category."

ula r for ma t for the ·

It refe rs to tran sfer ring the classified. dat a in a tab
es sor ting of data
pur pos e of analysis and interpretation. It inv olv
of resp ons es that
into diff ere nt categories and cou ntin g the num ber
bel ong to each category.

V atia ble s in Tabulation

ere in onl y one
1. Univariate: The tabu lati on can be uni var iate wh
For example
variable is involved in tabu lati on suc h as '~o ys'.
to a particular
15 out of2 0 boys-75% hav e resp ond ed positively
qu~stion. .

Multi-variate : Wh en two or mo re var iabl es

are inv olv ed in
2.
tabulation. For
tab ula ting the dat a, it is called as mu ltiv aria te
10 (60%) in the
example, mal es and age gro up - 6 mal es out of
to a particular
age gro up of 13 to 19 yea rs res pon ded pos itiv ely
gro up of 20 to
que stio n, and 5 n:iales out of 10 (50 %) in the age
stio n.
39 hav e res pon ded pos itiv ely to a par ticu lar que

Me tho ds of Tabulation

Ma nu.al Tab ula tion : . Wh en tab ula tion is

don e man ual ly
1.
nua l tabulation.
wit hou t the hel p of com put ers, it is called as ma
oata Analysis and_ Interpretation . ~21
Manual tabulation is possible when the number of variables is
few and the sample size is limited.

2. Mechanical Tabulation: When tabulation is done with the help

of computers, it is called as mechanical tabulation. Mechanical
tabulation is nece~sary when the number of variables is ·large
and also the sample size is large. ·

Guidelines/ Principles of Tabulation

1. Title : Each and every statistical table needs to have a clear,
and suitable title. For instance, a report on the causes of decline
in sales -will have a·title "Causes of J]ecline in Sales.of Produ~t
ABC in the Year 2014". ·

The table title should give a clear and·accurate description of

. the data. It s~ould answer the three questions ~'what", "where"
and "~hen". Be short and concise, and avoid using verbs.
2. · Units of Measurement: The units of measurement under each
heading or-sub-heading must be always indicated for clear and
better understanding. -The measurement ·may be in terms of
amounts such as rupees, qollars, and so on, volumes such as
kil?s, quintals, tonnes, etc., percentages, or some pther un~ts.
3. Numbering of Tables : Every table needs to have a distinct
number to facilitate easy reference. The numbering of tables is
specially required when there is large number of tables.
. .
4. - Explanatory Footnotes: Explanatory foot notes relating to the .
table should be below the table along with .the reference symbols
used in the ·t~ble-. Footnotes, at the bottom of the table, may
provide any additional information needed to understand and
rise the data correctly '(e.g. d~finitions).
5. Source of Data : The table needs to indicate th~ source of data.
For instance, if the data is taken from Economic Survey of India
-2018-19, the table
. should indicate the source. Source indication
gives authenticity to the_·data.
6. · Apptoximation of Data: Generqlly, it is better to approximate
. the data before tabulating the same. Data figures can be briefed. ..
 1_22 Business Research Methods (5. Y.B.M.S. : SEM~Jv
vert ed . J
For exa mpl e, the data figures of millions can be con l11t0
·11· .
b1 ions as t~e-num ber of digi ts gets redu ced .
7.. Column Headings : The column headings and the row headin
the Iiterags
of the tabl e sho uld be clear and _briEf For instance,
ra?,
!ate in urb an area s can be ~hown in the table as "Urban Lite • Y
Rate". · ·
the table
8. Row Stubs : The row stubs -in the first colu mn of
table. '
sho uld iden tify the data pres ente d in each row of the
ral colu mns in a.
9. Num ber ing of Column : Wh en ther e are seve
Numbering
table, then ec:tch column mus t be.serially num bere d.'
. of colu mns facilitates easy reference. .
'
10. Pla cing of Columns : The colu mns who se dat a are to be
plac eme nt
com par ed sho uld be plac ed s~de by side. Suc h
facilitates pro per comparison. ·
sep arat ed by
11. Separation of Columns : The colu mns mus t be
tive. Thick
line s whi ch mak e the table easily read able and attr ac'
mns .
lines mus t be plac ed to sepa rate two unr elat ed colu
figu res in a
12. Alignment of Data :. It is i.rµpo~tant that all the
ativ e signs
colu mn sho uld be suitably aligned. Positive and neg
mus t also be in perf ect alignment. .
righ t in the
Align t~e num bers on the decimal poi nt (or on the
is clear. Do
absence of decima.1 places)· so thei r relative valu e
aII the same
not centre the num bers in a column, unle ss they are
ly restrict
mag nitu de. Use min imu m of decimal places, pref erab
max imu m to two.
d eith er by
13. Displaying of Oata : The data can be disp laye
e stan d~d
chronological ord er for tim e series or by usin g_som
app rop nate
classifi~ation. For longer time series, it may 'be mo re
use the reve rse chro nolo gica l ord er (i.e. star ting wit h the m 0st
to
suc h ~s for
rece nt p~riod and goin g bac kwa rds) in som e case s,
mon thly une mpl oym ent. . .

cell emptY•
14. No Empty Data Cells : Do not leav e any dat a
le" or ''not
Mis sing valu es sho uld be iden tifie d as "no t ava ilab
(not
applicable". The abb revi atio n "NA " can app ly to eith er

1
 1111111'

oata Analyszs· an dInterp.retatwn

· 123
available or not applicable), so it needs to be defined in the
l
· footnote.
'!

i The research data needs to be presented effectively for quick and

1

~: clear understanding. Bar graphs, pie charts, line graphs, histograms

-:. and other pictorial devices are an excellent means to present _the
\ data.
\

Pie Chart: A pie chart.is a circular chart used to compare parts

of the whole. It is divided into sectors that are equal in size to
the quantity represented. For instance, a -pie chart is divided
I•
int(? different parts to.indicate percentage sales/profits/market
\
share, etc., of various brands of a company during a particul~r
period. The reader can understand at a glance the relationship
between various parts of_a pie-chart.

0-BrandA
- Brand B

0 -BrandC
- Brand D

Percent of Sales Revenue of Brands

2. Bar Graphs : _A bar chart or bar graph is a chart with rectangular

bars with lengths proportional to the values that they represent.
·The bars can be plotted vertically or horizontally. A bar chart is
.,I
l
,
very useful for recording discrete data.
t·

i A bar graph is a chart ~hat uses either horizontal or vertical

bars to show comparisons among categories. One axis of the
chart shows the specific categories be1ng compared, and the
·,1
other axis represents a discrete value. .
11
'I
·I
 Business Research Met hods (S. Y.B .M .S. : SEM-Iv
124 )
Sim pie Bar Diag ram
45
40
~
e
35
~

u 30

·-=
VJ
~
~
25
20 ~
Q.
=
~
15
10
5
0 2012 . 2013 2014
2010 2011
Years

scaled
Bar cha rts hav e a disc rete ran ge. Bar cha rts.are usu ally
rt may be
so tha t all the dat a can fit on the cha rt. Bar s on the cha
hes t to
arra nge d in any.. ord er. Bar c~a rts arr ~ge d from hig
lo~ est inci den ce are call ed Par eto_cha rts. The re can 1?e
gro upe d
usually
bar gra phs and stac ked bar gra p~.· Gro upe d bar gra phs
upi ng.
pre sen t the info rma tion in the sam e ord er ih eac h gro
gro upin g. ·
The bars are colo ur-c ode d to rep rese nt' a par ticu lar
anc e of
For inst anc e, a bus ine ssm an ma y sho w the . per form
hav e a
diff eren t bra nds , the .bar rep rese ntin g eac h bra nd can
· ·
diff eren t colo ur.
sen t the info rma tion ih the same
'
Sta cke d bar gra phs pre
or· more
seq uen ~e on eac h bar. The stac ked bar can hav e two
ked bar
par ts. For inst anc e, th~ foll owi ng dia gra m sho ws stac
gra ph. . .
n in three
The foll owi ng tabl e ·sho ws sub je~t wis e -dis trib utio
·col lege s:

Sub ject s Num ber of Stu den ts

Col lege A · Col lege B Col lege C
· Accountancy 300 250 150
Man age men t 200 250 100
·Economics 250 . 150 . 250
· Total 750 650 500
Data Analysis and Interpretation 125
y /

(h;,) Accountancy
800
181 Management
700 D Economics
....VJc=
~
600
'C
:s 500
ri1
c.,.
0 400
Q
z 300
200
100
_--z~i~ir~
....__,_.......,____"--_....1..-.....,._.,_.1-__;.-'---'---➔
X

College A College B College C

3. Line Graphs:
A line grap h show s information that is connected in som e way
(such as chan ge over time). A line char t or line grap h is a type
of char t whic h displ ays ·information as a series of data poin ts
called 'mar kers ' connected by straight line segments. It is a basic
. type of char t com mon in man y fields.
Line Char ts· show how a parti cul~ r data chan ges at equa l
intervals of time. A line .char t is often used to visua lize a tren d
in data over intervals of time - a time series - thus the line is
often draw n chronologically.

. ~ 400
I,;.

e
CJ
~ 300
~
=
C
~
> 200
~
~

- 100
~
o:s
(:I)

0
--- -+ --- --- +- --- +- --- --+ -~ x
Jan. - Mar. April - June July - Sept. Oct. - Dec.
Period
 ◄
SE M- Iv )
126 Busin~~s Re_sea_rch Methods (S. ~-B .M .S.:
4. Ga nt t Ch ar ts :
a de ve lop ed by He nry G
A Ga ntt ch art is typ e ofba r ch art,
t sch ed ule . Fo r ins tan c;t t
in the 19 10 s, tha t ill us tra tes a pro jec fa
riz on tal or ve rti cal ba rs
. Ga ntt ch art ma y co ns ist of tw o ho
rio d of tim e/a cti y1 ty. On e ba r i
_ nd i~a t~s the pla nn ed ~
ea c~ _pe
er ba r md 1c ate s the actual
an tic ipa ted pe rfo rm an ce , an d the oth -
· .
pe rfo rm an c_e. -
. .
5. Hi sto gr am s :
gr ap h wr ,er e the intervals
A hist<;>gram is. a ·sp eci al kin d of ba r
a gra ph ica l rep res en tat ion
are eq ua l. In sta tis tic s, a his tog ram is
th e dis tri bu tio n· of da ta. It is an est im ate of ~he pro bab ilit y
of
an d wa s fir st mt rod uc ed
dis tri bu tio n of a co nti nu ou s va ria ble
by Ka rl.P ea rso n._
y

so
~
-=~ 40
·"C
...,=
r,J
'- 30
0
i-.
~
~
·s 20
z=
10
x·
0 60 80 100
0 20 . 40
ble . 100)
Score on Final Ex am (Maximum possi

n
Gu ide lin es for Gr ap hic Pr ese nta tio
lou rs: If the re is mo re tha n on e cu rv e or ba r, the y sh ou ld be
1. Co
er by· dis tin ct pa tte rns or
cle arl y dif fer en tia ted fro m on e an oth
co lou r 'to ·m ea n the saJlle
co lou rs. Be consistent; u~ e the sa me
thi ng in the ch art or·gra ph .
Data Analysis and Interpretation
127
2. ~o~izontal a~ d Vertical Axis : Generally
, the hor izo nta l ·axis
indicate t~e dis cre te val ues suc h as scores~
hei ght s, am ou nts ,
etc., and the ver tic al axis ind ica te the fre
quency or nu mb er of
stu de~ ts, ma les , femgles, tre es, etc., for
par tic ula r dis cre te
values . .

3. Numerical_Da ta Ta ble : Nu me ric al dat a up

on wh ich the gra ph
or cha rt is bas ed sho uld be pre sen ted in an
acc om pan yin g table.
4. Title ·: T~e gra ph or the cha rt sho uld hav
e a clear, concise an d
simple title. Th e title .sh ou ld describe the
nat ure of the dat a
pre sen ted.._Th e title ma y be pla ced either
above or bel ow the
cha rt/ gra ph. ·
5. Placement of Variables : The me asu rem
ent variables sho uld
be pla ced fro m left to rig ht on the hor izo nta
l line.a nd gen era lly
froin bo tto m to the top on the vertical line
.
6. Placement of the Graphs : Generally, the
graphs/chart mu st
fol low an d nq t pre ced e 't he general text ma
tter. Ex pla nat ion of
the gra ph / cha rt ma y follow the diagram.
7. Graphic forms sho uld be use d sparingly
: :roo ma ny gra phi c
forms det rac t rat her tha n illustrate the pre
sentation.
'
8. Define Target Au die nce : The researche
r nee ds to define the
tai:get aud ien ce. De pen din g on the tar get
audience, t~e gra ph s
or charts nee d to he use d. For instance, sim
ple cha rts or gra ph s·
may be use d for lay ma n. Fo r technical peo
ple, complex gra ph s
ma ny be use d.
9. Determine the me ssa ge yo u wa nt to
communicate : Th e
researcher sho uld determine whether the dat
a intends to convey
only on e J?'.lessage or mo re tha n one me
ssage. For instance,
wh eth er the gra ph / cha rt int end s to sho
w onl y the sales or
whether it int end s to indicate sales as well as
profits of particular
pro du cts /br and s du rin g a particular per iod
.·
10. Determine the Nature of the Message
: The researcher nee ds
to identify the nat ure of the message to
be com mu nic ate d to
· the audience..Fo r instance, the researcher
ma y wa nt to com par e
the per for ma nce of sales du rin g the two per
iods or com par iso n
of sales bet we en two bra nds , etc.
·
 · ◄

Business Research Met hod s (S. Y.B .M .S.: S£1iK VJ

128 . . iv1~1

on the 5 sta
No te to Stu den ts: If a que stio n is ask ed for 15 ma rks
cft ma y incz!~'
wri te abo ut 20 to 30 line s on eac h sta ge - whi
rel eva nt ma tt e
me ani ng, ess ent ials /gu ide line s or any oth er er
inc lud ing sign ific anc e. .

ME TH OD S. OF TA BU LA TI ON (For Reference)
Tab ula tion and (2)
Tab ula tion ma y be dor ie by (l)_Hand or Ma nua l ·
Ma chi ne Tab ula tion . ·

1. -. HA ND / MA NU AL TA BU LA TIO N
d of tabu lati offi n
Ha nd tab ula tion is a trad itio nal and old me tho
and cou nti ng of
wh ich the wo rk of sor tin g,· ·n um ber ing
le to tab ula te data
· q1:1estionnaires is don e by han ds onl y. It is sui tab
, the num ber s of
on a sma ll sca le i.e., the size of the sam ple sur vey
tha n 500. Han d
res pon den ts and que stio ns'a ske d is sma ll, say , less
e.
tab ula tion is eas y and sim ple to foll ow and exe cut
t pre par es aplan
Pro ced ure of'H and Tab ula tion : Th~ rese arc her firs
wit h the hel p of a
for tab ula tion pur pos e. The pla n is pre par ed
on, the object of
num ber o"t ·r esp ons es, prin cip les of clas sifi cati
stio nna ires are
res ear ch etc. Aft erw ard s dul y com ple ted que
res pon se she ets
num ber ed ser iall y. Aft er num ber irig , the
cifi ed categories.
(qu esti onn aire s) are sor ted out and' gro upe d into spe
ts, the n the y can
For exa mp le, afte r num ber ing say 100 res poh den
gor y. Afterwards
be clas sifi ed and gro upe d into ma le & fem,a le cate
) counti11:g of ind ivid ual ans wer s is tak en up. Co unt
ing is oft~ n done
llin g wo rk eas y and
_ in sets of five tall y ma rks_thu s, ma kes the tota
V ·q~ ic~ .
At the end , the res ults of cou ntin g are sho wn qn
cou ntin g she ets for
ula tion is given
eac h res pon se sep ara tely . A. sam ple of han d tab
bel ow.
low ing questions.
Sup pos e a gro up of 45 res pon den ts is ask ed the.fol

Wo uld you pre fer bra nde d pro duc t ove

)

r non -br and

.
ed? O Yes
0 No
\ Data Analysis and Interpretation
. 129
i1 Afterwards th e researcher rece
ives all th e 45 Questionnaires
1
1 co m pr isi ng
ye s or n~ an sw er s. Fo r ta bu la tin back,
1
separate 'Y es ' re sp on se fro g ·pu rp os e he will
m 'N o' re sp on se an d~co un t
Afterwards th e sa m e da ta will th em .
b~ ta bu la te d as un de r.
-
Q ue s_tio n
No Yes
hI
Would yo u pr ef er br an de d
product ov er no n- br an de d? 4-H1 -1-+tr m t J..H1 -H11
l--H1 _µ ff J-H1 J-H1
Total
20 25
It is seen fro m th e above table
th at th e responses (Yes/No)
~ been re co rd ed by us in g fou! ha ve
vertical an d _on~ diagonal strok
· is calle~ as_ (Five-
Barred ~a~e) m et ho d of reco e. This
1 rdin~. ~a~h ga te
; repres~nt Five an sw er s an d if
th e 'G at es ' are ar ra ng ed m line
: 10, the totalling w or k is ~r th of 5 or
. . er s~mplified.
Merits of H an d Ta bu la tio n:

