You are on page 1of 317

MB-207

RESEARCH
METHODOLOGY

MBA OIL AND GAS MANAGEMENT

OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~; k~ e
k ; k ‘ k f D r o’

COLLEGE OF MANAGEMENT AND ECONOMICS STUDIES (CMES)


Course Code: MB-207

Course Name: Research Methodology

© UNIVERSITY OF PETROLEUM & ENERGY STUDIES (For this print run)


Contents

Unit 1 Introduction .......................................................................................................... 1

Unit 2 Defining Problem ............................................................................................... 31

Unit 3 Relationship between Management and Research Problem ........................ 69

Unit 4 Research Design ................................................................................................. 75

Unit 5 Hypothesis Design ........................................................................................... 109

Unit 6 Sampling Procedures ....................................................................................... 141

Unit 7 Data Collection Methods ................................................................................ 169

Unit 8 Data Analysis–Methods .................................................................................. 203

Unit 9 Report Writing ................................................................................................. 275


1

Unit 1 Notes
__________________
__________________
Introduction __________________
__________________
__________________
__________________
__________________
__________________
Objectives __________________

After reading this unit you will be able to: __________________

y Understand the meaning of research.

y Define the objectives of research.

y List the features of a good research study.

y List types of research studies.

y Explain the role of research in business.

Introduction
Research may be defined as a documented prose work.
Documented prose work means organized analysis of the
subject based on borrowed materials with suitable
acknowledgment and consultation in the main body of the
paper. Research in management is particularly important
to find out different phenomena. At the outset we should
distinguish between research in different areas. Management
research comes within the purview of social science research
and there are other different types of research which broadly
fall into the category of physical science research. Carrying
out research in social science subjects, i.e., commerce,
management, economics, sociology, etc., is basically different
from physical science because, here we need to study the
society based on certain trends and for this the laboratory is
the society.

Research in management is particularly difficult because of


its convergence with different disciplines. As we know,
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
2 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes management is not a particular discipline and in any study


__________________ on management we need to integrate the different approaches
__________________ borrowing suitably from different disciplines. Similarly,
__________________ before we understand the complexity of research in commerce
__________________ and management, we need to define certain important
concepts.
__________________
__________________ First is variable. What exactly is a variable? Variable is the
__________________ quantity, in which we are interested, that varies in the course
__________________ of the research or that has different variables for different
__________________
samples in our study. In one word, we can define variable as
a factor whose change or difference we study. Now, there
__________________
are two types of variables. The first one is dependent variable
and the second is independent variable. Dependent variable
is that quantity or aspect of nature whose change at different,
stages the researcher wants to understand or explain. In
cause and effect investigation, the effect variable is the
dependent variable.

Now, what exactly is an independent variable? Independent


variable is a variable, whose effect upon the dependent
variable we try to understand. There may be several
independent variables. For instance, we may simultaneously
investigate the effect of mother’s cigarette smoking, mother’s
exercise, parents’ weights and other variables upon the
weight of a baby. In this case, mother’s cigarette smoking,
mother’s exercise, parents, weights and other variables are
independent variables, which we want to study, upon the
weight of the baby which is, dependent variable.

Now, there are certain other areas. First is universe. We can


define it as the total population. It is the laboratory for the
research. In our research we may have or we may take the
entire population of India. In that case, as it is obvious, no
researcher can carry out research on the entire population
of India to find out the truth or to find out some areas of his
research interest. In some cases, universe or population may
be a particular group. To clarify this point further, let us
assume we want to study some effects on some particular
group of people, religions; Hindu, Muslim, Christian, Jain
etc., or certain particular age-group, the age of 25 to 35. In
that case our universe is getting limited to that particular
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 3
number or to that particular age-group of people. Even on Notes
that particular group it of people having those religious. __________________
Similarly even then we find that carrying out research on __________________
the universe, i.e., on the entire population of that particular __________________
group may not be always possible because of the time factor __________________
and the money involved. In that case, what we usually do is
__________________
to take out samples selected from the entire population. In
__________________
selecting samples, we use the available sampling techniques
__________________
to draw from the total population. Apart from these, we need
to clarify certain other concepts. For example, what does __________________

the term empirical mean? __________________


__________________
Empirical means the observations and propositions which
are primarily based on some sense experiments or derived
from experience by methods of inductive logic including
mathematics and statistics. This technical definition is
difficult to understand. To be more clear we can define
empirical research as that type of research where we try to
deduce some logic and principles based on our survey
reports. In other words, when we want to analyze the survey
report using some mathematical and statistical tools and
deduce logic to authenticate our findings, we are said to
follow the empirical research method.

Other important things are attitude and opinion. Attitude


as we all know is the psychic change and it can neither be
demonstrated publicly nor can it directly be observed,
because most of the psychological variables as we know are
dormant or latent and to can never be measured unless we
use the most sophisticated tools, like scales etc., with suitable
weight factors. To be more clear if we try to measure it, use
the technical scales fixing or affixing weights for each and
every individual response. There are different schools of
thought which define attitude in different ways. We are not
interested to go in to those details but the measurement of
attitude is primarily done on the scoring.

What is an opinion? Opinion and attitude differ principally


in, that opinion is a verbal statement of a person and given
merely for the sake of responding or rather to answer a
particular question without any object or aim and based on
superficial and or partial information of the object or
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
4 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes situation about which the opinion is passed. But, attitude,


__________________ as we have already defined, is a state of mind. So it can well
__________________ be understood in the context of our discussion, that in
__________________ carrying out research, based on survey responses, it is better
__________________ to rely more on attitude than on opinion, because, through
attitude in a survey we can really understand the inner mind
__________________
or the inner psychology or inner perception of the individual
__________________
or the respondent, while, through opinion survey we can
__________________
simply provide superficial answers from the respondents
__________________ which obviously do not serve the purpose of any research.
__________________
Now, before we proceed into the depth of research we need
__________________
further clarifications on social science research. As has
already been mentioned, management research falls within
the category of social science research. Social science
research as we all know, is carried out on the society,
considering society as the field of experiment or to be more
precise as a laboratory. Research on management or research
on social sciences may be fundamental in nature when we
want to build a new theory and approach on any particular
discipline. For example, in the case of management, before
the Hawthrone experiment was carried out by Elton Mayo,
it was not known that work environment and human
relations have an effect on workers’ productivity. But Elton
Mayo and his colleagues by varying the work environment
could substantiate with their findings for the first time that
work environment has got some direct relationship with
workers’ productivity. This is certainly a fundamental
research which helped us to build a theory on management
science. But there are certain other areas of research on
management too, where we cannot simply develop any theory
or where we cannot put our contribution in to the
development of certain basic theories, etc. But we can
contradict, modify or alter any existing theory for better
understanding. For example, consider Maslow’s hierarchy
of needs theory, which defines various needs of a human
being in order of priority. To be precise, in our approach the
workers are having certain needs in order of their
importance or in order of their requirements. Maslow
believed fulfillment of one need leads to satisfaction and
motivate the workers and after the lower level need is
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 5
fulfilled, the workers strive for fulfillment of higher level Notes
needs. To contradict Maslow’s need hierarchy theory, __________________
Alderfer Clayton with his existence, relatedness and growth __________________
approach to motivation, substantiated that there may be __________________
certain need factors important for workers, which may even
__________________
arise before the needs of the lower level had actually been
__________________
fulfilled. For example, we know that there are physiological
needs related to shelter, clothing and food, etc. Unless these __________________