1. H an d ta bu la tio n is a sim pl e,
easy an d convenient m et ho d
tabulation. of
,,,
2. M an ua l pr oc es sin g or tabulatio
n is flexible. It ca n be do ne by
any ju ni or staff un de r th e guid
ance of a supervisor. It do es -no
require technicians. t
3. H an d ta bu la tio n can be do ne
wi th ou t co di ng or processing.
4. M an ua l ta bu la tio n is suita
ble wh er e there is a great de
coded in fo rm at io n or wh er e ai of
few re sp on de nt s are in te rv ie we
extensively. d·

Demerits of H an d Ta bu la tio
n:
1. Manual ta bu la tio n is a time co
ns um in g method. It takes tim
to analyse, gr ou p, co un t ~ d re e
cord individual responses.
2. There ar e chances of cleric
al an d hu m an errors as it relie
hu m an ha nd s an d memory. s on

3. It is no t suitable an d co nv
en ien t to handle large and ex
research projects. tensive
 130 Business Research Me_thods (S. Y.B .M.S.: SEM~rv
4. Manual tabulation does not f acilitat.e cr·o ss classificatio )
establish inter relationship between different variables. n to

5. . It can not be used. in periodic

. types of surveys i.e. (continuous.
data collection work) . · . ·

2. MACHI-NE TABULAT ION·

The need for speedy analysis of vo~umes of data has resulted into
substitut~on of hand-tabula tion by machine tabulations . Mechanical
method uses automatic and fast·power m<:1chines for sorting,
counting, classifying a,nq. even tabulation of bulk data. Some
machines like computers, are.equippe d to perform most complicated
statistical jobs of calculation, analysis, interpretati on arid even
printing. In modern times, computers are being extensively .used
for s·mall and big research works for data analysis. ·

Merits of Machine(M echanical Tabulation :

1. Machines can be used to handle extensive and vast volumes
of data convenientl y.
2. lt facilitate~ cross-classi fication and ~tudy of many variables
I •
simultaneou sly.
3. , It is extremely useful .in periodic surveys requiring addition -
and upgradatio~ of already collected d<:lta at regular inte~als.
4. Machine tabulation is more accurate and there are less cha~ces
of errors in the process of sorting and counting.
5. It is time saving and economic~l if_used for large scale surveys.
Demerits of Machine/ Mechanical Tabulation :
1. Machine tabulation involves much clerical. work and
specialised operations.
- .
2.· It needs prior sorting and coding of data before it is fed _to_
machine. In· olden days punch-card s were used for coding
- p:ur_pose.
. 3-. Machine tabulation is .less fle~ible as machines can only act
but cannot think. •
Data Analysis and Interpretation 131
4. Machine tabulat ion requires skilled operators. It needs trainin g
and superv ision. Machines need regula r. mainte nance and if I

n.o t service d prop~r ly may result in freque nt break- downs

contrib uting to waste of time and resources. .
s. The speed and accuracy of machines may induce researc hers
to-collect large data on unrela ted topics- and deviate from the
original objectives of research.

The analysis of data is the most skilled task in the researc h process.
the
It calls for resea:(cher' s own judgem ent and s_kill. Analysis_means
a critical examination of the assembled and groupe d data for studyi ng
the characteristics of the object under.s tudy and.fo:r; determ ining the
patterns of relationships among the variables relatin g to it. Both
quantitative and non-qu antitat ive metho ds are used. Bowev er,
social ~esearch most.often requires quantitative analysis involv ing
the application of variou s s·tatistical techniques·:.
Analysis of data.sh ould be done by the researcher himself. A correct
analysis needs familiarity with the backgr ound of the·survey and
all the stages of research. The analysis does not necessarily be a
statistical one. · Quanti tative and non-qu antitat ive metho ds of
analysis can be done.
According to Selltiz and others, .".Analysis of data do not make such
a precise differentiation. Analysis is a comprehensive process,
involves processing." _
The dividirl.g line betwee n analysis of data and interpr etation of
data is ·difficult to draw. These ·two are symbio tic and merge
imperceptibly. If analysis involves organising the data in a particular
manner, it is mostly the interpretative ideas that govern this task. If .
the end produc t of analys_is is the setting up of certain genera l
conclusions, then what these conclusions really mean and reflect is /
the bare minim um that the researcher must feel obliged to know.
Interpretation is _the way to this knowledge. Thus the task to analysis
~an_hardly be said to be complete withou t interpretation coming to
1llumj.na
te the results. ·
 .S.: SE!vf-IvJ
13 2 Business Researc_h Methods (5. Y.B.M
~ Da ta Analysis f-
V

- 1
Int erp ret ati o1
. ·of
~ res ult s
~
u . . . L.. Resea rch

g
u . I
D1scuss1on 1 · questio
answered?
n

Go od nes s Hy po the ses

~ Ge ttin g da ta rea dy Fe el tor tes tin g
dat a of dat a
~ for ana lys is
Reliability Ap pro pri ate . .
Editing'"'1ata J. Mean sta tist ica l
2. Standard ·va lid ity
- Ha nd lin g bla nk ma nip ula -
res po nse s deviation
tio ns
Co din g .dat~- 3. Co rre lat ion s.
Categorizing data 4. Fre qu enc y
Cr eat ing dat a file distribution,
Pro gra mm ing etc.

oc ~s
Fl ow Di ag ra m-ofDa ta An aly sis R,

Fo llo ~i ng -ar e the ma in ch aracteristics~of an~lysis of da ta:

'
.

sis of da ta is on e of th e m
· o.s t im po rta nt as pe cts of _research.
• Analy
ical job, it sh ou ld be carried
S~ ce it is hi gh ly skilled an d techn clo se-supervision. It
res ea rch er h
. im se lf or un de r hi s
ou t by th e
s a de ep an d in ten se kn ow led ge on th e _pa rt of the
de ma nd
ea rch er ab ou t th e d~ ta· to be an aly se d. Th e re se ar ch er should
· res
o po ss es s ju dg m en t sk ill, ab ili ty of ge ne ra lis ati on an d should
als
.objects an d hy po th es is of
be fam ili ar wi th th e ba ck gr ou nd
stu dy .

Da ta, fac ts ~ d fig ur es ar e sil en t ·an d th ey ne ve r sp ea k f~r

• . It is th ro ug h systema~c
bu t th ey ha ve co mp lex iti es
themselves
th at the -im po rta nt- ch ara cte ris t1c s \Vhich ar e hi dd en .lll
_an aly sis
ta ar e bro :a,igh t ou t an d va lid ge ne ra lis ati on s ar e dr a~·
th e da
is · d~ m an ds a th or ou g~ kn ow le dg e of ·on e's. data-.
~n-~ lys
p

oata Analysis a~d Interpretation . 133

r

Without deep knowledge, the analysis is likely to be aimless. It

is only by organising, analysing and interpreting the r~sear~h
data that we know their important features, inter-relation ship
and cause-effect relationship. The trends and sequences ·
inherent in the phenomena are elaborated _by me.ans . of
generalisation. ·

1
According to~- V. Young, "The function of systematic analysis
is to•build an inteHectual edifice in which properly sorted •and
shifted facts and figures are placed ·i n their appropriate settings
and broader generalisations beyond the immediate contents of
the facts under study, consistent relationships, or that general
inferences can be drawrt from them-the aim of ~ mature
. ,,
science.
,The data to be analysed and interpreted should : (i) be
• reproducible, (ii) be readily disposed to quantitative treatment,
and (iii) ha~e significance for some systematic theory, and (iv)
can serve as a basis for br~ader generalisation.
We should reme:r;nber that the steps envisaged in the analysis
• of data will vary depending on the type of study. A set of clearly
formulated hypothesis to start with the study presents a norm
prescribing a certain action to be taken. The more specific is the
hypothesis, the more specific is the action and in such types of
.studies, the analysis of data is almost completely a mechanical
procedure.

• If the data are collected according to.vague clues rather than

according to the specific hypothesis, the data are analysed
inductively or invested during the proces~ and not by means
of new prescribed se~ of rules.
The task of analysis in incomplete without interpretation . In
fact, analysis · of data and interpretatio n of data are
complementa ry to .each other. The end product of analysis is
the setting up .of certain general conclusions while the
interpretation deals with what these conclusi9ns really mean.
• · Since analysis and interpretation of data are interwoven the
interpretation should more properly be conceived of as a special
aspect of analysis rather than a .distinct operation. Interpretation

'
I
134 Busin ess Research Methods (S. Y.B.M .S.: SEM-IV)
i_s the proc ess ~o esta blish relat ions hip ~etw een varia bles which
.~re expr esse d 11:1 the find ings and why such relat ions hips exists.

• For any su~c essfu l stud y the task of anal ysis and inter preta tion
sh?u ld be desi gned befo re the data are actu ally colle cted with
the exce ptio n.of form ulati ve stud ies whe re the rese arch er had
no idea as to wha t kind of answ er he wan ts. Othe rwis e there is
~lwa ys a dan'g er to bein g too late and the chap ces of missmg
imp orta nt rele vant data . · .

Stati stica ~ anal ysis of data se.r ves seve ral majo r purp oses·.
It sum·m ariz es ·la:r~ge mas s of data into und erst and able and.
·
mea ning ful form . This is the role of stati stics . The redu ctio n of data
facil itate s furth er .analysis.
Data anal ysis piak es exac t desc ripti ons poss ible. For exam ple, when
,
we s~y that the educ ation leve l of peop le in')( ' d·i stric t is very high
the desc ripti on is not specific; but whe n stati stica l mea sure s like the
perc enta ges·of liter ate amo ng male~ and fema les, the perc enta ges
.
of ·4eg ree hold ers amo ng male s and fema les, and the like are
avai lable , the desc ripti on.beco mes ~xact~
. .
~s
· Data anal ysis faci litat es . iden tific atio n of the. ca_u sal facto
h
und erly ing com plex phen ome na. · Wha t are the facto rs whic
dete rmin e a vari able like labo ur prod ucti vi~y or a·cad emic
perf orm anc~ of stud ents '? Wha t are the relat ive cont ribu tion s of the
c~us ative facto rs? Ans wers to such ques tion s can be obta ined from
stati stica l mult ivar iate anal ysis.
Dat a anal ysis aids -the dra~ ing of relia ble infe renc es from
obse rvat ions . Data are colle cted and anal ysed in orde r to pred ict 111
?r
tnak e infe renc es abou t situa tion s that have ·not been mea sure d
ng
full. Wha t can be the grow th rate of indu stria l prod ucti on duri
a·
the com ing year ? Wha t wou ld be the prob able dem and fo~
ire
part icula r prod uct iri the com ing year ? Que stion s of this kind requ
~rit
pred ictio ns of futu re state s to be mad e on th~ basi s of cu~r
know ledg e. Such pred ictio ns are esse ntial in any strat egi\ decisiort

r 111
oata Analysis and Interpretation
135
elating to man age men t of an-enterprise or the nati ona l eco
nom y or
r s·ocial action forum. The statistical prediction is one of the
functions
~ inferential statisti~s.
Data analysis also help s mak ing estimations ~r generalizat
ions from
the results of sam ple surv eys. Sample statistics base d on prob
abil ity
samples may giv e goo d esti ~at es of par ticu lar pop
ula tion
parameters. Any esti mat e will deviate from the true valu
e due to
sampling error. The. process of statistical inference enab
les us to
evaluate the accuracy of the estimates.
Inferential statistical analysis is ~seful for asse ss~ g the sign
ificance
of specific sample results und er assumed.population conditio
ns. This
type of analysis is called hypothesis testing.
·

Analysis of surv ey or exper~mental data involves esti mat

ing the
values of unk now n para met ers of the pop ulat ion and test
ing of
hypotheses for draw ing inferences. Analysis ~ay be categori
sed as :
1. Oescriptiv~ Analysis : It is largely a stud y of dist ribu
tio~ of
one or·mor e variable. Such ~tudy provides with profiles
of a
business grou p, ·w ork grou p, persons or othe r subjects on
any
of a mul titu de of cha ra_cteristics such as size, com pos ition
,
efficiency, preferences etc. Various measures that sho w the
size
and shap e of distribution alongwith the stud y of mea suri
ng
the relationship between two or more variables are ~vailab
le
from this ~nalysis .
2. · Inferential Analysis : It is concerned with the various
tests of
significance for testing hypotheses in orde r to determine
with
wha t vali dity the .d ata ca~ indi cate som e con clus ion
or
conclusions. It is also con cern ed .with the esti mat ion
of
population vaiues. It is mainly on the basis of inferential anal
ysis
that the task of interpretation is performed.
3. Correlation Ana
lysis : It studi~s that joint variation ·_of two or
more variables for dete rmi ning the amo unt of corr elat
ion
between two or more variables.
136 Business Research. Met hod s (S. ·Y .B.M .S.: SEM -IV)
.
.
dy of how one or
4. Ca sua l An aly sis : It is· con cer ned wit h the stu
iab le. It is a stu dy
mo re var iab les aff ect cha:11-ges in ano the r var
en tw o or mo re
of fun cti ona l rel ati ons hip exi stin g bet we ,
var iab les .
ilit y of com put er
5. Mu ltiv ari ate An aly sis : Wi th · the ava ilab
te ana lys is wh ich
fac ilit ies, the re is a dev elo pm ent of mu ltiv ari~
e mo re tha n two
me ans use of sta tist ica l me tho ds wh ich ana lys
inc lud e:
var iab les on a sam ple of obs erv atio n~. The s~

(a) Mu ltip le Dis ~ri min ate An aly

sis : It is sui tab le wh en the
t can not be
res ear che r has a sin gle.dep end ent var iab le tha
mo re gro ups ·
me asu red , bu t can be cla ssif ied int o two or
thi s ana lys is
on the bas is of som e attr ibu te. Th e obj ect ive of
pos sib ilit y _of
hap pen s to be to pre dic t an org ani sat ion 's
era l pre dic tor
bel ong ing to a par ticu lar gro up bas e~ on sev
var iab les .
(b) Mu ltip le Re gre ssio n An aly sis : It is
· sui tab le wh en the
is pre su_m ed
res ear che r has one dep end ent var iab le wN .ch
var iab les . The
to be a fun ctio n of two or mo re ind epe nde nt
tio n abo ut the .
. obj ect ive of this ana lys is i~ to. ma ke a pre dic
e wit h all the
dep end ept var iab le bas ed on its cov ari anc
con cer ned ind epe nde nt var iab les . .
.)

ult i-A nov a): This

.

(c) Mu ltiv ari ate An aly sis of Va ria nce (M

wh ere in the
ana lys is is an ext ens ion of two wa y AN OVA,
up var ian ce is
rat io of am ong gro up var iab l~ to wit hin gro
·
wo rke d out on a set of var iab les .
be use d in cas e of
(d) Ca non ica l An aly sis : Thi s ·a nal ysi s can
iab les for the
bot h me asu rab le and non -m eas ura ble var
of dep end ent
pur pos e of sim ~lt ane ous ly pre dic tin g a set
th a set of
var iab les fro m ·t hei r joi nt cov ari anc e w:i
ind epe nde nt var iab les .

. . .·.\ii!
nce s by ~ly sin g data.
Inte rpr eta ti:o n o_f d~t a ref ers t~ dra win g inf ere
ver t the sta tist ica l dat a
It can ~e sai d.th at int erp ret atio n hel ps to con
 137
Data Analysis-and Interpretation
. to information. Analysis and interpretation of data are closely
trllated. Anal ysis and inter preta tion of data can be done
1 a t ~s
· tion of data f ac1·1·t
re•multaneously. Analysis and inter preta
::search findings, recommendations and conclusions. ·

1. Consistency : There must be consistency in inter preta tion of

data. If there is more than one analyst draw ing inferences, then
they shou ld ·follo w the same patte rn of anal ysis and
interpretation of data. If there are differences in interpretations, •
the analysts need to discuss·the differences, and best inference
needs to be agree d upon. ·

2. Reliable Data: T}:te data collected must be reliable for draw ing
inferences. The data must be up-to-date. Outd ate9 data must
be discarded as such data do not serve any purp ose for the
purpose of interpretation. , ·.

3. Accuracy : Observations must be correctly perfo rmed ~nd

recorded, and data coll~cted shou ld be accurate. Error s in
observation, experiments or interviews and faulty recor ding of
the information can affect the <;::onclusions drawn.
4. Avoid Bias: The interpretatipn must avoid bias while draw ing
inferences. For_i:r;istance, if the inter prete r is in favq ur of
advertising, he may -interpret the data in favour of advertfsing,
even thoug~ the collected data may indicate otherwise.
5. Homogeneous Data : For analysing and interpreting the data,
homogeneous data is required. For example, if the resea rcher
collects data on brand preference from two or more areas, the
data variables must be the same in ~11 areas. If the data are
collected from three age group s in one area, then the same ·
criterion must be applied !n all areas.
6· Uniformity: There shoul d be uniformity in interp retati on of
data. For instance, if productivity is to be interpreted, the unit
. of measurement of outpu t anq input should be the same, i.e.,
either in terms of units or in terms of value. ·
 ◄
13 8 Bu sin es s Research M eth od s
(5. Y. B. M .S .: SEM
,JV)
· bl e to th e re se
7. Su1"t ab 1"lity : D at a co lle ct ed m us t be su ita
an ce , if st ud y is co nd uc te d on br an d lo arch
pr ob le m . Fo r in st iti ze ns ar e ch os en ctfalty
nk s an d if se ni or -c
of so ft -d ri dr aw n m ay be wronge to
os e, in fe re nc es
co nv en ie nc e pu rp er to co ns um e so ft- dr i~ lts
or -c iti ze ns do no t pr ef
ge ne ra lly se nj us t co ve r r
· ep re se nt at iv e ca Sess.
at io ns m
In ot he r w or ds .
, ob se rv
rt .a t e sa m pz·in g techniq
. Ap pr op
d ra w n fr om a spe,cific universe
r th is pu tp os ~ su ch as st ra tif ie d ra nd om samplin;~ .
ca n be us ed fo 1

D at a : To dr aw pr op er in fe re nc es , th e da ta ·
8. A de qu ac y of . Fo r in st an ce , if on ly 20 to 3Q
ed m us t be ad eq ua e
t·
co lle ct l un iv er se of 10 la kh people,
e se le ct ~d fr om a to ta
re sp on de nt s ar ie nt en ou gh f? r ~rawing
lle ct ed m ay no t be ,su ff ic
th e da ta co e ml !-s t be ap pr op ri at e sample
es . Th er ef or e, th er
pr op er.inferenc tio n of da ta . ·
l:k e m ea ni ng fu l in te tp re ta
si ze to .ml
isa tio n :-G en er al iz in g th e fin di ng s m us t be done
9. Proper General le ge ne ra liz e . th at "H ig h Jo
b
. Fo r in st an ce ,. pe op
ca re fu lly m an ce ' bu t th e re ve rs e is
s to H ig h Pe rf or
Sa tis fa ct io n le ad ce ca n al so le ad to H ig h Job
e th at ,.H ig h Pe rf or m an
eq ua lly tru
Sa tis fa ct io n. ' ·

at io ns m ad e sh ou ld no t be ge ne ra liz ed to ap pl y to
Th e ob se rv Fo r in st an ce , if th e st ud y is
d in th e sa m pl e.
ca se s no t co ve re lty of yo un g fe m al es for
·k no w th e br an d lo ya
co nd uc t~ d -to ~i on s dr aw n fr om su ch a
, th en th e co nc lu
re ad ym ad e ga rm en ts al es , or el de rly la di es , an~
ap pl ie d to yo un g m
st ud y ca nn ot be
d ·i n th e sa m pl e.
so on as th ey ar e no t co ve re
Da ta : Be fo re an al ys is an d in te rp re ta tio n·
10. Proper Pr oc es si ng of ec k if th e· da ta co lle ct ed for .
ar ch er m us t ch
of da ta , th e re se ec t to - ed iti ng , cod.~g,
k is pr oc es se d w ith re sp
re se ar ch w or ,
io n an d ta b~ la tio n. W he n da ta is pr oc es se d pr op er ly
classificat po si tio n to an al ys e an~
ld be :in a be tte r
th e re se ar ch er w ou · ·
t th e da ta co rre ct ly .
•in te rp re

11. O th er Pr ec au tio ns :
re se ar ch in m in d.
• K ee p .th e m ai n ob je ct iv e of
om m or e fu nd am en ta l as pe ct s.
• A na ly si s of da ta m ay ~tart fr
 JIii"'
139.
Analysis and Interpretation
paw .
• Give au~ attention to ~ignificant·questions.
Do not miss '°the significance of some responses, because
•• they are av~ilable from few respondents.

.'
I There are two methods of drawing inferences:
'

\
Inductior{ Method

Deduction Method
.

Induction Method
I •
Meaning:
It is one of the scientific methods. It follows 'the logical reasoning
process. It is a process of reasoning whereby the researcher arrives
at universal generalizations from particular facts~ In other words,
this method involves studying several individual cases and drawing
a generalization. Thus, it involves two elements - observation and
generalization. Conclu_sions drawn from induction are tentative
inferences and they are subject to further co.n firmation based on
more evidence.

When Follow~d:
The induction method is followed when new facts are studied, new
truths are uncovered and new generalizations are formed based on
~ research wo~k. For e?'ample, in a motiv~tional study it was
proposed to test the hypothesis - "The provision of-good working
conditions motivates the .workers, which in turn increases labour
· product~vity." For this purpose~ workers were divided into two
groups - test group and control group. In the test group, the working
conditions were improved significantly and in the control-group,
the working conditions remained the_~ame as before. After a certain
period, the productivity of workers in the_test group was compared

t t
. 140 Bu sin ess Research Me tho ds (S. Y.B .M .S.
. . .
: SE As I
-lVl- V)
wi th tha t of contr<?l gro up , in or de
r to _fin ~ ?u t the im pa ct of th
wo rki ng co nd iti on s on the lab ou r pro
du cti vit y. It wa s fou nd the ~
the pr od uc tiv itr im pr ov e~ in the
~~st gro~up , wh ich pro ve s tl~t r
hy po the sis tha t go od -w ork ing co nd
. . iti on s _mo tiv ate s the wo rke e
en ha nc e pr o d uc ttv 1ty . . . . rs to
· .
. Es sen tia l Co nd iti on s
. ')
!.'1
. . ~
Ac co rdi ng to Cl ov er an d Ba ls} ey, fou f
r co nd iti on s ar~. ess en tia l f ~·\e
va lid ind uc tio n. Th es_e are :
or f
1. Ac cu ra cy : Ob se rv ati on
s mu st be co rre ctl y pe rfo rm ed an
r_e co rd ed , ~n<;l da ta· co lle cte d sh ou d f
ld be ac cu rat e. Er ror s in f
ob s~ ~a ti9 n, ~x pe rim en ts or int erv iew
s an d fau lcy rec ord ing of
the 1n for rp..at ion ca n aff ect the co nc
lus ion s dr aw n. .·
2. Re pr ese nta tiv e: Ob ser va tio ns
mu st co ve r rep res en tat ive cases
dr aw n fro m a sp ec ifi c un ive rse . ·Fo
r ihs tan ce , · to co nd uc t a
su rv ey on br an d loy alt y am on g yo u.r
pr od uc t cat eg o~ y, the n the re sh ou ld a
ig fem ale s for · pa rtic ula r I
be pr op er rep res en tat ion
. of yo un g fem ale s co mp ris ing of co
lle ge go ing fem ale s, yo un g
wo rk in g fem ale s, gr ad ua tes , no
n- gr ad u.a tes , an d oth er 1
. rep res en tat ion s. Ap pr op ria te sam
pli ng.te ch niq ue ca n be use d.
for thi s ·pu rp os e, su ch as str ati fie d ran
do m sam pli ng .
3.. Ap pr 9p :ri ate Sa mp le Siz e:·
Ob se rv ati on s mu .st co ve r an
ad eq ua te nu mb er of ca ses .- ·T he sa
mp le siz e ·m us t be lar ge
en ou g~ _to ma k~ it rep res en tat ive so
as to ob tai n rel iab le r~sults.
4. Pr op er Co nc lu sio ns :· Co nc
ius ion s mu ~t be co nf ine d· to
inf ere nc es dr aw n fro m· the fin din gs
. Th e co nc lus ion s mu st be
dr aw n on ly aft er pr op er an ~ly sis of.
the co lle cte d da ta rel ati ng
to a pa rti cu lar stu dy . Th ey sh ou ld no
t be gel ).e t~l ize d to apply
to ca ses no t co ve red in the sam ple .
Fo r ins tar tce , if the stu dy 15
· co nd uc ted to kn ow the br an d loy
~lt y of yo un g fem ~le s fo r:
pa rti cu lar. pro du ~t, the n the conclu
si<:>ns dr aw n fro m_5 u~ h d
stu dy can.J)Ot be ap pli ed to yo un g ma
les ! or eld erl y lad ies , an
so on , as the y are no t co ve r~d in the
sam ple . ·
 oata Analysis and Interpretation 141
es of Induction ·
TYP ·
: . f. Theod?rson and A. ~eodorso n have identified two basic types
of induction - enumerati ve and analytic.
Eoumerati~~ induction is th~ most common form of induction used
in social science_res~arch. It mvolves generalizations from samples
and the generahza hons are usual~y derived through the analysis of
data. For instance, if in a number of cases, it is observed that young
females are not brand loyal in respect of readymad e garments, then
it can be concluded that all young females are not brand loyal in
respect of reac;lymade garments. Induction by enutnerati ?n gives us
only probable conclusion. . , .
~
Analytic Ind~ction involves case-by-ca.se analysis of specific facts. ·
It is a process of establishing a universal propositio n by exhaustiv e
x enumeration of all the instances of the type covered by the universal
proposition. Therefore; analytic induction can be .called as scientific
I: induction. The scientific indu_ction gives us defini_te conclusions.
]
Donald Cressey has stated a systematic procedure in the case of
analytic induction: · ·
~: • Define the phenomen on to be explained.
• Formulate hypothesi s to examine the phenomen on.
• Study or observe a case to determine whether or not hypothesi s
fits-in the case.
• If the hypothesi s does. not fit the case, either reformula te the
hypothesis or redefine the phenomen on precisely, so that the
case is excluded.
• Study a small number of cases to attain practical certainty; but
whenever a negative case disprovin g the explanati on 1:,
discovered, reformula te the hypothesis.
• Continue the procedure of examining cases, reformulating the
hypothesis and redefining the phenomen on, until a·universal
relationship is established.

.-1. i
 142 _ Business Res~arch Methods (S. Y.B.M.S'.: SEM-IV.)
For purposes of proo.f, examine ca·s es outside the area·
• c~cumscribed by-the definition to detennine whether or not
the final hypothesis applies tot.hem.

Merits
• Induction. provides the µniversal. premise and is helpful in
finding out material truth.
• - It points out the relativity of generalizations. It indicates that a
particular generalization is valid in ~ertain situation.
• It is scientific in nature. Some of the important·laws or principles
, of physical .and sociai". sciences have been developed through
the use of i?ductive appr.oach.
• Inductive generalizations ~re precise and accurate as _they are
based on observed facts and realistic foundation.

·Demerits
-• The collection of data for induction is a complex job~ This
method requires high degree of competence on the part of ~he
researcher. -
• Induction is not useful without deduction. Without deduction, I

induction produces only·a _mass o( unrelated facts. .

• It is a time consuming and expensive process; as lot of time,
effort and money is · required to collect and analyse a large
I volume of data to arrive at generalization. _ ·

II. Deduction Method

Meaning
It is a reasoning pro~ess of applying a general accepted principle to
a specific individual case falling under the general principle. It is
conside_red as reasoning from the general phenomenon to a
particular phenomenon. According to John Best, the deductive
method establi~hes a logical relationship between a major premise,

1
(-
 oata Analysis and Interpretation 143
. and a_~ or pre ~se and thereby arriving at a conclusi.on. A majo
r
1 pren::11se 1s a p~evi~usly es_tablish
ed generalization or assu mpt ion.
A nuno~ prerms_e 18 a _particular case related to the major prem ise,
'fhe logical relationship of the major and minor premises lead s
to a
conclusion, such as follows:
Major Prem ise
• All adul ts are eligible to vote
Minor Prem ise · Mr. Sam is an adult.
Conclusion
• Mr. Sam is .an adult, and therefore eligible
to vote. ·
Use
The deductive meth od of applying ·a general assu mpt ion to
the
specific application is useful for ~olving certain problems or deal
ing
with cert ain situ atio ns. Dep end ing upo n the prem ises
or
assumptions, all deductive reasoning is either valid or invalid..The
main task of deductive ·reasoning is to clarify the nafure of relat
ion
-. betweep. premises and conclusions in valid •arguments. Ded ucti
on
,;_ can give conclusive evidence. However, it is not useful in arriv
ing
·· at new truth s. ·
I
Essential Conditions
The following are the essential conditions for valid deduction:
. .
- • The gene ral prem ise ·or assu mpt ion mus t be correct. If the
general prem ise is correct, then the conclq.sion draw n can be
right.
• The general premise mus t be applied only to .the cases that
properly con:ie und er it.

Merits
• · Deduction is relatively simpl~ and less time cons umi ng
as
compar.ed to inductive method.

• It is precise and accurate in generaliz~tion as it makes use of

( logic and mathematical tools of analysis.
I

\
 f~J
144 Business Research Methods (S. Y.B.M.S.: SE!v
is lim ite d Sc op e /
• In so cia l sc ien ce s, wh ere the re
s the on ly me tho d for til'
· ex pe rim en tat ion , thi s me tho d be co me e
de ve lop me nt of ge ne ral iza ti~ ns.

De me rit s
tio ns, if the pre mi ses are
• It ma y res ult in ina ccu rat e ge ne ral iza
inc orr ect or pa rti all y correct.
• Th is me tho d is ab str ac t. Th ere for
e, !f
a _large am ou nt of
iza tio ns , the n s~ch
ab str ac tio n ·we re us ed in ge ne ral
ge ne ral iza tio ns wo uld be useless.

-in fer en ces by an aly zin g the

Int erp ret ati on .of _cj.ata inv olv es dra wi ng
d~ ta are clo sel y interlinked.
da ta. Analysis an d int erp ret ati on of
sis of da ta oft en inc lud es sim ult an eo us int erp ret ati on of
An aly
· ·
results.
'
of res ear ch wo rk lar ge ly de pe nd s on the qu ali ty of
The suc ces s
it m us t~ pro ces sed in terms
interpretation. After ~ollecting the dat a;
, cla ssi fic ati on an d tab ula tio n. Th e pro ce sse d da ta 1s then
of-coding
lys ed, an d fro m the ana lys e,d dat a, the res ear che r dra ws inferences.
ana
efly-stated as fol low s:
The im po rta nc e of int erp ret ati on is bri
1. Recommendations:
ke rec om me nd ati on s, .
Int erp ret ati on of da ta he lps to ma
an d soc ial res ear ch . For
especially in the cas e of co mm erc ial
to fin d ou t inc ide nc e' of
ex am ple , if a res ear ch is co nd uc ted
res ea rch er wi ll dra w
po ve rty .in ba ck wa rd are as, the n the
ta, an d th en pr ov ide
inf ere nc es ba se d on an aly se d da
om me nd ati on s to ov erc om e the pro ble m of po ve rty . Major
rec
de bt an d so cia l e_vils
ca us es of po ve rty ma y be inh eri ted
cas e, the res ea rch er may
(ga ~b lin g an d dri nk ing ). In su ch a
est me asu res to red uc e the bu rde n of inh eri ted de bt and
su gg
ls, wh ich in tur n ~il l help
~o ov erc om e the pro ble m of social evi ·
to red uc e po ve rty .

d
I

:,oata Analysis and Interpretation 145

I
)z. · ·Decision :M;aking: .
Data interpretation helps in decision making. Fo·r example, a
research may be conducted to find out the causes of declin~ in
sales. The analysis and interpretation of data may indieate the
main cause of decline in sales, i.e., poor ·after-sale-service.-
Therefore, the management may take suitable decisio!ls to
improve after-sale-service such as:
• · · Training to- the after-sale-service staff.
' .
• Opening of more service centres.
• Appointment of _competent staff.
• M~nitoring custoµ1er feedback on·~egular basis.
. .