primary needs are fulfilled one cannot have the needs of __________________
higher level like social needs, etc. But today if we just try to __________________
find out from our society, we can get examples of different __________________
workers who get motivated simply by fulfilling certain social
__________________
needs. So in that case, these social needs or the fulfillment
of social needs is primarily important for that group of
workers, even though for them the physiological or other
needs have not been met to the required level. To give a
practical example, let us assume the case of journalists
employed by different newspaper houses. If we follow the
Maslow’s hierarchy of needs theory for their motivation, it
would obviously fail because journalists value their job
environment most. So obviously fulfillment of physiological
needs is not relatively more important for the journalists.
We have tried to analyze the difference between the
fundamental research, which help us to build a new theory
or a new approach to management and other types of
research which merely help us to contradict, alter, or modify
the existing research findings. We have focussed on the
example of Maslow deliberately, as his findings have been
altered, contradicted, or modified, in the context of new
findings or new research. Subsequent approaches to
motivation have been developed after contradicting Maslow.
Such approaches tried to define motivation from a more
scientific perspective.

Research, unless it is understood in its right spirit, cannot


be undertaken with success. What is important for
researchers is to take research more as a passion than for
career progress. From our experience, we see that most of
the students are interested in research merely to fulfil their
career prospects. In this way, certain researches are carried
out in a mechanical way. The researcher should have the
natural interest in the research.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
6 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes There are still certain things which we need to consider


__________________ before we go into the research matter in more depth. What
__________________ are the management areas and what is the scope available
__________________ for research in this field? Now to define management,
__________________ without going into the traditional approach we may put it in
__________________
more simplistic terms like: management is a deliberate effort
to achieve some defined goals. With definition alone we can
__________________
understand management from a broader perspective. Now
__________________
what is a ‘deliberate effort’? No management can succeed
__________________ unless the efforts or actions are initiated as a deliberate
__________________ process from the top. So the action should be initiated as a
__________________ deliberate policy on the part of the company or the industry
or the union whatever it may be. Now what is the ‘defined
goal’? We know that unless some one is having a defined
goal, the question of achieving it does not arise at all. So,
there should be some definite goal on the part of the
management or on the part of the country or organization,
whatever may be the case, and this definite goal obviously
should fulfil the requirements of law, legislation etc. The
foregoing discussions help us to understand that there are
different functional areas of management. Such functional
areas may be listed as follows:

1. Marketing

2. Human resource management

3. Production

4. Finance and accounting

5. Information technology

6. Material planning and production control

7. Purchasing function

8. Advertising and sales promotion function

9. Social functions, which enables a company to fulfil the


social responsibility, etc.

There may be certain other function too which, however,


depends more on the requirements of the environment
or particular environment that prevails in a company.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 7
So even beyond these functional areas, we have scope to do Notes
research. __________________
__________________
Let us try to understand what happens exactly in India. If
__________________
we try to understand the Indian economy as a whole, we get
a clear picture that we are lagging behind many other __________________

developed countries for obvious functional deficiency in __________________


management amongst other things. __________________
__________________
The functional deficiency in different management areas is
__________________
primarily responsible for the backwardness of economic
development of our country and may be grouped under the __________________

below mentioned broad categories. __________________

One is the financial management research. Research on


financial management includes different operations like:
financial analysis, capital structure, ratio analysis, etc.
Capital structure and ratio analysis enable a company to
ration the requirement of resources. Rationing of resources
means optimum utilization of different funds available for a
company. Unless the resources available in a company are
utilized we may need to find out various sources of funds
and costs of capital, liquidity constraint, for a more clear
understanding. As we all know unless a company is able to
maintain the required liquidity level for ensuring the
requirements of the working capital, the company is bound
to suffer from operational problems which in the long run
may affect profitability. Apart from such an affect, the non-
availability of working capital in time, and holding of excess
liquidity again affects the company’s profitability. So what
is needed is to ensure the minimum or the optimum holding
of the liquid assets of the company. Other areas of financial
management research may be the management of different
funds, management of assets and financial market analysis
etc. Now we all know that in any company, profitability
largely depends on the efficient handling of financial
resources. Financial Management Research enables us to
compare the operation of different companies with the help
of financial analysis. If a major portion of the working capital
is invested for purchase of raw-materials and inventories
alone, companies will unnecessarily hold capital in excess
of requirement and thereby strain their profitability. Apart
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
8 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes from holding of raw-materials and inventories which need


__________________ to be controlled, companies may have the problem of excess
__________________ storage costs for finished products. When we produce more
__________________ without taking in to cognizance the market potentialities,
__________________ we face the problem of storing finished products, which
occupy maximum godown space and involve a lot of costs.
__________________
__________________ Although rationing of resources is primarily a concern for
__________________ the cost management function, with the help of financial
__________________ management research, using different cost indicators or
__________________
using different ratios, we may find out what exactly is the
optimum resource holding and in what way a company is
__________________
able to hold that optimum level of the resources.
Management of funds and assets and analyzing the financial
market are also equally important for financial management
research. Management of funds requires an efficient planning
of borrowing long-term, short-term and medium-term loans
as per the requirements. When to borrow, how to borrow
and what to borrow is important for the company to
understand in the context of different data of the company
itself. The company may have some traditional way of
borrowing, like, borrowing from banks, borrowing from other
financial agencies like different development banks or, they
may borrow by issuing debentures in the market or they may
raise their capital by issuing shares, etc. All these
alternatives need to be compared and analyzed and the most
cost effective one needs to be chosen to ensure maximum
profitability.

Similarly, production research is vitally important for a


company specially those who are engaged in the
manufacturing operation. Unless goods or products are
manufactured right in time, or made available in time in the
market, the company may loose the market stake or the
market share. Thus, what is primarily important for a
production department is to ensure manufacturing of the
right goods and products in the right quantity and to make
it available in the market for the ultimate users. Thus there
is need for development of, different cost minimization
alternatives, work simplification, inventory control, product
design, process improvement, quality control, etc. With the
help of suitable capital budgeting unless the required
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 9
machines and equipments are made available right in time Notes
or unless the available machines are replaced with the __________________
modern available technology right in time, the production __________________
process itself will suffer. Product development and product __________________
design research is also equally important. The changing __________________
demands of the consumers heavily influence the product
__________________
design and product innovation. A good number of companies
__________________
gradually find themselves out of market simply because they
__________________
are not able to produce the right quantity or the right product
for the consumers as per their changing requirements. Free __________________

market or free competition to a great extent exerts pressure __________________


on the company. Thus unless a product is developed suitably, __________________
based on consumer requirements the production in the long
run is bound to suffer. Similarly, product development and
product innovation with work simplification and performance
improving approaches ensure cost minimization also. With
the help of industrial engineering, development of the
product with simultaneous reduction of the total costs is
possible. Cost efficiency, work simplification, work
measurement with the help of industrial engineering,
development of the product with value engineering approach,
among other things, enable us to find out the exact work
unit required for doing a job which in the long run helps us
to decide wage rate more scientifically. Unless inventory is
controlled rightly or the optimum utilization of inventory
resources is ensured, the production operation will definitely
suffer in the long run and will ultimately affect the company’s
profitability.

In marketing survey we have different areas of research like


marketing policy which involves, pricing, advertising service,
channels of distribution product research etc. Product
research helps us to find out the changing consumers’
requirements by carrying out survey. Improvement and use
value of the present product may be carried out, as also in
'packaging, based on consumers’ preferences. Marketing
research, as we all know, is primarily initiated to find out
the market potentiality of a particular company. With the
help economic and sales forecasts with the help of the survey
of consumers or the potential consumers of the society at
large, a company can take suitable decisions on expansion
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
10 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes or design or outline their production process exactly coping


__________________ with the requirements. Unless a company is undertaking
__________________ suitable marketing research it will definitely suffer in the
__________________ long run. As we all know, no company can afford to remain
__________________ in a static product or system indefinitely. To clarify this point
further, if a company remains in producing a static quantity
__________________
of some products keeping in view the market stake that they
__________________
are having for this much quantity alone, in future the
__________________
company may find themselves out of place because of their
__________________ competitors who may gradually grab the increased market
__________________ potentiality for that particular product or products.
__________________ Moreover unless suitable expansion programmes are
undertaken, from costing point of view, no company can
achieve production efficiency. Cost efficiency is only possible
when a company enjoys the benefit of producing an
economically run quantity and developing by-products. So,
undertaking suitable marketing research is of utmost
importance for a company. Apart from all these things, there
are other areas of research on marketing like, research on
channels of distribution, etc. A company sends its products
to the ultimate consumers through different channels of
distributions. Products may not reach the ultimate
consumers and users for reasons of inefficient channels of
distribution. Thus, a company needs to ensure that its
products reach the ultimate users on time. Efficient channels
of distribution also ensure development of further marketing
potential. Apart from the channels of distribution,
advertising as well as sales promotion research are
important areas of marketing.

Thus, a prospective researcher on management gets


immense scope to investigate into these three core functional
areas of management, i.e., finance, production and marketing.

Research in human resources strategy, its integration with


corporate planning and the development of human resource
policies are certain new areas in management research. The
research requires in depth understanding of Human
Resource Management in a business context, which addresses
a number of fundamental questions as under:

1. Is strategic HRM a reality?


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 11
2. Are all cross-sections of people involved in decision Notes
making? __________________
__________________
3. How can human resource strategies and practices be
__________________
developed to meet perceived threats and to reap
opportunities in a changed business environment? __________________
__________________
Strategic Human Resource Management (SHRM) requires
__________________
formation of HR objectives, strategies and policies. SHRM
__________________
also requires the provision of the skills and abilities matching
__________________
organizational overall objectives. In other words, SHRM
provides the framework, which ensures that an __________________

organization's people needs are met. __________________

Change strategies are inevitable consequence of many


human resource strategies. There are various types of change
initiatives that may interest a researcher. Turn around
change is financially driven to ensure corporate survival by
reducing unprofitable products and services. It involves
redesign of organizational structures, out-sourcing of non-
core activities and may even manpower redundancies.
Behavioural transformations are changing behaviour pattern
throughout the organization. Restructuring is the most
common form of major organizational change. Restructuring
need not be a defensive cost-cutting process but rather a
proactive attempt to achieve innovative products and
services. The goal should be synergy. Hence enthusiastic
researchers provided they understand the business issues
holisiticaly can do research on change strategies.

Research on competency development is also another


important aspect, which focuses on integration of business
plan of the company with selecting the right individual for
the job. Competency is defined as the smallest unit of on-
the-job behaviour that is observable, measurable, and
changes over time.

Similar contemporary areas of research interest are;


Innovation and creativity, Six-sigma practices, Supply chain
management, Knowledge management issues, Business
process outsourcing, Human resource outsourcing, Customer
Relationship Management (CRM), e-CRM, Web marketing,
Financial re-engineering, etc.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
12 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes In selecting a research topic, out of several potential topics,


__________________ the first task is to narrow down to the most potential one
__________________ out of the different alternatives. Selection of a research topic
__________________ is the first task of the research process. The research process
__________________ consists of a series of actions or steps necessary to carry out
research in a sequential and logical manner. Selection of
__________________
research topic although primarily depends on the guides’
__________________
instructions, the researcher should also take independent
__________________
decisions to a great extent for influencing the guide in this
__________________ matter. Certain pertinent factors in this regard are:
__________________
1. The selected topic should have some current implication.
__________________
2. There should be scope and amenities to carry out
the intended research, as for example; availability
of required data, feasibility of carrying out survey
etc.

3. The selected research topic should contribute to


knowledge in the desired field.

4. The selected topic should receive the necessary


concurrence from the research guide.

In our country most of the researchers, specially those who


carry out research under the universities’ or institutions’
affiliation, depend on the guides’ choice. Without depending
on the guides’ selection, the researcher should on his
own select the topic which has some interesting elements
in it. In cases where independent researches are carried
out, however, the question of guides’ consent does not
arise.

After selection of the research topic, the first and foremost


task for a researcher is to define the research problem. In
defining the research problem, what is important is to review
the existing literatures available in this field. The researcher
should study the existing or available literature with an
interdisciplinary perspective to base his research topic on
some acceptable background. The research problem should
also focus on the relevance of the present research with the
past works.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 13
The next task, after selection of the research topic and Notes
defining the problem, is to mention the objective of the __________________
present research. By the term, ‘objective’, we mean to what __________________
extent the present research differs from earlier ones by duly __________________
pointing out the deficiencies in the past research work __________________
related to this field. In simplistic terms the objective means,
__________________
what the researcher aims to achieve.
__________________
The third step involves survey of existing literature. In our __________________
earlier discussion we have mentioned, that research is a __________________
documented prose work resulting from an organized analysis __________________
of a subject. In all social science researches we need to
__________________
proceed through some pre-assumption which helps to
develop our premises for research. To understand the
premises for research, what is important is to review the
existing literatures from an interdisciplinary perspective.
Whatever literature is available, related to the topic of our
research we should focus on the basic principles, duly
summing them up. Literature review has two basic elements
in it: (a) to survey the existing books available in the field,
(b) to review all other published form of works like articles,
seminars, other research findings published in the form of
papers, etc. What is important in this case is to draw out a
suitable abstract of published works related to our research
topic. In drawing an abstract, we can take the help of
available published social science bibliographies for books
and various indices on social sciences for articles. In India
we have the problem of non-availability of an abstract
particularly for carrying out research on management. The
researcher here should prepare his own index duly going
through the related journals, etc., for a period in
chronological order in addition to his consultation of various
indices.