Forecasting: Data interpretation helps in forecasting the

trends:
-• In the case of social research, the trends in the growth rate
of population, literacy rate, income levels, etc., can be
forecasted. ·
• In case of commercial research, the trend in sales, profits,
market share, etc:, C<Ul be forecasted.
Based on forecasts, necessary action can be taken by the
concerned authorities.
4. Development of Models:
Data interpretation helps to develop new models: For instance,
a re~earch on brand loyalty can help to develop a new model
on brand loyalty.. The. model can be used for academic arid
commercial purposes. For instance, the AIDA (attention,
interest, desire, action) model can be used.by advertisers and
marketers. This model indicates that the customers attention '
must be attracted, interest must developed in the minds of
customers, strong desire must be created, and finally action can
be induced, i.e., purchase or acceptance of idea.
5. Development of Hypothesis:
A pilot study can help to develop hypothesis. The in_
te~pretation
of pilot'study enables the rese~rcher to modify the preliminary
. '

. \.
 ◄
146 Business Research Method s (S.. Y.B .M.S.: S£1iK
iv1~1V)

hypot hesis and accord ingly a new-hypoth esis can be develop

to proce ed with the resear ch activit y. · ect

6. · Valid ates Theor y:

rnterp retatio n of data may help to vali~ate_ or invalid ate the
theory . For instan ce, the resear ch may mdica te wheth er th
presen t theory is releva nt or not based on the inferen ces draw e
~om the resear cl) work. In other words , !he infer~ nces_ ma;
either suppo rt the curren t theory or prove it otherw ise.
For · examp le, i_t is norma lly assum ed that peopl e are poor
be~au se they are unemp loyed, but the resear ch may indicate
that people are -poor not becau se they are unemp loyed but
becau se they are addict ed to liquor and gambl ing, and also due
to the proble m of inheri ted debt. .
7. Futur e Refere nce:
The InfeFences drawn for a partic ular resear ch activit y can be
used for future referen c~. Furthe r· studie s can be condu cted_
based on t~e conclu sions drawn from the earlie r resear ch

, 8.
activit y.'
Motiv ation to the Researcher:
Prope r interp retatio n will genera te prope r recom menda tions.
01_1 the basis of recom menda tions, effecti ve decisi ons can be
taken. by the. organi sation . If there are excell e.n t results , the
:resea rcher/ interp reter may be rewar ded w"ith additi onal
incent ives.

I. Signi fican ce of Editi ng

1. Accur acy of Data :

Editin g checks the accura cy of data collect1.on. At times; the
respo ndent may provi de incorr ect. respo nses to certai n
►
· 147
oata Analysis and Interpretation
questions. The editor uses his judgement to correct the
jnconsistendes in responses given by some !espondents. For
instance, a respondent belonging to middle class may respond
that he buys premium priced products, which is not a possibility
·for a middle class consumer. Therefore, the editor may correct
the wrong responses and as such ensures the accuracy of data.

2. Uniforniity in Responses :
Editing ?lso helps to find out whether or not the respondents
have interpreted the questions uniformly. ·For instance, th~
question may have a scale of 1 to 5 where l == ex·cellent and
5 = poor. However; some respondents may have interpreted in
the other way.' In such·cases, the responses are defective.
Therefore, the editor checks into the uniformity in responses of
all respondents and make necessary changes.
3. Completeness of Data:
Editing en~ures completeness.of data. At times, the respondents
may provide response~ only to certain questions and not to all
questions. Also, the responses to certain questions may be
incomplete, especially in the case of open ended questions. The
editors may use judgement to deduce proper answer to
unanswered or ~complete responses based on other responses.
4. Coding of Data :
. .

Editing facilitates coding of data at the post data collection stage.

After editing the data, the researche_r assigns codes to the
responses provided by the respondents, especially in. the case
of open ended questions. In the case of open ended questions,
all responses are placed in different categories and each category
'is assigned a code.
For instance, the responses to a ·question -what are the buying'
motives for pu!chasing a car- may include status in'the society
(Statu_s - code 1), convenient mode of travelli~g to family
members (Convenience - code.2), claiming of depreciation in.
tax returns (depreciation·- code 3), and so on. Therefore, 'there
Would be at least three codes.
 <;
148 . . Business Research Methods (5. Y.B.M.S.: SEM-IvJ
. .
II. Signifi_cance of Co di ng · ·

1. Faci_lita tes Cla ssi fic ati on of Da ta:

er pro vid ing Codes
Co din g fac ilit ate s cla ssi fic atio n of dat a. Af~
ssi fie d int o various·
to var iou s res pon s~s , the dat a can· be cla
cla ssi fie d int o categories
' cat ego rie s. Th e cod ed res pon ses can be
om e lev el, are a wise, ,
suc h .as age , gen der , edu cat ion al lev el, inc
occ up ati on wis e, an d so on.
2. Fa cili tat es Ta bu lat ion of Da ta ·:
a, it 'be com es easier
· Sinc;e cod ing fac ilit ate cla ssi fic atio n ·of dat
e cod e res pon ses are
for the res ear che r to tab ula te the ·da ta. ·Th
ord ing ly the dat a is
cla ssi fie d int o dif fer ent cat ego rie s, an d acc
ted dat a can the n be
tra nsf err ed to sta tist ica l tab les . Th e tab ula
use d for ana lys is and int erp ret ati on.

III . Sig nif ica nc e of ·c las sif ica tio n

_
1. Protection and Ma na gem ent of J?ata :
cre ate d an d unt il it is·
Fro m the tim e inf orm ati on·is col lec ted or.
it -i s pro tec ted , stored
des tro yed , it sho uld be cla ssi fie d to ens ure
I

inf orm ati on ma y be

\

and ma nag ed app rop ria tel y. Fo r _inst anc e,

on for int ern al use
1 cla ssi fie d as -public inf orm ati on, inf orm ati
on. Th e inf orm atio n
onl y, and con fid ent ial / -restricted inf orm ati
pro per ly._Th e public ·

V nee ds too ~ pro tec ted , sto red and ma nag ed

. informatioi:,. can be pro vid ed to any on e ...:
ins ide rs and outsiders,
the inf orm ati on tha t is cla ssi fie d as 'in ter
no t be pro vid ed to ou tsi der s, a~ d the
nal 1;1se on ly' sho uld
inf orm ati on tha t is
tric ted on ly to top
cla ssi fie d as con fid ent ial ' ma y be res
I

aut hor itie s_in the org anj sat ion . ,

I
I

2. Facilitates Ta bu lat ion of Da ta :

tio n of dat a. In fact
Co din g and cla~sification fac ilit ate s tab ula
nt of classification.
cod ing is con sid ere d as an im po rta nt ele me
eit her at the pre -da ta
Th e res ear che r ass ign s cod es to re.s po nse s
lec tio n sta ge ,· and
col lec tio n sta ge or at the po st ·da ta col
o dif fer ent categories.
acc ord ing ly the res pon ses are cla ssi fie d int
pro pe r ana lys is an_d
Th e cla ssi fie d inf orm ati on is tab ula ted for
int erp ret ati on. · ·
 149
Analysis and Interpretation
vata '
Facilitates Speedy Searches of Data : · ·
3
· The classified data is easy to locate and to retrieve. The classified
data helps in speeding up data searches. For instance~ the
classified data can help an organisation to retrieve quickly
certain data which may be required for legal and regulatory
requirements.within a set time-frame.

4. Grouping of Data :
Classification facilitates grouping of data into different
categories. 1:he c~llected d~ta can be grouped in different
categories such as age, gender, education, etc. Each of the
categories can be further sub-classified.
For instance, the age group can be further sub-divided into
children (upto 12 years),·teenagers (13 to 19 years), young adults
. (20 to 39 years) middle-aged (40 to 59 years) and senior citizens
(60 years and above). Such grouping of data facilitates easy
understanding of the differences of the different groups in terms
of tastes, preferences, iikes and d~slikes, etc.

IV. Significance of Tabulation

1. Analysis and Interpretation of Data : .

Tabulation helps to arrange the classified data into statistical
tables. The statistical tables facilitat~ analysis and interpretation
of data. The tabulated data can facilitate comparative analysis ·
of two or more variables - such as different age groups, income
groups, different states, different periods, and so on. For ·
ins_tance, the tabulated data can be analysed in terms of buying
patterns of different age groups, in~ome groups, ~d so on.
2. Basis for Writing Research Report : .
· Tabulation provides the basis for writing research report. Based
on the tabulated data, the researcher ·c an analyse and interpret
the data. The analysed data in turn enables the researcher to
_draw conclusions and accordingly make recommendations. The
statistical tables, conclusions and recommendations form an
important part of research report. ·
 150 Busin ess Research Metho ds (S. Y.B:M .S.: Sf),K
. ·. . iv1~Jv)
3
· Corr elati on betw een Vari ables : .
. Tabu latio n facili tates c~rre latio n betw e~n two o~ more .Variables
, For exam ple,_data on inco me _and sav~ g habi ts place d in th~
. sam e table help to draw certa m conc lusio ns abou t the inoo
leve l and savin g habi ts of certait:t class es of peop le. Genera~e
high er the in~om e, high er ~ou ld be the savin gs, but suchy,
?en~ ral as_sumptiori may diffe r am~ ng diffe rent age group s an~
m diffe rent areas . . ·· ·

·4. Dete cts Errors in Cod ing and Clas sifica tion·:
Tabu latio n may help to find out erro rs in codi ng and
class ifica tion of data. For insta nce, certa in code d and classified
data may not .fit in the statis tical t~~le s._Ther efore , this may
requ ire chan ges in codin g and clc1ssification of data. . .

5. Ease in Unde rstan ding of Data : .

Tabu latio n helps the resea rch~ r to deter mine and communicate
the findi ngs in a form whic h can be easil y unde rstoo d by others.
For insta nce, the tabu lated ·data may indic ate high literacy in
II .
: one State as comp ared to anoth er. Ther efore , one can easily
unde rstan d that the form er State is -m ore.l itera te than the latter.
6. Facil itate s Loca tion of Spec ific D~ta :
Tabu latio n help s to· loca te spec ific data requ ir~d by the
resea rche r. For exam ple, cens us data prov ides a weal th of
geog raph ic and demo grap hic data, but.a·resea rcher migh t need
only certa in segm ents of the data from certa in locat ions. This
speci fic data abou t certa in segm ents from certa in ioca~ons
(reg ions /stat es/ distr icts) can be easil y iden tified from the
s~tis tical table s - with refer ence to dens ity of popu latio n, gender
ratio , life expe ctanc y, etc.
7. Supp orts Wri_tten Matt er :
Stati stica l table s sq.pp ort writt en·m atter . The writt en matte r gets
_more impo rtanc e due to statis tical table s. Th'i s may be because;
statis ti_cal table s gives a good feel of the writt en matt er; and
. also easy to unde rstan d. .
 151
oata Analysis and Interpr~tation
v. Sigriificance of Graphic Presentation·

Communication : .
1. · Quick
The graphs and charts can communicate the information at a
glance. It does n~t take much time to read and understand the
message. One can easily understand the.data presented in the
bar charts or pie diagrams, graphs and so on. For instance, a
graph can indicate·at.a glance the trends in sales over a period
of time, either increasing or decreasing or showing a mixed
trend.
Graphs ·enable to present information easily and quickly. At
times, one needs to get certain information or point across fast.
This -cannot be done·when trying to explain hundreds of lines
of data. Therefore, graphs are essential in these scenarios. With
graphs, one can more easily explain the information clearly and
quickly. .

2. Effective Appeal :
The graphic presentation may have an effective appeal to the
readers. For instance, the pie diagrams, bar charts, graphs, etc.,
can be illustrated with the help of effective colours. The colours
in the graphs or charts easily attract a,ttention and may create a
good impact on the mind of the readers with special reference
to understanding the data.
3. Condenses·Large Volumes of Data: ·
Graphs and charts condense large amounts of information into
easy-to-understand formats ..The graphs or charts can be
expressed in terms of frequencies, percentages, ·or some ·other
variables. For instance, .a pie. diagram ·can condense the sales
data of various brands of a firm in the form of percentage of
sales of each brand in the total sales basket of a firm.
4. Educative Value:· ·
Graphs provide educative value to the audience. For ii-lstance,
graphs and charts can be used in training sessions'. It is more
visually appealing to show a colourful graph then·to explain
with hundreds of pages of raw data. Graphs and charts are not
 ◄
152 . ~usiness Research Methods (S. Y.B.M.S. : SEM~IV) ·
onl y. app eal ing but cre ate a good memory impact on the
au d 1ence.

pro ve. or disprove the

The basic logic of hypothesis testing is to
quantitative research;
research question. When a researcher conducts
que stio n or hypothesis
·he /sh e is at_tem pti ng.. to ans we r a research
lua ting this research
tha t has been formulated. One me tho d of eva
s tes tin g, which is
que stio n · is via a pro ces s cal led hyp oth esi
ing.
som etim es also referred to as -significance test

Ex am ple :
t the y use the best
Two iecturers, San dy and Ma ndy , ~hink tha
50 statistics students
method to teach their stud~nts. Each lecturer has
em ent . In Sandy's
wh o are stu~ying a gra dua te degree in ma nag
seminar class every 1
class, students have to attend one lecture and one
y hav e to attend _one
· week, whilst in Ma ndy 's class stu den ts onl
wit h lec tur es are an
lecture. Sandy thin ks that· sem ina rs alo ng
ieves tha t lectures are
important teaching me tho d; whilst Ma ndy bel
den ts ·are better off
/ I
sufficient by themselves and thin ks tha t stu
) time. This is the first
solving problems by themselves in their ow n
the y tak e up a lot of
year .that Sandy has given seminars, but since
not wa stin g her time ,
her time, she wants to ~a ke sur e tha t she is
formance.
. and tha t seminars imp rov e the stu den ts' per

The Research Hy pot hes is

ear ch hypothesis . In
The first step in hypothesis testing is _to set a res
e the effect that two
Sandy and Ma ndy 's stu dy, the aim is to examin
lectures and seminar
different teaching me tho ds - pro vid ing bot h
themselves (Mandy)
classes (Sandy), and pro vid ing only lectures by
re specifically, they
- had on the performance of the stu den ts. Mo
different between the
wa nt to dete_rmine wh eth e_r per for ma nce is
is skeptical about
two different teaching me tho ds.- Whilst Ma ndy
believes tha t giving
th~ effectiveness of sem ina rs, San dy clearly
den ts do better than
seminars in addition to lectures hel ps the stu
fol low ing research
tho se in Ma ndy 's class. This lea ds to the
hypothesis :
 oata An~lysjs and Interpretation
153
Research Hyp oth esis : When students attend seminar
classes, in
addition to lectures, their performance increases..
·
. .
By taking a hyp othe sis testing approach, San dy and Man dy :wan
t to
generalize thei r ~esult~ to a pop ulat ion (total studei:ts) rath
er th~
just the stud ents m their sample. However,·in order to use hyp
othesis
testing, one nee ds to re-state the res~arch hypothesis as a
null and
.I alternative hypoth~sis.
·,
.
: Null Hypothesis: The null hyp othe sis (H ) is a hyp othe'
0 sis whi ch
: · the researcher tries tq disprove~ reject 'or null ify. A null hyp othe sis
1
is "the hypothesis that there is no relationsp.ip between two
or mor e
· variables, sym boli zed as H . · ·
0
. ,
Alternative Hypothesis: The alternate, or research, hyp othe
sis
proposes a relationship between two or mor e variables, sym
boli zed
·. asH1.

: Decision Errors

Two types of errors can resu lt from a hypothesis test.

I .