The next step after extensive literature survey is to develop


a hypothesis. Hypothesis is an assumption on which we base
our research premises and which we all test in the light of
our findings. In any social science research, the development
of a hypothesis is important as this helps us to authenticate
our research findings with due quantification and test. In
formulating the hypothesis what is important to remember
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
14 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes is to first ensure conceptual clarity. Apart from this,


__________________ hypothesis should be empirically testable, closest to the
__________________ observable, specific in nature, related to the body of theory,
__________________ simple and brief. Testing of hypothesis involves application
__________________ of both parametric and non-parametric (which helps us to
quantify the qualitative data) statistical tools after analysis
__________________
and interpretation of collected data. The various forms and
__________________
methods of hypothesis testing are left for subsequent
__________________
discussion.
__________________
__________________
Next task for a researcher is to draw out the methodology
for the research subject. Methodology here means in what
__________________
way the research is going to be carried out. Researcher here
should mention the logical sequence as follows:

1. Selection of samples with reasoning.

2. Sampling methods.

3. Survey outline.

4. Relating the survey results after necessary analysis and


interpretation with the main body of research.

5. Building of model, if any.

6. Interview methods and techniques, etc.

Next task for the researcher is to conclude his research by


restating his objective with due mention whether hypothesis
is accepted or rejected. Conclusion and suggestion should
be with respect to the objectives of the study. Conclusion
should also focus/suggest the scope of further research in
the area, possible implication of the study for building a
theory, model and should also focus on the extent to which
this research enriches knowledge of a particular discipline
or disciplines.

There are other different ways to formulate research steps.


But the above form is the most acceptable one. In some
research cases, analysis and interpretation of data is covered
under a separate head.

After designing the research steps in the above order, the


researcher should chalk out his chapter planning keeping
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 15
in view the research area. The most pragmatic approach to Notes
chapter planning first involves analyzing the research matter __________________
from a historical perspective. It should not be forgotten __________________
that no present is possible without a past and unless we __________________
analyze the past by maintaining due sequences, we cannot __________________
understand the process of development of the subject in its
__________________
present.
__________________
The next task in chapter outline is discussing the definitional __________________
aspects. After this, the research should mention the present __________________
trends from an international perspective. To relate the
__________________
present theories with the subject for research in chapter
__________________
planning, the researcher should duly focus on the extent to
which the present theory suffers from its own limitation and
in what way the researcher can contradict and rebuild the
theory/theories.

The next task for the researcher is to review his research


topic from an inter-disciplinary perspective. After outlining
the chapters in the above form or order, the researcher
should come to his own area of research with a more specific
approach. Here he should mention about his study
specialities and survey pattern. While elaborating the survey
data in chapters, he should duly point out survey
methodologies, particularly sampling procedures, interview
techniques, etc., and his survey area. He should analyze and
interpret the survey results and relate the survey findings
with the hypothesis.

The next task in chapter planning involves stating the model


developed after survey findings. Conclusions and suggestions
should follow later. Since research is documented work, it
should not be forgotten that the researcher should duly
authenticate his findings whenever he is taking references
from any existing publications for theoretical reasoning. He
should also authenticate his survey findings duly
acknowledging the sources or mentioning the survey pattern/
structure as and when it is necessary. To introduce
uncommon terms, he should take the help of foot notes
which should either be in the body of his research paper or
at the end of the research paper in the form of explanatory
notes.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
16 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Research steps and chapter outline, unless discussed with a


__________________ practical example, cannot be understood in their true spirit.
__________________ For that purpose we have given a sample research topic and
__________________ discussed the same in the light of our above discussion. The
__________________ intention is to prescribe an acceptable framework for
research. Its application varies, and also to a great extent
__________________
depends, on the research guides’ prescriptions.
__________________
__________________ In our subsequent chapters we will discuss writing
__________________ procedure of synopses, research paper, practical selection
__________________
of research topic, survey procedure, testing of hypothesis
etc., in more elaborate form. Our main focus is to develop
__________________
research so that the reader can become an independent
researcher by following a structured map.

Meaning of Research
Research has been defined by various authors in different
ways. It always begins with a question or a problem. It's
purpose is to find answers to questions through the
application of systematic and scientific methods. Thus,
research is the systematic approach towards purposeful
investigation. This needs formulating a hypothesis,
collection of data on relevant variables, analyzing and
interpreting the results and reaching conclusions either in
the form of a solution or certain generalizations.

Research is an academic activity and a systematized effort


to gain new knowledge.

Objectives of Research
The objective of research is to find out answers to questions
through the application of systematic and scientific way.

Though there is a specific purpose behind each research


study, however, the objectives can be broadly classified as
under:

u To obtain familiarity of a phenomenon.

u To determine the association or independence of an


activity.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 17
u To determine the characteristics of an individual or a Notes
group of activities and the frequency of its (or their) __________________
occurrence. __________________
__________________
Features of A Good Research Study __________________
i. Objectivity: A good research is objective in the sense __________________
that it must answer the research questions. This __________________
necessitates the formulation of a proper hypothesis, __________________
otherwise there may be lack of congruence between the __________________
research questions and the hypothesis.
__________________
ii. Control: A good research must be able to control all __________________
the variables. This requires randomization at all stages,
e.g., while selecting the subjects, the sample size and
the experimental treatments. This shall ensure an
adequate control over the independent variables.

iii. Generalizability: We should be able to have almost the


same result by using an identical methodology so that
we can apply the result to similar situations.

iv. Free from personal biases: A good research should


be free from the researcher’s personal biases and must
be based on objectivity and not subjectivity.

v. Systematic: A good research study must have various


well planned steps, i.e., all steps must be interrelated
and one step should lead to another step.

vi. Reproducible: A researcher should be able to get


approximately the same results by using an identical
methodology by conducting investigation on a
population having characteristics identical to the one
in the earlier study.

Hence, the following points must be ensured:

u Purpose clearly detailed.

u Research design thoroughly planned.

u High ethical standards applied.

u Limitations frankly revealed.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
18 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes u A complete and proper analysis.


__________________
u Findings presented unambiguously.
__________________
__________________ u Decision based conclusions.
__________________
__________________
Types of Research Studies
__________________ In carrying out research, we follow scientific norms. Science
__________________ is an organized knowledge. The essential feature of science
__________________ is the application of scientific method. In any science,
__________________ concepts, theories and other accumulated knowledge develop
from hypotheses, experimentation and analysis. In our
__________________
earlier discussion we have distinguished between physical
and social science and also introduced the concept of
fundamental research which helps us to build an original
theory and applied research aimed to contradict, modify
and is after an existing theory on social science. In other
words, research may be grouped under two major
categories:

1. Fundamental research

2. Applied research

Fundamental research here means research of a scientific


nature which has practically no connection or if at all there
is any connection, it is absolutely remote in nature as far as
social science research is concerned.

Applied research, on the other hand, means application of


basic scientific principles in different areas of social science
research.

The science of psychology has certain assumptions regarding


human behaviour which, in general, may be termed as
fundamental in nature. But, when we apply these
principles in different fields of management, say, personal
management (where we are more concerned with human
relations, etc.), marketing management (where we need to
study the marketing behaviours of consumers,
advertising response, etc.) our research becomes applied in
nature.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 19
Without going into the rigours of technical jargon, in social Notes
science research, we are more concerned with application __________________
of basic scientific principles in different functional areas of __________________
management on the one way and testing of all existing __________________
theories of management on the other. Social science research __________________
may be categorized under the following major heads:
__________________
1. Fundamental or basic research: Fundamental or __________________
basic research is primarily intended to find out certain __________________
basic principles, viz; John Robinson's imperfect __________________
competition theory in economics, Maslow's hierarchy of __________________
needs theory in motivation, Elton Mayo's hawthrone
__________________
experiment relating productivity with work
environment, etc.

2. Applied research: Applied research, as already stated,


is the application of available scientific methods in social
science research which helps to contradict, alter or
modify any existing theory or theories and helps to
formulate policy. Applied research, thus, is more
concerned with actual life. It also suggests remedial
measures to alleviate social problems. Example of
applied research may be:

a. John Huberman's: Discipline without punishment


theory.

b. Alderfer Clayton's: Existence relatedness and


growth theory which contradicted Maslow's
hierarchy of Needs theory and suggested a better
alternative.

3. Descriptive research: Descriptive research is usually


a fact finding approach generalizing a cross-sectional
study of the present situation. For example, a study on
problems of industrial relations in India with an inter-
disciplinary approach. It is classified under conclusive
research.

4. Historical research: Historical research is the


research on past social forces which have shaped the
present. For example, to study the present state of
Indian labour we may research on past historical forces.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
20 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 5. Formulative or exploratory research: Formulative


__________________ or exploratory research helps us to investigate any
__________________ problem with suitable hypothesis. This research on
__________________ social science is particularly important for clarification
__________________ of any concept and throwing new light for further
research on principles of developing hypothesis and its
__________________
testing with statistical tools.
__________________
__________________ 6. Experimental research: Although experimental
__________________ research is primarily possible in areas of physical
__________________
sciences, with the help of hypothesis, may also be carried
out in social sciences if such research enables us to
__________________
quantify the findings, to apply the statistical and
mathematical tools and to measure the results thus
quantified. It is also classified under conclusive
research.

7. Ex-post-facto research: Ex-post-facto research is an


empirical enquiry for situations that have already
occurred. For example, market failure for any
companies' product if studied or researched later may
be categorized as ex-post-facto research. Apart from
this, declaration or slow rate of growth in national
income when studied to reformulate the future policies
on this account, is classified as ex-post-facto research.

8. Case study approach: Case study approach to social


science research is particularly initiated at the micro
level. For example study of a particular industrial unit
or units, study of some banking units, etc., may be
categorized as a case-study. This type of study is
intensive in nature and data compilation requires
exhaustive study of the units with utmost sincerity. We
have discussed case study method separately.

Apart from the above research methods, there are certain


other general methods in social sciences. As for example,
survey research. Survey research, in one way, may be
categorized as a separate research method but in other way
may be defined as a tool which we follow in carrying out
social science research of any other methods already
mentioned above. Survey research invalues study of
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 21
population or sample based on some questionnaire to find Notes
out intended characteristics. When it is not possible to study __________________
the population because of obvious time constraint and cost __________________
factor, sampling method is applied, which ensures the __________________
probability of inclusion of each and every part of the __________________
population. Apart from research on management and in
__________________
business, survey research is widely used in different other
__________________
areas of social sciences like economics, sociology, psychology,
__________________
etc.
__________________
Some other types of research like evaluation research, etc. __________________
are directed to study or evaluate the performance of any
__________________
programmes/projects that have already been implemented.

Assessment research describes a situation that prevails


without any suitable explanation and suggestion for any
remedial actions. Comparative research is intended for
comparative study by collecting data from units under
comparison.

It may be mentioned that it is very difficult to categorize


any particular research under any above major heads. Thus,
our effort to nomenclate any research on social sciences
under above categories, in reality, is meaningless. However,
for the sake of our clear understanding such categories are
important. Commonly, we find that research on social
sciences are carried out with suitable development of a
hypothesis. After compilation and interpretation of data
collected for the purpose of such research, hypothesis is
tested and necessary conclusions are drawn. Whatever may
be the nature or methods of research, we should not forget
that inter-disciplinary treatment of research problem is
essential. By inter-disciplinary treatment, we mean
borrowing ideas from related disciplines connected with the
research topic for more authenticity. Research on
management and commerce discipline particularly requires
an integral approach to various disciplines for obvious
reasons as management is not a separate discipline in its
own right.

As has been said at the outset, research is a documented


prose writing. Documentation here means acknowledgement
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
22 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes of different sources particularly from where we have taken


__________________ references.
__________________
Organized analysis of a subject selected for research is the
__________________
prerequisite of success. While writing a research paper, the
__________________ subject of first and foremost importance is to select the
__________________ research topic. In doing so, we should also find out the
__________________ feasibility of locating information about the subject. Whatever
__________________ may be the method employed in carrying out research, we
__________________ should not forget that this should invariably be symptomatic
__________________
in nature.
__________________
Scientific Method
The application of valid and reliable research methods is
called scientific method. It has three distinct characteristics:

i. Objectivity: The scientific method should enable us to


classify facts accurately and carefully, without any bias.

ii. Accuracy of measurement: A mere collection and


classification of the facts may not be sufficient, one must
be able to make observations of their correlation and
sequence, which can be denied as a result of dissipation
imagination and painstaking efforts of the scientists.

iii. Self criticism: Scientist should critically examine their


own research as they are a group of people who are
never sure that they have can found the ultimate truth,
thus their studies are continuing an exhaustive. If
researchers are completely objective, their
measurements are accurate and their studies are
exhaustive, then their results will be valid and reliable.

Comparison of the Scientific Method and


Non-scientific Method
i. Objectivity: The scientific method is more objective as
compared to the non-scientific method, the reason being
that its development is normally a result of long years
of experience and good deal of thought by the scientist.
The hypothesis, a back bone of scientific method, can be
verified with the help of statistical techniques and
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 23
principles. This ensures better and more objectivity as Notes
compared to subjectiveness in decision making by non- __________________
scientific methods. __________________
__________________
ii. Degree of accuracy: The scientific method is more
precise than non-scientific methods. Though __________________

measurement is not always attempted in every scientific __________________


investigation, but wherever attempted even the __________________
quantitative concept can be defined precisely, measure __________________
effectively and helps in achieving results with higher __________________
degree of accuracy. __________________

iii. Continuing and exhaustive: The scientific method __________________


takes cognizance of the existing knowledge. The results
achieved by various researchers lend to the same
conclusions. This helps in accumulation of systematic
knowledge which is continuous and unending. However,
the same is not true in case of non-scientific method.

Research Methods and Research Methodology


u Research methods are the methods or techniques
employed by researchers in conducting research
operations.

u Research methodology is a scientific and systematic way


to solve research problems. A researcher has to design
his methodology, i.e., in addition to the knowledge of
methods/techniques, he has to apply the methodology
as well. The methodology may differ from problem to
problem. Thus, the scope of research methodology is
wider than research methods. In a way, research
methodology deals with the research methods and takes
into consideration the logic behind the methods, we use.

Organizing the Research Function


The research function as in research & development for a
manufacturing unit can be organized within an organization
of medium or large size but a marketing research
organization shall be very much different in nature as
the requirements and emphasis on research will be
different.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
24 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Many a time, companies do not set up their own departments


__________________ but prefer to engage outside agencies for their marketing
__________________ research requirements. The following points should be kept
__________________ in mind while hiring the services of outside agency:
__________________ u Technical expertise.
__________________
u Objective(s) of research studies.
__________________
__________________ u Economic considerations.
__________________
u Standing/reputation of the agency.
__________________
__________________ Some advantages of using an outside agency are:

i. It may propose a research proposal without any


constraint and may be worth considering.

ii. Timely advice can be sought which may have far


reaching effects on the business policy.

iii. The relevant problem to us today may have already been


studied by consultants and hence it shall help in saving
company’s time, money and energy.