Type I error : A Typ e I error occurs whe n the researcher rejects

a
. null hypothesis when it is true. The probability of com
mitt ing a
: Type I erro r is called the significance 'level. This probabil
ity is also
: called alpha-, and is often denoted by a.
~

: Type Il error: A Typ e II error occurs whe n the researcher fails .

to
· reject a null hypothesis, which is false. The probability of committ
ing
a Type II erro r is called Beta, and is often denpted by
p. The
probability of not committing a Typ e II error is called the
Pow er of
~~~
.

1. Iden tify the Res~arch Pro blem :

The first step is to state the research problem. The rese
arch
problem nee ds to identify the pop ulat ion of interest (uni
verse
of researc;h), and the variables und er investigation.

............
 • SE""
154 Bus ine ss Research Me tho ds (5. Y.B .M. S " iVl- f\l)
d ou t the eff ect iv
Ex am ple of Re seahrch Pro ble m: To fin r . eness 0If:
hv o t h. tho d, an d lec tur e-cu ,
. eac ing me_t _o~ s - on ly lecture me
m ma rks of the studen7s...
sem ina r me tho d - wi th reference_ to exa
lat ion of int ere st refe
In the abo v~ res ear ch pro ble m, the po pu
e the tea chi ng metho~s
to the stu de nts , an d the var iab les inc lud s
an d the ma rks . ·

nly def ine wh at is to be

Th is ste p ena ble s the res ear che r no t ·o
us ed in sam ple data
tes ted bu t wh at va ria ble (s) wi ll be
er cat ego ric al, discrete
co~ ect ion . Th e typ e of variable(s), wh eth
ica l tes t wh ich can be 3,
or co nti nu ou s, fur the r def ine s the sta tist
- pe rfo rm ed -o n the_col lec ted dat a.
Hy po the sis :
2. Sp eci fic the Nu ll an d Alt ern ati ve
co nv ert ed int o a nu ll
Th e res ear ch pro ble m or qu est ion is
sis . Th e hy po the ses are
hy po the sis an d an alt ern ati ve hy po the
lly exc lus ive . Th at is, if
sta ted in suc h a wa y tha t the y are mu tua
on e is tru e, the oth er mu st be false. ·
sis (H0) is a sta tem ent
(a) Nu"Il Hy po the sis : A nu ll hy po the
du e ton cha nce ." It
tha t dec lar es the ob ser ved -dif fer enc e is
rej ect or disprove.
I- is the hy po the sis the res ear che r ho pes to
is no rel ati on shi p
A nu ll hy po the sis sta tes tha t: the re
sim pli sti c def ini tio n
I . be tw een-tw o or mo re var iab les . Th e
ati ve hy po the sis
of the nu ll is -as the op po sit e of the alt ern
(H1)- ,
en the tw o me tho ds
Example : ,'/There is no dif fer enc e bet we
tur e-c um :.se mi nar
of tea chi ng (on ly lec tur e me tho d, an d lec
de nts ."
me tho d) on the sco rin g of ma rks of stu
ate hy po the sis pro po ses
(b) Alt ern ati ve,H ypo the sis : Th e alt ern
iap les , sy1:1-bolized
a rel ati on shi p be tw een tw o or mo re var
as H1 .
tho d im pro ve s the
Example : "T he lec tur e-c um -se mi nar me
are d to the on ly
sco rin g of .m ark s of stu de nts as co mp ·
·
lec tur e me tho d."
ose to tes t _m us t be
"N ote tha t the tw o hy po the ses we-pr op st
i.e. , wh en on e is tru e the oth er mu
mu tua lly exc lus ive ;
 155
ta Analysis and Interpretation . . h ·
Da . . b exhaustive, t ey
be false. And we see t~at they must e" - Rosenthal &
must include all possible occurrences.
Rosnow
th is is a hypothesis
From the above, it is clear ~hat the null h ypo . es othesis is
of no difference. The main problem of testing of hyp t·
h . The alterna ive
to accept or to reject the null hypot esis. h two
hypothesis specifies a definite relationship between t_ et th
variables. Only one alternative hypothesis is tested agams e
nuli hypothesis. ·

3. Significance Level : · t
After formulating the hypotheses, the rese_a rcher m~s
determine a certain level of significance. ·The confidence with
which a null hypothesis is accepted or rejected depends on the
level of significance.
Generally,_the level of significance.falls between 5% and 1 % :
A significance level of 5 %.means the risk of making a wrong ·
• decision is 5%. This means, the res~archer may make a
wrong decision in accepting a false hypothesis or in
rejec_ting a true hypoth~sis by 5 times out 100 occasions.

• A significance level of 1%means the risk of making a wrong

decision is ·1 %. This means the researcher may make a
wrong decision in accepting a false hypothesis or in
rejecting a true hypothesis is once out of 100 occasions.
Therefore, a 1 % level of significance provides greater
confidence with which null hypothesis is accepted or
rejected as compared to 5 % level ?f significance.
4. Test Statistic :
A statistic used to test the null hypothesis. The researcher needs
to identify a test statistic that can be used to assess the truth of
the null hypothesis. It is used 'to test whether the null hypothesis
set up should be accepted or rejected. ·

Test statistic is calculated from the collected data. There are

different types of test statistics. For.instance, the z statistic will
compare the observed sample mean to an expected population
mean ~- Large test statistics indicate data are far from expected,
 156 B.usiness Research Methods (S. Y .B.M .S.: SEM 1
in fav ~ VJ
pi
pro vid ing evi den ce aga ins t the nul l hyp oth esis and
. or of
the alte rna tive hyp oth esis .
I Eve ry test in stat isti cs ind icat e th~ sam e. Bas ed
on the sa:rn
be observPle
dat a, it giv es the pro bab ility (P-Value) tha t can
dat a are v:d,
Wh en the P-v alu e is low , it me ans the sam ple
esis is wronry
sign ific ant and it ind icat es tha t the nul l hyp oth
lect ed dat a a.g.
Wh en the P-v alu e is hig h, it sug ges ts tha t the col re
w1 ·th·m t h e nor ma1 ran ge.
. . '

:
5. Reg ion ?f Acceptance and Reg ion of R~jection
test statistic
The reg ion of acc ept anc e is a ran ge _o f val ues : If the
oth esis is not
fall s wit hin the reg ion of acc ept anc e, the nul l hyp
t the chance
reje cted . The reg ion of acc ept anc e is def ine d so_tha
(a) level of
of ma kin g a Typ e I erro r is equ al to the Alp ha
sign ific anc e.
Typ e I error - A rejection of a .true nul l hyp oth esi s
e is call ed the
The set of val ues out sid e the reg ion of acc ept anc
the reg ion of ·
reg ion of reje ctio p. If the test stat isti c falls wit hin
cas es, we say
reje ctio n, the nul l hyp oth esis _is reje cted . In suc h
, tha t the hyp oth esis has bee n reje cted at the Alp
sign ific anc e. . ·
ha (a) level of

g to take in
r [Al pha (a) - the pro bab ilit y the researcher is wil lin
fal sel y rejecting a true nul l hyp oth esis .]
6. . Sel ect an Ap pro pri ate Tes t : . .
. Wh et~ er the
A hyp oth esi s test ma y be_one -tai led or two -tai led
tive hyp oth esjs
. tes t is one sid ed or 2 sid ed dep end s on alte rna
and the nat ure of the pro ble m. ·
ion of reje_ction
A tes t of a stat isti cal hyp oth esis , _wh ere the reg
,-is call ed a one-
is on onl y one sid e of the sam plin g dist ribu tion
stat es that
tail ed test . For exa mp le, sup pos e the nul l hyp oth esis
ive hyp oth esis
the me an is less tha n or equ al to 10. The alte ma~
ion of rejection
wo uld be tha t the me an is gre ate r tha n 10. The reg
the righ t side
wo uld con sist of a· ran ge of num ber s loc ate d on.
s gre ate r than ·
of sam plin g dis trib utio n; tha t is, ~ set .of num ber
10. .
Data Analysis and Interpretation 157
In sim ple wor ds, in one-tailed test, the test statistic for reje
ction
of null hyp othe sis falls only in one side of sam plin g dist
ribu tion
curve.

Reject H0
One-tailed Test
Example: A refrigerator com pan y claims that the
mea n life of
its new refr iger ator m~del is 30 years. Now the
rese arch er
formulates the hypothesis the refrigerator life is 3.0 yea
rs.
A test of a .statistical hypothesis, where the regio~ of
rejection
is on both side s of the sam plin g distribution, is call
.tailed test. For example, sup pos e the null hypothesis
ed a n:vo-
states that
the mea n is equ al to 10. .The alternative hyp othe sis
wou ld be
that the mea n is less than 10 or greater than 10. The
·regibn_of
rejection wou_ld consist of a range of num bers loca ted
on bot h
sides of. sam plin g .distribution; that is, the region of
rejection
would consist part ly of num bers that wer e less tha
n 10 and
partly of num ber s that wer e greater than 10.
In simple wor ds, in two-tailed test, the test statistic for
rejection
of nul l hyg oth esis falls on bot h the tails of the
sam plin g
distribution curve, as sho wn in the diagram:

Reject H0 ' Reject H0

Tuo-tailed Test
 ess ~esearch Methods (S. Y.B.M.S.: SEM.-JvJ
158 Busin
n to apply one-tailed and two-tailed tests?
Whe
y on e- ta ile d te st o t tw o-tailed test depends on
Whether to appl ne -t ai le d te ~t is used when th
le m . O
th e n a_ture of ·the prob il y on on e si de of the issue. e
is pr im ar
researcher's interest
t sa le s pr om ot io n sc he me more effective
E~ample : 'Is the curren otion s~heme?' _. ·
pr op os ed sa le s pr om
th an the
is ap pr op ri at e w he n the researcher has no
A· two-tailecf test ·
e issue.
focus on one side of th
o m ar ke ts ·..: M um ba i an d Chennai _
Example : 'Are the. tw unch a new product?'
t from ea ch ot he r to la
differen
st is that a product is m
arketed
e of tw o ta ile d te
Another exampl m ot io n sc he m es - one offering a
le s ,p ro
with the help of two sa a di sc ou nt to promote the same
he r of fe ring
free gift,·and the ot T hi s is a tw o- ta iled test, because
ar ke ts .
brand in two different m o sc he m es used for.promoting ·
th at 'th e tw
the null hypothesis is
significan~ly.'
the brand do not' differ
riterion:
7. Select the Test C
' .

ed s to se le ct th e te st cr iterion. If the pothesis

hy
The researcher ne y le ss th an 30 ,_th e t-test is used.
pl e sa
pertains to a smaller sam la rg er sa m ple say more than 30,
er ta in s to a
If the hypothesis ·p ·
the z-test is used.
g or fa il in g to re je ct the null hypothesis ·:
8. Rejectin
al ly to th e qu es tio n of whether we reject or fail
Let's return fin
esis.
to reject the ·null hypoth
sh ow s th at th e sign ificance level is
If our statistical analys
is
lu e w e ha ve se t (e .g ., either .05 or .01), we_
be~ow the cut-off va ac ce pt th e al te rn ativ e hypothesis.
an d
reject the null hypothesis e le ve l is ab ov e the cut-off.value,
gn ifi ca nc
Alternatively, if the si po th es is an d ca nn ot accept the
ll hy
we fail to reject the nu ou ld no te th at you cannot accept
is . Y ou sh
alternative hypothes fi nd ev idence against it.
th es is , bu t on ly
the null hypo
 oata Analysis and Interpretation 159

'~ ~ ,
The types of hypothe sis testing can be broadly_divided into two
gi:oups:
Parametric Tests
•
Non-Parametric Tests
•
In the literal meaning of the terms, a param~tric statistical test is
one that makes _as~~mp tions about the _parame ters (~efinin g
properties) of the p(?pulation distribution(s) from which one's--dqta
are drawn, while a non-parametric test is one that makes no such
assumptions.

Parame_tdc Test
It is a statistical test that depends on an assump tion about the
· distribution of-the data, that the data are normally distributed. When
considering a normal distribution of a populati on th_ese features are
known as parameters. Parametric analysis relies on the data being
normally (or nearly) distribu ted so that an estimat ion of the
undedying ·po.pulation' s parameters can be .made. These can then
be used to test the null hypothesis. As only quantita tive data can
have a normal distribution, if follows that p·arametric analysis can
only be used on quantitative data. · ·
Provideq they are appropriately used, parametric tests derive more
information about the whole population than non-parametric ones.

. N_on-Parametric Tests
The non-parametric tests of the null hypothesis .do not assume any
particular distribution for the data. Instead they look at the category
or rank order of the values and ignore the absolute difference ·
between them. Consequently non-parametric analysis is used on
nominal and ordinal data as well as' quantitative data that are not
normally (or nearly normally) distributed.
Non-parametric an~lysis can be used on any data but paramet ric
analysis can only be used when the data are normally distributed.
. .
 5
160 Business Research Methods (S. Y.B .M. S.: EM-rv)
,_ it is mo re like ly to
If a dif fer enc e ex~ sts bet we en the stu dy gro ups
ore im po rta nt to kno w
be fou nd usi ng par am etr ic tes ts. It is the ref
ute d. Yo u can som etim es
for cer tai n if the dat a are nor ma lly dis trib
cur ve of the plo t~e d dctta. -
d~ ter mi ne thi s by che cki ng the dis trib uti on
sho w ho w pre cis ely the
A mo re for ma l wa y is to use a com put er to_
·
da ta fit wi th a nor ma l dis trib uti on.
em pts ~re oft en ma de to
.W hen dat a are no t nor ma lly dis trib ute d att
be car r~e d out . _The mo st
tra nsf orm it so tha t par am etr ic ana lys is can
orm ati on. .Th is_has the
com mo n me tho d use d is log ari thm ic tra nsf
me ans an d con fid enc e
ad de d ady ant age of all ow ing geo me tric
e uni ts as the orig ina l
int erv als to be cal cul ate d tha t hav e the sam
dat a. ·
on for par am ~tr ic tests
Th e dif fer enc e is bet we en nor ma l dis trib uti
me tr1c tes ts. Th e sam ple
and no t nor ma lly" dis trib uti on fo;r _no n-p ara
par am etr ic tes t is use d.
siz e als o pla ys a rol e i.e. if sm all num ber non
l be mo re nor ma lly
Sin ce wi th lar ge nu m~ er of · sam ple ( wil
am etr ic te~ ts cou ld be
dis trib ute d and var iati on wil l be les s and par
uti on of dat a are ma jor
use d. Th us, sam ple siz e and nor ma l dis trib ·
fac tor s. . · .

z TE ST , t TE ST AN D f TE ST
e int erv als . Fo r mo st
t, z, and f tes ts are use d to cre ate con fid enc
all y, the re is alw ays a
eve nts tha t can be des cri bed ma the ma tic
rgi n of err or is wh at a
_m arg in of err or. 'The me asu re of the ma
·
con fid enc e int erv al des crib es.
Th ese tes ts are use d fo:r; _dif fer ent sam ple s.
sam ple siz e is less
• Th e T tes t is gen era lly use d wh en the tot al
tha n aro und 30. · . · · .

• Th e Z tes t wh en it is abo ve 30.

t is use d to ana lyz e
• , Th e F tes t can be use d for eith er, bu t thi s tes .
·
the rat io o! the var ian ce of 2 sam ple s.
 161
. Analysis and Interpretation
oata '
zfEST
___, - , o ulation ve~sus a
st is used for testing the ,mean of a p Pl . with large
A z- te . f tw popu attons,
standard, or comparmg the means o . o o ulation standard
. ( >30) samples whether or not one knows the P P . of some
n- •ati· on It is also used for testing the proportion . . g the
dev1 · • comparm
racteristic versus a standard proportion, or
, cha .
proportions of two popu1ations.
Example: Comparing the average salaries . of m ale employees versus·
female employees in a production u~t.
Exa;,,ple: Comparing the fraction defectives from two manufacturing
futes.
z-test is a statistical test where normal distribution is applied and is
basically used for dealing with problems relating to large samples
when.n~30.
n = sample siz~
For example suppose a person_wants to test if both tea an¢ coffee
are equally popular in a particular town. Then he can take a sample
of size say ·500 from th~ town out of which suppose 280 are tea
drinkers. To tes~ the 4ypothes1s, he can u_se Z-test.

Z-Test's for Different Purposes

There are different types of ~~test used for different purposes. Some
of the popular types are outlined below:

1. z tes~ ~or single propor~on is used to test a hypothesis on a

specific_vaJue of the population proportion.
Stat_istically speaking, we test the null hypothesis Ho : p =
against t~e alternativ~ hypothesis H1 : p > < Po where p is tK~
popula~1on _propo~hon and Po is a specific value of the
population proportion w~ would like to test for acceptance.
~e ex~ple on tea drinkers ~xplained ~hove requires this tes
r that e~ample, p0 = 0.5. Notice that in this particular exam :·
P oportion refers to the propo~tion of tea drinkers. P1 '
 162 Busines s Research Method s (S. Y.B.M. S.: SE"A
lVl-f\,1)
.
2. z test for differe nce of propo rtions is used to test the hypoth e .
-that two popul ations have the same propo rtion. sis

For examp le suppo se one is intere sted to test if there i~

signif icant differe nce in the habit of tea drinki ng betwe en any
~d female citizen s ~f a town. _In su~h a situati on, Z-tes~ ~e
differe nce of propo rtions can be applie d. · r

One would have to obtain two indep enden t sampl es_from the
towri- one from males and the other from female s and determin
the propo rtion of tea drinke rs in each sampl e in order to perfor e
this test. · · .rn

3. z -test for single mean is used.t o test a hypot hesis on a specific

value of the popul ation mean. .

Statist ically speak ing, we test the null hypot hesis H 0 : µ == µ

agains t the altern ative hypot hesis H 1 -: µ >< µ 0 where µ is th~
popul ation mean and µ 0 is a specifi c valu~ of the popula tion
that we would like t<? test for accept ance. . ·
. .

Unlik e the_t-test for ·single _m ean, this test is used if n > 30 and
t popul ation standa rd deviat ion is know n.
.
.
4. z t~st for single varian ce is used_to test a hypoth esis on a specific
V value of the popul ation varia11._ce.
Statist ically speak ing, we test the null hypot hesis H 0 : cr == cr0
agains t H 1 : cr >< cr0 where cr i,s the popul ation mean and cr0 is a·
specifi c value of the popul ation varian ce that w~ would like to
test for accept ance.
ln other words , this te~t enable s us to test if the given sample
has been di-awn from a popul ation with specif ic varian ce O"o-
Unlike the chi square test for single varian ce, this test is used if
n>30.

5. z test for testin g equal ity of varian ce is used ·tq test the
hypo.t hesis of equali ty of two popul ation varian ces when the
sampl e size of each sampl e is 30 or large~ ;
 Data Analysis and Interpretation
163
Assumption:

Irre spe ctiv e of the typ e of Z-t est use d it is

ass um ed tha t the
populations fro m wh ich the sam ple s are dra wn
~re nor ma l.
1nus_tration :
In Mu mb ai Cit y, in 2014, 20% of the stu den ts stat
ed tha t the y rea d
college Times ma gaz ine . In 2018, the pub lish er of Bus.i
nes s Wo rld
found out tha t circ ula tion has inc rea sed . A sur
vey wa s con duc ted
to find out wh eth er or_not , the circulation inc rea
sed significantly. A
sample of 1000 stu den ts wa s con tact ed and it
was fou nd tha t 210
respondents con firm ed tha t the y sub scri be to Col
lege Times.
Solution:
Null ~yp oth esis is H
0 ·µ = 20%
Alternate Hy pot hes is H > 20%
1 ·µ
This is one -tai led test . .

210 - 0.20
P - µ _ .100
Z=
Jµ( ln- µ) _~ __~_;=0.=20=(1=-=o=.20=.)
1000
Z = 0.21 - 0.20 ·- 0.01 0.1 0.1
= = 8.33
0.2 X 0.8 (o.16 0.4 0.012
1000 vwoo 31.62
As the val ue of z at 0.05 = 1.64 (as per Z table),
and the calculated
value of z (8.33) falls in the rejectiqn region, the
nul l hyp ~th esis _is
· rejected, and the alte rna te hyp oth esis. is acc~pt
ed and therefore, it
can be con clu ded tha t the c~rculation of College
Times increased
significantly.

tTEST
- .

At-test is use d. fo~ test ing the me an of one pop

'

_ulati?n aga inst a

standard or ·c om par ing the me ans of two pop ula
tion s if you do not
 <
164 Business Research Methods (S. Y .B.M.S.: SEM-IV)
know ·the populations' stand~rd deviation and when you have
limited sample· (n < 30). If you know the populations' standar~
deviation, you may-use a z-test.
' .

· Example: Measuring the average diameter o~ shafts fr.om a certain

machine w~en you have a small sample.
An F~test is used to compare 2 populations' vartances. The samples
can be any size..It is the basis of ANOVA. . ·

Example: Comparing the .v~iability _o f bolt diameters from two

machines.
Matched pair test is used to compare the means before and after
something is·done to the samples. At-test is ofteh :used because the
· samples are often·s~all. However, a z-test is us_e.d when the samples
~e large. The variable is the difference between the before and after
measurements. ·
Example: The average w~ight of subjects before anq. after following
a diet for 2 mon_ths. ·

Uses oft-distributio n Test:

We have ~een that if the sample is large, the use is made of the
tables of the normalpro~abilit y integral in interpreting the results
of an experiment and on the basis of th_a t to reject or ac_c ept the null
hypothesis. · ·

If, however, the sample.size~' is smail, the non:p.al probability tables

will n<;> lo~ger .b e use~l. Following~are the uses oft-distribution :
(a) To test the significance. of the mean of a small random sample
from a p.ormal popul.a tion. ·
(b) To test .the significance of the ·difference between the means of
two .samples .t aken fr~m a normal population. ·
' ~

(c) To test the significance of an ~bserved coefficient of correlation

including partial and rank correlations.
(d) To test the significance of an o b~erved regression coefficient.
 oata Analysis and Interpretation 165
f TEST,
An F-test is any statistical test in whi ch the test statistic h~s an
·F-
distribution und er the null hypothesis. It is mos t ofte n used whe
n
cornparing statistical mod els that hav~ been fitted to a data set,
in
order to identify the mod el that best fits the pop ulat ion from whi
ch
the data_were sam pled . Exact "F-tests"_mainly arise w);len the mod
els
have been fitted to the ~ata usin g least squares. The nam e was coin
ed .
by George W. Snedecor, in hon our _of Sir Ron ald A. Fisher. Fish
er
initially deve lope ~ the statistic as the variance ratio in the 1920
s.
common Examples of F-Tests
Cornmon exam ples of the use of P.-tests· are :·
, The hyp othe sis that the mea ns of a gjve n set of norp.,.ally
distributed populations, all having the same stan dard deviation
,
, are equal. This is perh aps the best -kno wn F-test, and play s
an
imp orta nt role in the analysi~ of variance (ANOVA).
, The hyp othe sis that a prop osed regression mod el fits the
data
well. See J;.,ack-of-fit sum of squares.
. . .
, The hyp othe sis that a data set in a regr essi on analysis follows
the simp ler of two prop osed line ar mod els that are nest ed
within each othe r. . - . .
• In addi tion , som e stati stica l proc edur es, such as Scheffe' s
method for mul tiple comparisons adju stme nt in linear mod els,
also use F-tests.

Formula and Calculation

~ost F-tests arise by ~onsid~ring a decomposition of the variabili
ty
ma collection of data in term s of sum s of squares. The test stati
stic
in an F-test is the ratio ·o f two scaled sum s of squa res reflecting
different sou rces of vari abil ity. The se sum s of ·squ ares
are
constructed so that the statisti~ tend s to be greater whe n the
null
hypothesis is not true . In orde r for the statistic to follow the
F-
distribution und er the null hypo thes is, the sum s of squa res shou
ld
be statistically inde pend ent, and each shou ld follow a scaled
chi-

l t
 ◄
Y.B .M.S.: ~EM.
. Busine~s Research Methods (S.
16 6
io n. Th e la tte r co nd iti on is .g ua ra n_tee d if th ~fvJ
sq ua re d di st rib ut ly di st rib ut ed w ith a co e datan
de pe nd en t" an d no rm al lluno
va lu es ar e in · · ~ ,
va ria nc e. • ,

C h i- S q u ar e T es t

In tr od uc tio n
re w hi ch ev .a lu at es ex te .n t :o whic~ ~ set of
C hi sq ua re is th e measu a sa m _P le de vi at es fr om the
.
fr eq ue nc ie s of
th e ob se rv ed . It is the
eq ue nc ie s of th e sa m pl es
nd in g se t of ex pe ct ed fr
co rr es po be tw ee n ac tu al an d expected
at e di sc re pa nc ie s
m ea su re of ag gr eg
·
fr eq ue nc ie s.
lle d x 2 i_
-d s ~i bu t~ o~ , ~a s .fir st dis_covered by ·
th is di ~t rib ut io n, ca de riv ed 1t m de pe nd en tly _m · 1900 and
Helmet in 1875. Karl Pear so n
\
of H go od ne ss of fit ". It is us ed as a test statjc in
ap pl ie d it as a 't es t eo retical frequencies w ith whi~h
es is th at pr ov id es th e th
testing a hypoth
ed.
obse_rved frequencies ·are compar
II _D ef in iti on of C hi Square
·(x 2) ·

rv ed fr eq ue nc y an d fe th e co rr es po nd in g
· · If fo de no te s th e ob se
s in te rv al or ce ll, th en w e de fin e x2 by
ex pe ct ed fr eq ue nc y of a cl as
th e re la tio n
. 2
2 (f0 _ - fe). . ... (1)
X =L--- ·
fe
ho le se t of cl as s intervals or
I ex te nd s to th
_ e w
w he re th e su m m at io n
is gi ve n by
cells. A no th er us ef ul fo rm _
I . £2 2)
2 ... (
x = 1:_Q_ - N
fe
nc y ..
w he re ~ is th e to ta l fr eq ue
th at u~ e~ th e ch i sq ua re di st ri bu tio n ca n be called
~ y sta tis tic al te st th fo r la rg e an d sm al l· samples-
pl ic ab le bo
ch i sq ua re test. It is ap ·
de pe nd in g on th e co nt ex t.
oata Ana_lysis and Interpretation
_ 167
r,xatnple: .Su pp os e a pe rs? n wa nts to tes t
the hy po the sis·tha t success
rate in a pa rti cu lar En gli sh tes t is sim
ila r for stu de nts stu die d in
private schools an d Go ve rnm en t sch
ools. ·
. ,
If we take ran do m s~ pl e of say siz e 80 stu
de nts an d me asu re bo th
types of schoo~s as we ll as su cc ess
/fa ilu re sta tus of e_a ch of the
student, the ch i sq ua re tes t ca n be ap
pli ed to tes t the hy po the sis .
Tests for Di ffe ren t Purposes
' '
There are dif fer en t typ es _of ch i sq
ua re tes t ea ch for dif fer en t
purpos~s. So me of the po pu lar typ es
are sta ted as follows: _
Chi square test for testing goodness offit: It
is us ed to de cid e wh eth er _
there is an y difference be tw ee n the ob
ser ve d (ex pe rii ne nta l) va lue
and the ex pe cte d (theoretical) va lue .
Example: Gi ve n a sam ple , we ma y like to
tes t if it ha s be en dr aw n
from a no rm al po pu lat ion . Th is ca
n be tes ted us ing ch i sq ua re
goodness of fit pro ced ur~ . ,
Chi square test for independence of -tw
o attribt!tes: ·su pp os e N
ob ser va tio ns are co ns ide red an d cla ssi fie d ac co
characteristics say A an d B. We ma y rd ing tw o
be int ere ste d t~ tes t wh eth er ·
the two characteristics are ind ep en de
nt. In su ch a case,·we ca n us e
Chi square tes t for ind ep en de nc e of
tw o att rib ute s.
The example co nsi de red ab ov e tes tin
g for ind ep en de nc e of suc ces s
in the English tes t - Stu de nts Pri va
te schools ve rsu s Go ve rnm en t
schools - is a cas e fit for an aly sis us ing
.this test. ·
Chi square tes t f~r sin gle va ria nc e is us e~
to tes t a hy po the sis on a
specific va lue '9f the population_variari
ce. Statistically spe ak ing , we
test the nu ll hy po the sis Ho : cr = cr ag
1-Ii: cr # cr wh ere cr 0 a~ st the res ear ch hy po the sis
0 is the po pu lat ion me an an d cr is a specific
of the po pu lat ion va ria nc e tha t we wo 0 va lue
uld ~ike to test"for acceptance.
This test enables us to tes t if the giv en
samp~e ha s be en dra wn_fro m
a Population wi th specific va ria nc e cr
• Thi~ is a ~mall sam ple tes
be used only if sam ple size is le~s tha0n 30 m t to
generq.1. -

\.
168 Bus ine ss Research Me tho ds (S. Y .B. M.S.: SE1\1,JV)
.
As su mp tio ns
ha s an_· ass um pti on th
Th e ch i ~q uar e tes t for sin gle va ria nc e
ha s be en _is no rm al at t~e
po pu lat ion fro m wh ich the sam ple
ch i squ are goo dne s~ Th~s
no rm ali ty a$ sum pti_o n ne ed no t ho ld for of fit
d tes t for ind ep en de nc e of att rib ute s.
tes t an
o tes ts; on e ha s toe
Ho we ve r wh ile im ple me nti ng the se tw
t les s tha n 5. If it is so~ i:e
!ha t ex pe cte d fre qu en~ y·in an y ~el l is_ no
or suc cee din g cell 80 th n
1t ha s to be po ole d wi th the pre ced mg · at
is at lea st 5.
_e xp ect ed fre qu e~ cy of the po ole d cel l

No·n Pa ram etr ic an d Di str ibu tio n Fr ee

s to be no ted tha t the chi squ ?tT e go od ne ss off it ~est and test for
It ha
on_the set of obs erv ed and
ind ep en de nc e of att rib ute s de pe nd on ly
edo m. Th ese tw o tests do
ex pe cte d fre qu enc ies an d de gre es of fre
ed an y ass um pti on i-~ gar din g dis tri bu tio n of the p,arent
no t ne
ta~ en.
po pu lat ion fro. in wh ich _the sam ple s ar~ . .
.
.
po pl; ila tio n par am ete rs or
Sin ce the se tes ts do no t inv olv e an y
rac ter isti cs,·the y are als o ter me d as no n par am etr ic or distribution
cha
t on the se-tw o tes ts is the y are
. fre e tes ts.· An add itio nal im po rta nt fac
d for ·an y sam ple size as
sam ple siz e in~ ep en de nt an d can be use
ect ed cel l fre qu enc y is met.
lon g as the ass um p.t ion on mi nim um exp

Ste ps in Co mp uta tio n of Ch i Sq ua re

(1) Ind ica te Nu ll hy po the sis•an d .Al
ter nat e hy po the sis .
the inv est iga tio n.
(2) Lis t the ob ser ved fre qu enc ies fro m
bas.ed up on the application
(3) Ca lcu lat e the exp ect ed fre qu enc ies
of the ore tic al dis trib uti on .
ob ser ved an d exp~cted
(~) Ca lcu lat e the dif fer enc e bet we en the
fre qu enc ies cor res po nd ing ly.
r abo ve ste p.
(5) . Ad d tog eth er all the ·fra cti ons as pe
As cer tai n the app rox im ate va lue fro m the tab ie at a .particular
(6)
lev el of sig nif ica nce .

\
 . taAnalysis and Inte-rpretation
va . . . .
(7) Finally t~ke the decisio n of accepti ng or reiec_tmg of the
hypothesis. ,
2
conditions for the Applica ti?n _of X ~est .
Before ~sing x2 distribution as a test of significance, the followin g
precautions are necessa ry to observe : _
(a)' The total frequen cy, N, should. be fairly large. It. s~ould b.e at
· least 50, howeve r small the number of cells may b_e:
(b) No theoretical cell frequen cy should be small. It should not be
less than five. It is better if the smalles t cell frequen cy is 10 or
greater. When the cell frequel)cies are small in two or more
cells, they should be combin ed _into a single cell.
(c) All the individu als in a sample must be indepen dent.
(d) The number n, of classes or cells, should be neither too small
nor too large. It is better if 5 < n < 20. Sometim es the numbe r of
cells less than 5 can also be used provide d the cell frequen cies
are not small. · ·· · ·

(e) The constraints impose d on the ceil frequen cies must be linear.

Examples: .

1. In 120 throws of a single dice, the followin g distribu tions•of

faces was obtained . ·

Faces: 1 2' 3 4 5 6 - Total

fo: 30 25 18 10 22 15 120 -
Do these results constitu te a refutatio n of the "equal .p robabili ty"
(null) hypo~hesis ? . . ·

On the basis of equal probabi lity i.e. 1/ 6, the theoret ical

. · 120 .
frequencies for each face will be 6 = 20.
 Y.B.M 's...SE-,..,f
170
Business Rese. arch Methods (S. i.v1~JvJ .
2
(30 - 20)2 + •(25 ._ 20
) + (18 - 20)2
H en ce : x2 =
20 20 20 -

2 2
+ (10 - 20) + (22 - 20) . + (15 :- 20)2
20 ~
20 ·

_!__ [100 + · 25 ·+· ·4 + 100 + 4 + 25]

=
20

129
= = 12.90
10

N ow v = 6 - 1 = 5 ·
, w e ge t fr om th e tables
Fo r 5 de gr ee s of fr ee d_o m

x2 = 11.070 at 5 pe rc en t level.
x 2 is gr ea te r ·th an this value, the
Since th e ca lc ul at ed -y alue · of