Importance of Research in Management Decisions


The role of research has greatly increased in the field of
business and economy as a whole. The study of research
methods provides you with the knowledge and skills you
need to solve the problems and meet the challenges of today’s
modern pace of development. Three factors stimulate the
interest in a scientific research to decision making:

i. The manager’s increased need for more and better


information.

ii. The availability of improved techniques and tools to


meet this need.

iii. The resulting information overload.

The usefulness and contribution of research in assisting


marketing decisions is so crucial that it has given rise to the
opening of a new field altogether called ‘marketing research’.
Market research is basically the systematic gathering,
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 25
recording and analyzing of the facts about business problems Notes
with a view to investigate the structure and development of __________________
a market for the purpose of formulating efficient policies for __________________
purchasing, production and sales. Research with regard to __________________
demand and market factors has great utility in business. __________________
Market analysis has become an integral tool of business
__________________
policy. Once sales forecasting is done, the Master Production
__________________
Schedule (MPS) and Material Requirement Planning (MRP)
__________________
can be efficiently done within the limits of the projected
capacity based on the MPS. Budgetary control can be made __________________

more efficient, thus, replacing subjective business decisions __________________


with more logical and scientific decisions. __________________

Modern industry with its large scale operations tends to


create a gulf between the customer and the manufacturer.
Particularly when business is too big and operations are too
far-flung, one cannot depend upon casual contacts and
personal impressions. Research methodology has been
developed as the tool by which business executives keep in
touch with their customers. If an entrepreneur has to make
sound decisions, he must know who his customers are and
what they want. To a certain extent he relies on his salesmen
and his dealers to supply him with market information but
in recent years, more and more firms/executives have
turned to research methodology as a medium of
communication between the customer and the company.

Marketing research is the link between the manufacturer


and the consumer and the means of providing consumer-
orientation in all aspects of the marketing function. It is the
instrument of obtaining the knowledge about the market and
the consumer through objective methods which guard against
the manufacturer’s subjective bias.

Many researchers define marketing research as gathering,


recording and analyzing of all facts about problems relating
to the transfer and sale of goods and services from producer
to consumer.

Research methodology is an essential prerequisite for


consumer-oriented marketing. It is necessary for developing
the marketing strategy where in factors under the control
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
26 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes of the organization viz., product, distribution system,


__________________ advertising, promotion and price can be utilized so as to
__________________ obtain maximum results in the context of the factors outside
__________________ the control of the organization viz., economic environment,
__________________ competitor and laws of land.
__________________
Role of Research Methods/Research Methodology
__________________
in Business/Industry
__________________
__________________ Before undertaking any research it is essential to delimit
__________________ the primary objectives of the project and then to define the
methodology in as much detail as possible. This shall help
__________________
the researcher to utilize the research for decision making.
Any research done on an arbitrary basis, i.e., without
having explicit objectives or applying one’s personal
convenience in data collection and interpretation etc. shall
be of no use.

Common Uses of Research in Decision Making


The following are the areas in which research methodology
can help in better decision making:

1. Marketing research:

i. Product research: Assessment of suitability of


goods with respect to design and price.

ii. Market characteristics research (qualitative): Who


uses the product? Relationship between buyer and
user, buying motive, how a product is used,
analysis of consumption rates, units in which
product is purchased, customs and habits affecting
the use of a product, consumer attitudes, shopping
habits of consumers, brand loyalty, research of
special consumer groups, surveys of local markets,
basic economic analysis of the consumer market,
etc.

iii. Size of market (quantitative): Market potential,


total sales quota, territorial sales quota, quota for
individuals, concentration of sales and advertising
efforts; appraisal of efficiency, etc.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 27
iv. Competitive position and trends research Notes
__________________
v. Sales research: Analysis of sales records.
__________________
vi. Distribution research: Channels of distribution, __________________
distribution costs. __________________

vii. Advertising and promotion research: Testing and __________________

evaluating, advertising and promotion. __________________


__________________
viii. New product launching and product positioning.
__________________
2. Government policies and economic system: Research __________________
helps a decision maker in a number of ways, e.g., it can __________________
help in examining the consequences of each alternative
and help in bringing out the effect on economic
conditions. Various examples can be quoted, such as;
problems of big and small industries due to various
factors – upgradation of technology and its impact on
labour and supervisory deployment, effect of
government’s liberal policy, WTO and its new guidances,
ISO 9000/ 14000 standards and their impact on our
exports allocation of national resources on national
priority basis, etc.

3. Solving various operational and planning


problems of business and industry: Various types of
researches, e.g., market research, operations research
and motivational research, when combined together,
help in solving various complex problems of business
and industry in a number of ways. These techniques help
in replacing intuitive business decisions by more logical
and scientific decisions.

Market research helps in deciding forecasting which


helps in efficient production and investment
programmes such as purchasing and financial plans
which ultimately result in projected profit and loss
account.

4. Social relationships: Research in social sciences is


concerned with both – knowledge for self and knowledge
for helping in solving immediate problems of human
relations. It is a sort of formal training which helps an
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
28 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes individual in a better way, e.g.


__________________
i. It helps professionals to earn their livelihood.
__________________
__________________ ii. It helps students to know how to write and report
__________________ various findings.
__________________ iii. It helps philosophers and thinkers in their new
__________________ thinkings and ideas.
__________________
iv. It helps in developing new styles for creative
__________________
work.
__________________
__________________ v. It may help researchers, in general, to generalize
new theories.

Other areas have been already highlighted in our


introductory discussion. Also in subsequent chapters we
have discussed various contemporary areas encompassing
all functional areas of management.

Current Status of Research in India


Research in India can still be said to be in the growing phase.
Although there is an enormous interest evinced in research
by our business, they are not as yet completely sold on its
value of or as an aid to marketing. This is due partly to the
fact that in India, many product groups are still enjoying a
sellers’ market and consequently management is preoccupied
with problems relating to production and raw materials.

Secondly, there is a general misconception among Indian


businessmen that research is a very costly affair and hence
a luxury that they can ill afford. This is particularly true of
smaller firms and industries but paradoxically, it is they who
require research even more than the larger firms in order
to make marketing effort more efficient and to reduce or
eliminate wastage.

However, it is important for businessmen to realize that no


product can have a sellers' market in perpetuity, particularly
in a developing economy. When the reversal occurs, the
manufacturer who has kept in touch with the consumer all
through and practised consumer-oriented approach will reap
the rewards. Further, it should be realized that research
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 1 Introduction 29
increases the efficiency of marketing efforts, reduces wastage Notes
and thereby pays. __________________
__________________
Limitations of Research __________________

In India, research based decision making has just made a __________________


beginning. There are still a number of gaps in understanding __________________
between the researchers and its users. __________________
__________________
It may be made clear that research is there to help in decision
making and is not a substitute of decision making. Some of __________________

the limitations are given below: __________________


__________________
i. Research can provide a number of facts, but it doesn’t
provide actionable results.

ii. Some problems just cannot be researched, e.g., a precise


estimate of sales directly attributable to advertising.

iii. It cannot provide the answer to any problem but can


only provide a set of guidelines. Thus, management will
have to consider research findings in the light of other
factors such as production, finance and personnel, in
arriving at the decision.

iv. Managers normally do not include research in the


overall problem or total decision making. Partial
decisions on specific issues and without proper timing
and budget may not help in satisfactory solutions.

v. Managers rely more on intuition and judgement rather


than on research.

Review Questions
1. What is “research”? What are the two broad categories
in which it can be divided?

2. Distinguish between:

i. Basic research and applied research.

ii. Exploratory and conclusive research.

iii. Research methods and research methodology.

iv. Scientific and non-scientific method.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
30 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 3. Do you agree with the statement “research is much


__________________ concerned with proper fact finding, analysis and
__________________ evaluation?” Give reasons in support of your answer.
__________________
4. A multinational company is thinking of introducing a
__________________ general type of credit card. Before its launch, it wants
__________________ to know more about the current methods in use for
__________________ goods or services costing more than Rs 1000 . The
__________________ company wants to have the following information:
__________________
i. Method preferred by consumers : cash, cheque,
__________________ credit card or electronic fund transfer.
__________________
ii. Which method is mostly used in a 30-day period?

iii. Do people of different age, education, sex, and


income matter?

iv. Do people use different methods of payments for


personal expenses versus business expenses?

5. Explain the role of research in management decision


making.

6. Explain how research helps in making better decision


in following areas:

i. Marketing research

ii. Motivational research

iii. Government policy and economic decisions

iv. Solving business and industry related operational


problems.

v. Social sciences.
31

Unit 2 Notes
__________________
__________________
Defining Problem __________________
__________________
__________________
__________________
__________________
__________________

Objectives __________________
__________________
After reading this unit you will be able to:

y Have an insight about various components which affect a decision.

y Describe various types of models classified under (i) certainty (ii)


risk (iii) uncertainty and steps to be followed while calculating
expected monetary value.

y Understand decision tree diagram.

Formulation of the Problem


Formulation of the problem means defining the problem
precisely. In other words, A problem well defined is half
solved. In operations research, we say that formulation of
problem is often more essential than its solution because
when the problem is formulated, an appropriate technique
can be applied to generate alternative solutions. Choosing
the best alternative is the best decision under the given
circumstances.

Steps involved in defining a problem are:

i. Statement of the problem in a general way.

ii. Understanding the nature of the problem.

iii. Surveying the available literature.

iv. Developing the idea through discussions.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
32 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes v. Rephrasing the research problem into a working


__________________ proposition.
__________________
Once the problem has been selected, the same has to be
__________________
understood thoroughly and then the same has to be reframed
__________________
into meaningful terms from an analytical point of view. The
__________________
best way is to discuss the problem with friends or colleagues
__________________
or with those who have the knowledge of it. Both parties,
__________________
the researcher and/or the concerned manager or customer,
__________________
must agree on the specific nature of the research problem.
__________________
Ideally, both the parties must ascertain the priorities of the
__________________ issues involved, scope or potential benefits, cost as well as
the time required to conduct the study. Researcher must
specify the exact issues being examined and the underlying
logic in setting the priorities. Moreover, the research must
define the boundaries of the population covered in the
proposal.

In essence, a proper formulation of the research problem


starting with objectives would enable a researcher to go
ahead in the proper direction. Finally, it may be noted that
problem formulation would have focus on what sort of
decision issues are tackled. In other words, it gives a clear
idea of the research itself.

Formulation of Hypothesis
Concept of Hypothesis
A hypothesis is a proposition – a tentative assumption which
a researcher wants to test for its logical or empirical
consequences. Hypotheses are more useful when stated in
precise and clearly defined terms. It may be mentioned that
though a hypothesis is useful it is not always necessary,
specially in case of exploratory researches. However, in a
problem-oriented research, it is necessary to formulate a
hypothesis or hypotheses. In such researches, hypotheses
are generally concerned with the causes of a certain
phenomenon or a relationship between two or more variables
under investigation.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 33
Hypothesis Testing Notes
__________________
A number of steps are involved in testing a hypothesis:
__________________
i. Formulate a hypothesis. __________________
__________________
ii. Set up a suitable significance level.
__________________
iii. Choose a test criterion. __________________
__________________
iv. Compute the statistic.
__________________
v. Make decision. __________________
__________________
u Formulate a hypothesis: Let us discuss about
introduction of a new drug. The drug is tested on a few
patients and based on the response from patients, a
decision has to be made whether the drug should be
introduced or not. We make certain assumptions about
the parameter to be tested – these assumptions are
known as hypotheses.

We start with a ‘null hypothesis’: H0 : m = 100. This is a


claim or hypothesis about the values or population
parameters.

This is tested against alternate hypothesis , m1 : m ¹100.

The null hypothesis is tested with available evidence


and a decision is made whether to accept this hypothesis
or reject it. If the null hypothesis is rejected, we accept
the alternate hypothesis.

u Setting up a suitable significance level: There are


two types of errors that can be committed in making
decisions regarding accepting or rejecting the null
hypothesis:

– Type I error: An error made in rejecting the null


hypothesis, when in fact it is true.

– Type II error : An error made in accepting the null


hypothesis, when in fact it is untrue.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
34 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The level of significance signifies the probability


__________________ of committing Type I error and is generally taken
__________________ as equal to 5% (a = .05). This means that even after
__________________ testing the hypothesis, when a decision is made, we may
__________________ still be committing an error in rejecting the null
__________________ hypothesis. Sometimes, the value of ‘a’ is taken as
__________________ .01 but it is the discretion of the investigator, depending
__________________ upon the sensitivity of the study.
__________________
u Choose a test criterion: This means selection of
__________________
a suitable probability distribution that can be used
__________________ for the available information under consideration.
The different distributions that are normally used
are :

– Normal distribution: Z characteristic, this is


most often used, when the samples are more
than 30.

– T-distribution: ‘T’ test is used for small samples


only.

– F – test and chi-square test.

u Compute the test characteristic: This involves the


actual collection and computation of the sample data.
For the case under consideration , we have to find the
sample mean () and then compute the calculated ‘Z’. This
calculated value (absolute) is compared with tabulated
value obtained from normal distribution table against
the decided criterion ( value of ‘a’ and one tail or two
tail).

u Make a decision: If the calculated value of the test


characteristic is greater than the tabulated value, the
null hypothesis is rejected and the alternate hypothesis
is accepted. Talking in terms of critical region, the value
of calculated characteristic falls outside the acceptance
region.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 35
Notes
__________________
__________________
__________________
__________________
__________________
__________________
__________________
__________________
__________________
__________________

Figure 2.1

Developing the Research Plan


The next step of the research process calls for determining
the information needed, developing a plan for gathering it
efficiently and presenting the management of the
organization. The plan outlines sources of secondary
data and spells out the specific research approaches,
contact methods, sampling plan and instruments that the
researcher will use to gather primary data. First of all
research objectives must be translated into specific
information needs. To meet the management information
needs, the researcher can gather secondary data and primary
data or both.