hy po th es is of eq ua l pr ob ab ili ty is rejected.
.
n 96 tim es an d th e nu m be r of times 4, 5 or
w
2. Five dice w et e th ro
, 6 w as th ro w n w er e :
6: 5 4 :3 2 1 0
ce sh ow in g 4, 5 or
.No. of di
y . 8 ·18 35 24 10 1
Fr eq ue nc
ob ab ili ty of ge tti ng th is re·s ul t by chance.
· Fi nd th e pr
ge tti ng a 4, ~ or 6 in a si ng le th ro w of 9 single
Th e pr ob ab ili ty of
ce th e th eo re tic al fr eq ue nc ie s ar e th e successive
dice is ½. H en
pa ns io n.
te rm s of th e binomial ex

(1 1J
5
96 - ·+ -
· 2 2
►
Analysis and Interpretation I 171
oata
or 3, 15, 30, 30, 15, 3
Since the borde r frequ encie s are le~s than 5, we comb ine them
with the adjac ent frequencies ·as follows : . .
fo: 26 35 24 11
fo : 18 30 30 18
2
Hence X2 = (26 - 18) + (35 - 30)2 + (24 - 30)2 + (11-18) 2
18 30 30 18
64 . 25 36
= - + - + - +49
-
18 · . 30 30 18
=8.31
v==4 -l=3
For three degre es of freedom, the value of P correspondµ1g to
x2 == 8.31 is found from the tables to be 0.041. ·
Hence the probability of getting as bad or worse a fit in rando m ··
41
sam~ling is ie. ~bout 1 in 25._
100
Significance of Chi Square (X2) Test

(1) It is based on frequ encie s: The most impo rtant characteristic

of Chi Square Test is that it is based upon ·frequencies of an
item and not on the param etreal equations such as mean and
standard deviations. Thus , the values generated by Chi Squa re
Test are comp arativ ely more reliab le and signi fican t as
compared to other mean s of testing hypothesis .
(2) It possesses addit ive properties: The Chi Square Test possesses
additive properties such as-testing of goodness of fit or judgi ng
the significance of association between its attributes. Thus, if
the independent ·variables x and y confirms to the Chi Squar e
.. distribution and y and y degrees of freedom then they are
1 2
distributed like the requi red Chi Square with y1 and y .
2
3
( ) Application in comp lex conti ngenc y table : The ·Chi Squar e
Test is applicable to various complex contingency table with

l
 172 A
. Bus ines s Research Met hod s (S. Y .B .M.S .: S£1.~~~ ~
.
sev era l cla sse s and hen ce, it is a ver y usef1:11 test in rese
~h
wm k. .
-

is
(4) Fle xib le in nat ure : Thi s test an imp ort ant non par arn etr·
ary and le ic
tes t and no rig id a~s um pti ons are nec ess 88
ma the ma tica l calc ula tion s are jnv olv ed. ·
Chi Squ are Test .
(5) Mo st sui tab le me tho d for hyp oth esis : The
s at test ing ~;
mo st sign ific ant ly use d bec aus ~ jt pri ~ar ily ·aim
To1:1s, the test
hyp oth esi s rath er tha n gen era t~g est i~a tlon s.
d 1n a more
of pop ula ti~ n can be don e usi ng this me tho
sim plif ied ma nne r. ·

AN OVA-( AN AL YSI S OF VA RIA NC E)

two pop ula tion s
In stat isti cs, we ofte n wa nt to knp w if the me ans of.
n equ al wag es on
are equ al. For exa mp le, ·do me n and wo me n ear
-sa mp le t-te st for
ave rag e? This is an eas y thin g to test usi ng a two
t is we can not deal
the equ alit y of me ans . The pro ble m wit h tha~ tes
to kno w,p erso ns
wit h mo re tha n two pop ula tion s. Wh at if we wa nt
·sam e wag es on
, bel ong ing to thre ~ or mo re eth nic gro ups ear n the
d to use AN OVA.
ave rag e? To ans wer a que stio n ~e this , we nee
AN OVA me ans ana lysi s of var ian ce. .
stat isti cal rrio_dels
Ana lysi s of var ian ce (AN OY A) is. a col lect ion _of
me ans and their
use d to ana lyz e the diffete11cJle s b.e twe en gro up
ong and bet wee n
ass oci ated pro ced ure s (su ch as "'va riat ion " am
OV A set ting , the
gro ups ), dev elo ped by R.A : Fis her . In the AN
par titi one d into
obs erv ed var ian ce in .a pq1i;_~ici.tlar var iab le is
varia_tion. In its ·
com pon ent s attr ibu tab le to diff~re nt .sou rce s of
t of wh eth er or not
sim ple st form , AN OVA pro vid es a stat isti cal tes
refo re gen era lize s
the me ans of sev era l gro ups are equ al, and -the
le two -sa mp le t-·
the t-te st to mo re tha n two gro ups . ·A s doi ng mu ltip
mit ting a stat isti cal
test s wo uld res ult in an inc rea sed cha nce of com
(tes ting ) thr ee or
typ e I erro r, AN OV As are use ful in com par ing
sig nifi can ce.
mo re me ans (gr oup s or var iab les) for stat isti cal
esi s tes ting hea vily
AN OVA is a par ticu lar for m of stat isti cal hyp oth
isti cal hyp oth esis
used_ in the ana lys is of exp erim ent al dat a. A stat
a. A tes t res ult
test is a me tho d of ma kin g dec isio ns usi ng dat
sam ple ) is called
(calculated fro m the nul l hyp oth esi s and the
Data Analys is and'Interpretation . 173
statistically signif icant if it is dee~e d unlike ly to have occurr ed by
c~an~~, assum ing the truth of the null hypot hesis. ·A statist ically·
signif icant r_es~lt, wh~n a proba bility (p-val ue) is less than a
thresh ol~ (signi ficanc e level) , justifi es the rejecti on of the null
~ypo th~sis, ~ut only if the a priori probab ility of the null hypot hesis
1s not high. :
. '

In the typica l_ applic ation of ANOVA, the null hypot hesis is that.al l
group s are ~rmpl y r~do m ·samp les of the same popula tion. For
examp le, when study ing the effect of differe nt treatm ents on simila r
sampl es of patien ts,.the null hypot hesis would betha t all treatm ents
have th_e same effect (perha ps none). Reject ing the null hypot hesis
would rmply that differ ent treatm ents result in altere d effects.
By constr uction , hypot hesis testing limits the rate of Type I errors
(false positi ves leadin g to false scienti fic claims ) to a _signifi cance
level. Exper iment ers also wish to limit Type II errors (fals~ negati ves
resulti ng in misse d scient ific discov e~ies). The Type II error rate is a
functi on of sever al thing s includ ing· sampl e size (posit ively
correl ated with exper iment cost)~ signif icance level (when the
standa rd· of proof is high, th~ chanc es 0£ overlo oking a discov ery
are also high) and effect size (when the effect is obviou s to the casual
observ er, Type II error rates are low).
The termin ology of ANOVA is l~rgel y from the statist ical design of
exper iment s. The e~per iment er ,.a dj?sts factor s and mea~u res
respon ses in an atteII1:pt ~o determ i~e ~ effect: Factor s ~~e ~ss1gn ed
to exper iment al units by a combi nation of r~do1 :ruzah on and .
blocki ng to ensur e the validi ty ~f. the results . Bhndm g keeps the
weigh ing impar tial. Respo ns~s sll~w a variab ility that is partia lly
the result of the effect and is partia lly rando m error.
'. . .
_

ANOV A is the synthe sis of severa l ideas and it is u~ed for ~ultip le
purpo ses. A s a conse quenc e, it .is difficu lt to define
. concis ely or
precise ly.
Classi cal ANO VA for balan ced data_does three things at once:
t ry data analys is, an ANOVA is an organi zation of
1. As exp1ora O · . f · d'
. • d ta decom ppsiti on, and its sums o square s m icate
dd
a 1t1ve• a e of f h d · · ·
the varian c each comp onent o t e ecomp osihon (or,
.
. 1 tly each set of terms of a lmear
equ1v a en , · model ).
 174 Busin ess Research Meth ods (5. Y .B.M. S.: SEM-IV)
2. Com paris ons of mea n squa res, alon g with F-tes ts allow testing
of a neste d sequen-ce of mod els. · ·

3. Clos ely relat ed to the ANO VA is a linea r mo.d el .fit With

coeff icien t estim ates and stand ard error s. In shor t, ANO VA is
a statis tical tool ~used ~ seve ral ways to deve lop and confinn
an expl,a natio n for the obse rved data,.
5. ·. It is com puta tiona lly elega nt and relat ively robu st aga·inst
viola tions of its assu ~ptio ns.
6. ANO VA prov ides indu stria l str~n·g th (mul tiple samp le
· . comp ariso n) statis tical analy sis ..
7,. · It has been.adap ted to the ~aly sis of a varie ty 'o f expe rime ntal
· desi ~s.
As a resul t ANO VA "has· long enjoy ed the _status of bein g the most
used statis tical techn ique in psychologi<?al resea rch." ANO VA "is
pro_b ably the mos t usef ul tech niqu e in th_e field of statis tical ·
infer ence ." ·

ANO VA is difficult to teach , parti cular ly for comp lex expe rime nts,
with split -plot desig ns. In some cases the prop er appl icati on of the
meth od is best deter mine d-by probler,n patte rn recog rutio n followed
by the cons ultat ion _o f a cla.ssi~ authdrita tive test. .. . · , .

1. State the mean ing qf data p i6tes s~g. Expl ain iri brief the stage s of.
data proce ssing. . _ . . .
2. What do you mean by data analy sis? Expl ain its chara cteris tics.
3. Discu ss the purpo ses of .data analy sis .
. 4. State and expla in the differ ent types of data analy sis. ·
5. What is ~ata_interp retati 9n? Wha t are its essen tials?
6. WJ:tat are the meth ods of data interp retati on? .
7. . Expla in the signif icanc e of data -inter preta ti~n.
8.. · Discu ss the imp~ rtaric e and significanc;e of data proce ssing .
9. · Expla in the conce pt of hypo thesi s testin g. . . .
10. What are_the steps invo lved ~ hypo thesi s testin g?

I:.,_
' oataAnalysis ~nd Interpre~~tion 175
11. What are the types of hypothesis testing?
1z. Write short notes on:
· (a) Data editing
(b) Coding of data .
.(c) · Data tabulation
(d) ANOVA · ·
(e). Multi-variate analysis
13. Solve the following problems:
. (a) In experiments on pea breeding, Mendel obtained the following
frequencies of seeds: · ·
Round and Yellow 315
Wrinkled and Yellow 101
Round and Green 108
Wrinkled and ·Green 32
Total 556
Theory predicts that the frequencies should be in proportion of 9:3:3:1.
Examine the correspondence between theory and experiment.
(b) Find the value of chi-square test for the following table:
Class .A B' C D E
Observed frequency · 8 · 29 47 15 4
Theoretical frequency 7 24 38 24 7

I. Fill in the blanks with the :correct option :

1~ ___ is a proces~ of skillfylly O !,~~sing of data for the purpose of
data analysis and interpretation.
(a) Scaling ~..1r(b) Sampling
(!2) Data processing _ (d) Data collection .
2. ___ is . the process of thecking errors and omissions ih data
collection, and making corrections, if required. ··
(a) E_d iting . (b) Coding
(c) · Classification . (d) Tabulation
3. · ___ is· the· process of grouping of collected ?-ata into different
· categories. .
(a) . Editing (b) Coding
. (c) Classifkation (d) · Tabulation
 ◄

176 Business Research Methods (S. Y.B.M·s...·SEA~

1v1-IV)
.
4. _.,. __ char t shows how a particular data changes at equal inte
in~ . . ~
(a) Pie . (b) Bar
(c) Line (d) Histogram
--- ·
5. The probability of committing a Type I error is called the
(a) significance level ·(b) . beta . .
(c) sample .. (d) alph a · ·
6. . erro r occu rs whe n the rese arch er fails _to reject a nun
hypothesis, which is false .
(a) Type I (b) Type II
(d) Type IV · .
(c) Type III _
l and
7. _ _ _ is the mea.sure of aggregate discrepancies between actua
expected frequencies.
(a) Chi-square (b) . ANOVA
(c) T-distrbution (d) F-test
8. A 11 A2, A3, etc. is called as __ _·coding.
(a) alpha-numerical (b) alphabetical
(c) numerical . (d) scientific
er-wise,
9. When different categories are. acco rded age-wise, gend
occupation-wise, etc. it is called as - - - .
(a) Editing · ,,'{b) Classification
. (c) Cq_ding (d) Tabulation
10. . The categories must be _ __
(a) excl,usive (b) mutu ally exclusive ·
(c) inclusive· /) (d) mutu ally inclusive ·.
counting
11. ·· __ _ involves sorting of data int.a different categories and
the number of responses ~hat belong to each category.
(a) Editing (b) _Classifi~atio,n
Ic.1 J
(c) · Tabulation (d) Coding .
e.
12. A __ _ is a circular chart usect_to ,com pare part s of the whol
(a) Pie chart . ~(b) ' Bar grap hs
(c) Line graphs . (d) Gnat t charts
called
13. Bar charts arranged from highest to lowe st incidence are

(a). Pie charts (b) Pareto char ts

(c) Line graphs (d) Histogram
nuous
14. - - - is an e~timate of the probability distr ibuti on of a conti
varia b~e and'was first introduced by KarLPearson.
(a) Pie chart . . (b) Gnat t char t
(c) Line graph ,(d) Histo gram
 l
1
'· ~ oata Analysis and Interpretation l77
\

~ 15. The concepts of data analysis and _ _ _ are interwoven.

, (a) data collection (b) data·organisation
(c) data pr~cessing (d) data interpretation
16. ----- aids the drawing of reliable inferences from observations.
(~) Data collecti'on (b) Data an~lysis
(c) Data pr9cessing . (d) Data interpretation
17. - - . - analysis is concerned with the study _of how one or more·
variables_affect changes in another variable. ·
. (a) . Descriptive (b) Inferential .
1
(c) _Correlation (d) Causal ·
18. - - - analysis uses statistical methods which analyse more than
ltt two vari~bles on a sample of observations. .
(a) Casual . (b) Multivariate
(c) Correlation (d) . Inferentia:~
19. ______ helps to convert the statistical data into information.
(a) Data collection (b) Data organisation
,(c)' Data processing · . (d) Data interpretation
20. _ _ _ ensures completeness of data.
(a) Editing . (b) Coding_ ,
(c) Classification (d) Tabulation
21. _____ facilitates classification of data.
(a) Editing . (b) Coding
(c) Gr~phic presentation (d) Tabulation
22. _ _ _ facilitates grouping of data into 4ifferent categor~es.
(a) Editing '(b) . Coding ..
(c) Classification . (d) Tabulation
23. _ _ _ provides the basis for writing rese·a rch report.
(a) Editing (l?) Coding
(c) Clas~ification . . .(d) Tabulation
· 24. _ _ _ can be illustrated to 1create a good impact on the mind of the
readers with special referertck to understanding the data. . .
(a) Graphic presentation , (b) . Editing
(c) Tabulation . (d) Coding
25. _ _ _ aims•- to prove or disprove the research question.
(a) Literature survey (b) ·Hypothesis testing
(c) · Data analysis · . · (d) Sampling ·_
i 26. · depends on an assumption abou~ the distributio11: of the data,
. that the data are normally distributed. ·
(a) Data collection (b) Parametric tests
(c) Hypothesis (d) Non-parametric tests
 ◄
178 Business Research Met hod_s (S. Y.B.M.s ••. S£ll.. r
LVl-J\7
27. _ _ _ is wh en the sam ple is mor edthad- n 30 an_d whe n th~ pro p . )
- port·ion or com pari ortio
n ll
of som e cha ract eris tic ver sus a stan ar pro
· ns.- gthe
.
tion s o f two pop u1atto
pro por
(a) z-te st · (b) · t-te st
(c) £-test (d) chi- squ are test
anc e of the mea n of a sma ll rand orn·
28. _ _ _ is use d to test the sign ific .
~ 30) from a nor mal pop u1atio n.
sam ple (n
(a) z-te st (b) t-te st
(c) AN OV A _ (d) £-test
29. _ _ _ is app lied as "good
nes s o ff1.t" . . -
(a) t-te st , . (b) Chi -squ are test
(c) · AN OV A (d) £-test
of stat isti c_a l mo del s· use d to ana lyse the
30. _ _ _ is a coll ecti on ciat ed_proc edu res.
diff eren ces betw een gro up mea ns and thei r asso
(a) z-te st (b) £-test
(c) AN OV A (d) t-te st

II. . State whe the r the foll owi ng stat eme nts are Tru e or False:
dat a collection and
1. Dat a pro cess ing is the inte rme diar y stag e betw een
data ana lysi s. . :
-is call ed as central ·
2. Edi ting und erta ken at the tim e of fiel d sur vey
edit ing. -
re.
3. The clas sific atio n of data mu st be exh aus tive in natu
4. · M_a nua l tabu lati on is pos sibl e whe n the sam ple size is larg e.
data . - · ·
, 5. A bar cha rt is ver y use ful f~r reco rdin g disc rete
rov e the_research
6. The basic logic of hyp othe sis test ing is to pro ve or disp . ·
que stio n.
7. Non -par ame tric ana lysi s can be use d onl y whe n the data are:normally ·
·
dist ribu ted.
8. . T-tes~ is us_e d whe n the sam ple size is abo ve 30.

.\ A
►
179
Oflta Analysis and-Interpretation
Ill· Match the following:
Gr<;>upA Group B· ·
,

1. Coding (a) Ho
2. Histogram (b) Large sample size
3. Graphic presentation (c) H1 -

4. Null hypothesis (d) Kar1Pearson

5. Alt'ernate hypothesis (e) Effect~v~ appeal

B. . Group A Group B
1. z-test (a) Fisher
2. t-test (b) Based ·o n frequencies
3. £-test (c). Sample size above 30
4. Chi-square (d) . Means of two populations
5. ANOVA which ar~ equal
(e) Sample size is less than 30

Atl$:

IV. Explain the concepts:

1. Editing
2. Coding
3. Classification
4. Tabulation
5. Data analysis ·
6. Data processing
?. Data interpretation-
s. T-test
9, F-test
lO. Z-test
ll. Chi-square test
12• ANOVA
 180

' .
Meani ng of Repo_r t Writin g
•
Impor tance/ Functi ons ofRepor t/Rep ort Writ~ng
•
Essentials ofA Good Repor t
•
Conten ts.ofA Repor t__
•
Steps in•W riting A.Repor t
•
• Types of Repor ts .
Foofn:otes and Bibilio graph y
•
Ethics anil Research
• • I

• . Object ivity, Confid enliali ty and Anon ymity in Research

• Plagiarism
(

. 1· •

. .
A report is a statement. of facts and figures , prepa red for the purpo se
of inform ation and action. The Oxfor d Dictio nary define s a report
as ✓1a· record of ascert ained fac~s".
A report is a forma l statem ent of the result of an invest igatio n of
any matter on which definit e inform ation is requir ed, made by some
pers,on or body, instru cted to do so. ·
Advanced Techniques in Report Writing 181
;\lllerican Marketing Society _states that the purpos e of report is
''to convey to the interested persons the whole result of the study
· itt sufficient detail and so ·arranged as to enable each reader to
comprehend the data and determine for himsel f the validit y of
conclusions." .

Reports· can be in writin_g or oral. Rep9rts are usually in writing .

This is becaµse, they act as a reference for future. _O ral reports are
justified, especially when the matter is of urgent in nature or where
secrecy needs to be mainta ined. . . . .
The last phase of the journe y of research.activity is writing of the -
report. After the collected data has been analys ed and interpr eted
and various generalizations have been drawn, the report has to be
prepare9.-. Report writing is a brief stateme:nt of most significant facts
that are necessary for unders tandin g the generalizations drawn by
the researcher. ·
Report writin g is a means for comm un.i cating the resear ch
experiences to others and adding· to the pool of knowl edge.· It
narrates the proble m studied , method s used for studyin g it and the
findings and conclusions of the research study.

. .
The importance and functions .o f report / report writing can be ·
explained as follows:
1. Up-to-date Information: The research report provide$ up-to-
date information on import ant matters. For instance, the Census
Survey -Repor ts provid e .information on-age compo sition,
gender ratio, educat ion level, incom e level, densit y of
population, rural-u rban popula tion, etc. Such information is
vital for business firms as -w:ell as for non-business organisations ·
to take vifal decisions.
Also, in, the case of commercial research, reports may provid e
information on market share, competitors' strategies, custom er
 .,.
182 Business Research Me tho ds (S. Y.B·.M. S.: S£1-LVl~J\7
. orm ati•on c·.an be. use d by rn arketir\ )
· fer enc es, etc: Su.ch inf .
pre
ma nag ers to tak e vital dec·isionsJrela ting_to 1:1ar~etmg mix 84chg
• e, pro mo tio n, p ace of d1s _ tnb ubo n, etc· ·
as pro du ct, pnc
ed· on int ern al and externai
pre par . . .
2. Pla nn ing : Re por ts can be ~r~anisa~i~ns. The
env iro nm ent of bus ine ss and no n bus11:~ss
in orm ati on o:n_
int ern al en vir on me_n t ma y pro vt e es
ma nag em ent -la 'bo ur rel ati onwo s, ma np ow er. c_ompetenci
· ine s ·an d tec hn o1og y,· rki ng co nd itio ns,. Welfare'
·
ma ch ex te~ a. 1 environment
facilities, wa ges and sal arie s, etc . Th e
•d · for ma tio n on .Go vt po~ hc1 e_ s , cus
. tom er
·
rep ort s prov1 e 1n em abo na 1environment,
. ti' tor s' stra teg ies, mt .
pre1erences, compe .
.t.

etc.
SW OT analysis.
Th e int ern al and ext ern al rep ort s fac ilit ate
me d to_ achieve
Ac cor din gly , alt ern ati ve pla ns can be fra
al! ~rn ahv e plans
objectives. The ma nag em ent the n eva lua tes
fac ilit ate planning
and selects th~ best pla n. Th us, ,re por ts can
in the organisation.
red on_ the cau ses of
3. Sol ves Pro ble ms : Re por ts can be pre pa
b~ pre pa red on
v~ iou s pro ble ms . Fo r instance,. rep ort s can
i~a te the cau ses of
declining sales tre nd. The rep ort ma y ind
pri ce, pro mo tio n,
ded inµ lg sal~s suc h as defects in pro du ct,
e, the ma nag em ent
distribution, after-sale-sErrvi~e, etc. Th ere for
act ion .
can solve the_pro ble m by 'faJ' <ing. cor rec tiv e
es is du e to poor .,
Fo r· instance, if the pro ble m of dec lin ing sal
eff ort s to improve
after-sale-service, the com pan y ma y, ma ke
ion of ser vic e staff,
th~ after-sale-service throtig.Wsdentific·sel ect
ng the em plo yee s ·
eftective tra ini ng pro gra mm es, an d mo tiv ati
s.· .
w1th mo net ary and non-moni3ta:ry inc ent ive
4. Feedback to En,tployees ·: Re po rts can pro vid e fee db ack to the
'·
· t a~~ •h · sal rep· ort s
em plo yee s · Por ms e, _t e pe rfo rm an ce app ra~
indi
plo ye es Therefore,
h cate strengths ·ai:id ~e akn ess es of the em
s an d eh ha nc e. the
t e rep ort s ena ble to cor rec t we ak ne sse
stre ngt hs.
· d ica
rep o t8 m ·- · · h
For ins tan ce, if the
.m r te lac k of co mp ete
·
nc ies of t · e
· .
em p1oy ees the .-
ect ive train111g ·
programm~s t . an ag em en t can de sig n eff plo ye es. If the
-o rm pro ve com ~e ten cie s of th~ em
Advanced Techniques in Report Writing 183
workers are demotivated due to poor personnel policies, the
management may make changes in the personnel polices so
that the emplpyees are motivated and improve the performance.

5. Organising Resources : Reports facilitat_e s organising of

resources. For instance, a project report may µ1dicate the amount
of additional funds for the impl~mentation of project, This will
enable the company to obtain fti~ds from the right sources.

6. Reports on Job .Satisfaction : Resear~h. may be conducted to

study the job satisfaction l~vel of employees. Employee
per£ormance reports. can provide information on the level of
job satisfaction. If the reports indicate low level of job
satisfaction due to faulty placement, poor superior-subordinate
relat~onship, work overload, or some other reasons, the
manage:111ent_Cclll take suitable action to improve the level of
job satisfaction. ·

7. · Decisio.n-Making : Reports are useful for taking vital decisions.

For. exainple, market _research reports may indicate good
potential for a certain product due to the gap between customer
expectations and product._offering by the competitors. Such
reports can be utilised by the marketing firm to launch a new
product in the market.
8. ·coordination : Repqrts can fqcil~
6
tate co-ordination between the
various departments · of an ~rganisation. For instance, a
committee · consisting of representatives from various ·
dep~rtments can jointly ·deal:o/ith a problem affecting variou_s
departments.·The committ~et~~-ort can_b~ prepared by taking .
the view points of all repr~~entatives. T~s m turn leads to better
co-ordination arid co-op~.Jjltion. . . . ·
9. Approval o_f Resear~h Report: The ~e~earch report is required
for approval from concerned authorities:
In case of academic research, the research report is required
• to be·approved by the research guic;ie.
A report relating to commercial research needs to accepted
• ~d approved by the_top management of the concerned ·
organisation.
184 Business Research Metho ds (S. Y.B.M .S.: SE""
~V! ~IvJ

• · A repor t on socia l resea rch need s accep tance by th

conce rned autho rity that has spons ored the resear ch.· e

10. Ref~r ence: Repo rts can be prese rved for_ fu~r e refere n~e. If a
partic ular probl em re-occ urs, then the find~ gs stated in th
previ ous resea rch repor t can be referr ed for suitab le action. F/
instan ce, if there is a dispu te betwe en the mana geme nt and th r
work ers regar ding incre ase in wage s and salar ies, the
mana geme nt may refer the previ ous repor ts on agree ment8e
and accor dingl y try to resolv e the matte r. '

action , ·For
11. · Legal Actio n : Repo rts can be used to take legal
instan ce, a comm ittee may be set up to look into the malpractices
adop ted by a partic ular mana ger or depar tment ~ The report
prepa red by comm ittee can·be use~ to take legal action against
the guilty perso ns.
12. Tool of Comm unica tion : Repo rts act as an impo rtant tool of
intern al comm unica tion. The super iors can excha nge views,
ideas , and infor matio n with their subo rdina tes. The '
subor dinate s can also provi de sugge stions in respo nse to the
repor ts of their super iors.

The essen tial chara cteris tics of a good resea rch repor t are stated as
follow s: / ·

1. Infor mativ e : The resea rch repor ts must be inform ative. The
resea rch repor ts must provi de adequ ate infor matio n to the
conce rned autho rities to take appro priate decis ions. Inade quate
inform ation ~ay not facilitate. prope r decis ion-m aking on the
part of the mana geme nt.

2. Clari ty : The repor t must be writt en "in simp le and Iu~id

langu age. The reade r shoul d find no diffic ulty in under standU lg
the conte nts of the repor t. The reade r shou ld be able to
under stand the conte nts in the first readi ng itself. Technical
Iangu a?e may b~ us~d in excep tiona l cases . Ambi guous
(confus mg) word s and phras es shoul d be avoid ed in the reports.
Advanced Techniques in Report·Wr iting
185
3, ~on cis~ .: The repo rt mus t be writ ten brie fly. Max imu m
info r~a~ on mus t -be prov ided in min imu m wor ds. As far as
poss ible\ leng thy .repo rts mus t be avoi ded. This is beca use,
l~ng thy repo rts are ofte n conf usin g and they requ ire a lot of_
time for the read er,
to note the cont ents .
4. Accuracy : The repo rt shou ld cont ain accu rate facts and figu res .
This ~s beca use, the man ager s base their deci sion s on the facts
and figu res of the repo rts. If the repo rts are inac cura te or cont ain
wro ng facts and figu res, then it will Jead to·poo r deci sion s.
5. Reli abil ity: A goo d repo rt mus t be relia ble. The info rmat
ion
in the repo rt .mus t be colle cted from relia ble sour ces. Whi le
c?lle cting info rma tion for the purp ose of prep arin g repo rts, care
mus t be· take n . to chec k the valid ity and genu inen ess of the
sour ce. Rep orts mus t not cont ain outd ated data . If repo rts are
base d ·on seco ndar y sour ces, the rese arch er in.us t chec k the
· gen ~~e ness of the data .
6. Obj ectiv ity: The repo rt mus t be obje ctive . It mus t cont ain only
obje ctive facts .and figu res .. The repo rts mus t not be bias ed or
subj ectiv_e . The repo rts_mus t not be infl: uenc ed by pers onal bias
of the re·s earc h~r. Pers onal bias adve rsely affec ts the deci sion -
mak ing. The refo re,..rese arc;h · repo r~s mus t be .obje ctiv e,
espe ciall y in _the case of com merc ial rese arch . ·
. ·~ . .
7. Log ical Arr ange men t : The.rese arch repo rt _m ust be writ ten in ,
a .syst ema tic man ner. The diffe rent part s of the repo rt mus t be
arra nged in logi cal sequ ence , as follo ws:

• Title of the repo rt

. •J

• Tab le of cont ents or inde x ·

Intr odu ctio n:of rese arch prob lem
•
Met hod olog y used in c_o nduc ting rese arch
•
Find ings of the rese arc~
•
Lim itati ons of rese arch , if any
•
 ◄
186 Business Research Methods (S. Y.B.M:s.: SEM-IV)
• Con clus ions

• Rec omm end atio ns

• App end ix
,
• Bib liog raph y
·
• Sig natu re of the rep orte r/C hair man of the rese arch
com mitt ee.
For
8. Sec recy : Cer tain repo rt find ings mus t be kep t secr~t.
ces
inst ance , a com mitt ee may be set up to look afte r malpracti
leak
·of a cert ain man ager . The com mitt ee mem bers sho uld not
ided
out the find ings . The repo rt find ings mus t be stric tly prov
only to the top man agem ent.
t be
9. Tim ely Sub mis sion of Rep orts : Cer tain repo rts mus
ee
subm itted with in a part icul ar time limi t. For inst ance , committ
take
repo rt mus t be subm itted ·with in. the time limi t so as to
r suit able actio n. ·
ce
10. Ref eren ces: The repo rt sho uld quo te refe renc es of the sour
in,
' from whi ch the info tma tion is obta ined or coll ecte d. Aga
inte d·
refer~nce has.to be maq.e to the auth orit y whi ch has appo
ee
the com mitt ee to repo rt, proc edu re foll owe d by the com mitt
whi le coll ectin g i:}nd repo rtin g data , and so on.
nal
-11. I1:11personal Styl e: The repo rt mus t be draf ted in an imp erso
style. The repo rt sho uld be ".Vritten in thir d pers on such as
"The
be
com mitt ee recom1!'ends.. ~~ tfowe ver, ce.r tain repo rts can
1

writ ten in the first pers on, su_c h as· n I reco mm end... "
format.
12. Proper _Format: The repo rt mus t-be writ ten in a suit able
ly
The repo rt mus t be div ided into para grap h~, pref erab
h.
num bere d and be givr n a suit able hea ding Jor each para grap
The repo rt mus t also con tain a suit able title . ·
sign ed by
13. Signatu~e and Dat e: .A repo rt sho uld be date d and
rt.
th_e Cha rrm an of the Com mitt ~e or the pres ente r of the repo
 187
Advanced Techniques in Report Writing
• ':'!" ,. .,,,..,,,.,.. • -;~"' -~ >lo":;·~- .... • ..., •• ...,.,., Y\: - '(" ·+ .:' . i~\~-- --~--, , , ~ ;,
·, STEPS - IN WRITING, X REPOR~ ; ~ -... ' 'i ,{ • ... -~~ '

The report.must be effectively drafted. The researcher must _follow

the general·guidelines for drafting a good report. The following are
the steps in writing research report:
1. Definitio~ of Target Readers: The researcher must define the
target readers. Target readers can be:
• In the case of academic research - the guide and the re~eree,
and also .general readers.
• In commercial research - top management.
• In social research ·_ sponsoring authority and the general
public.
The researcher should understand the nature of target readers,
whether they understand the technical language, whether they
have enough time to go through a detailed report, etc.
2. Proper Format: The report·m ust be written in a suitable format.
The report must ·be divided into paragraphs, preferably
numbered and be given a suitable headipg for each paragraph.
The·report must also contain a suitable title.
3. Logical Arrangement: The report must be written in a ·
systematic manner. The ideas or views or findings must be
arranged in a logical sequenc~ to ensure cohe~ence. Related
paragraphs or ideas must be written in successive paragraphs.
The different parts of the report must be .arranged in the
following order. First the title of the report, followed by a brief
introduction, then the procedure adopted in collecting data,
then the findings and recommendations and finally the date
and signature of the reporter or Chairman of the reporting
Committee. ·
4. PropeJ"..Di:afting: The report must be drafted in an impersonal
style. The report should be written in third person such as "The
committee recommends .. However, certain reports can be fl

writtet1· in the first person, such as "I recommend... fl