Implementing the Research Plan


The researcher next puts the research plan into action.
This involves collecting, processing and analyzing the
information.

Interpreting and Reporting the Findings


After collecting, processing and analyzing the information
collected, the researcher must now interpret the
findings, draw conclusions and then report to the
management.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
36 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
How To Prepare A Synopsis
__________________
Synopsis is an abstract form of research which underlines the research
__________________ procedure followed and is presented before the guide for evaluating its
__________________ potentiality. In one sentence it may be described as a condensation of
__________________ the final report. The structure of synopsis varies and also depends on the
guides’ choice. However, for our understanding a common structure may
__________________ be framed as under:
__________________
1. Defining the problem: In defining the problem of the research
__________________ objective, definition of key terms, general background information,
__________________ limitations of the study and order of presentation should be
mentioned in brief.
__________________
__________________ 2. Review of existing literature: In this head, researcher should
study the summary of different points of view on the subject matter
as found in books, periodicals and approach to be followed at the
time of writing.

3. Conceptual framework and methodology: Under this head the


researcher should first make a statement of the hypothesis.
Discussion on the research methodology used, duly pointing out
the relationship between the hypothesis and objective of the study
and finally discussions about the sources and means of obtaining
data should also be made. In this head the researcher should also
point out the limitations of methodology, if any, and the natural
crises from which the research is bound to suffer for such obvious
limitations.

4. Analysis of data: Analysis of the data involves testing of hypothesis


from data collected and key conclusions thus arrived.

5. General conclusions: In general conclusions, the researcher


should make a restatement of objectives. Conclusion with respect
to the acceptance or rejection of hypothesis, conclusion with respect
to the stated objectives, suggested areas of further research and
final discussion of possible implications of the study for a model,
group, theory and discipline.

Finally the researcher should mention about the bibliographies and


appendices. The above format is drawn after a standard framework
followed internationally in preparation of a synopsis. However, in our
country, keeping in view the object of research, style and structure of
synopsis varies and quite often it is found that the research guide
exercises his own discretion in synopsis preparation than following some
acceptable international norms. A standard format for preparation of
synopsis commonly used in management and commerce research in
India may be drawn as follows:

1. Introduction: This includes definition of the problem and its review


from a historical perspective.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 37
2. Objective of the study: It defines the research purpose and its Notes
speciality from the existing available research in the related field. __________________

3. Literature review: It includes among other things, different sources __________________


from which the required abstract is drawn. __________________

4. Methodology: It is intended to draw out the sequences followed in __________________


research and ways and manners of carrying out the survey and __________________
compilation of data.
__________________
5. Hypothesis: It is a formal statement relating to the research problem __________________
and it need to be tested based on the researchers’ findings. __________________
6. Model: It underlies the nature and structure of the model that the __________________
researcher is going to build in the light of survey findings. __________________

And finally the chapter outlines.

Any chapter outline follows a logical sequence right from reviewing the
research problem from a historical perspective to the conclusions and
suggestions part. In the chapter outline, we also need to review research
problem duly drawing it in several supporting chapters with inter-
disciplinary connotation.

Most universities require the researcher to submit a synopsis duly drawing


out the research problem, following suitable norms, in approximately a
thousand words at the time of registration. The research guide, after
thorough checking of the synopsis, suggests necessary changes, if any,
and after evaluating the potentiality, clears the registration. The
researcher, after sometime, which varies from university to university, is
directed to submit the said synopsis with some elaboration containing
approximately 4000 words. In some universities, a seminar is arranged
for the said research topic with some experts and the researcher
is required to answer the different questions raised by those experts.
After successful seminar, the researcher is given final clearance for
going ahead with the research problem in consultation with the
research guide from time to time. While preparing the synopsis of 4000
words before the seminar, the researcher needs to include at least some
portion of the survey results which he has carried out for the purpose of
research.

In some other universities, of course, the seminar is not required and


researcher is given final clearance for going ahead with the research
problem merely after submitting the synopsis of thousand words. In such
cases where an abstract of 4000 words is required, the researcher needs
to incorporate very carefully the survey results of some units or some
samples with due condensation, following the structured framework. The
researcher also needs to submit the details of books and journals he has
consulted for satisfaction of the experts.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
38 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Marketing research provides information about consumers,


__________________ channel members, competitors, changes, and trends in the
__________________ marketplace, and other aspects of the firm’s environment.
__________________ The purpose of marketing research is to assess information
__________________ needs and provide the relevant information so as to improve
marketing decision-making.
__________________
__________________ Marketing research may be conducted internally or may be
__________________ purchased from external supplies. The statistical decision
__________________ theory in the present context refers to value of information
__________________
under conditions of uncertainty. In this case the probability
theory has a vital role.
__________________

Introduction
Decision in any aspect refers to the selection of the best or
favourable alternative out of several alternatives. In earlier
days decisions were made mainly on personal judgement.

Now-a-days in marketing one has to consider various aspects


such as planning, organization, direction, command and
control. While performing so many activities, marketing has
to face many situations from which the best choice is to be
taken. This choice making is known as decision-making.
Simply, decision is a selection from two or more courses of
action. Decision-making may be defined as–‘a process of best
selection from a set of alternative courses of action, which is
supposed to meet objectives upto the satisfaction of the
decision maker’.

The knowledge of statistical techniques help to select the


best action. Measures of central tendency show the average
value, dispersion measures the average deviations of
variables. Techniques of correlation and regression help in
establishing functional relation of variables, where
forecasting helps in scientific prediction. Again, time series
analysis gives us an estimation of variables subject to
condition. The test of hypothesis procedures were designed
to test a statistical statement about a population (the null
hypothesis) for a given level of significance.

The statistical decision theory (also known as Bayesian


decision theory) in the present context refers to value of
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 39
information under conditions of uncertainty. In this case the Notes
probability theory has a vital role. As such, this probability __________________
theory will be used more frequently in the decision-making __________________
theory under uncertainty and risk. __________________

The statistical decision theory tries to analyze the logical __________________

structure of the problem in alternative actions, states of __________________


nature, possible outcomes and likely payoffs from each such __________________
outcome. __________________
__________________
Components in Decision-making __________________
The management of an industry or an organization is involved __________________
in taking decisions about the development or expansion of
the concern. No doubt the concern has to face many problems
from time to time. The management has to overcome or solve
these problems. These problems may be regarding the
availability of resources, quality of production, costing,
publicity or promotion, etc. To illustrate for the publicity of
goods produced, the manufacturer or seller may take the
help of TV, radio, newspapers or door-to-door promotion.
Many such endless problems are there. In each of these
decision-making problems, there are some common
components. These are discussed as follows.

Course of Action
Decision-making problems deal with the selection of a single
act from a set of alternative acts. If two or more alternative
courses of action occur in a problem, then making a decision
is necessary to select only one course of action. In the above
example of advertising goods produced, there are so many
alternative courses of actions. The final choice of any one
will depend on the payoff (or money value) of each strategy
under the circumstances.

Let the acts or actions be symbolised by a1, a2, a3, a4, ...........;
then the totality of all these actions is known as action space
denoted by A. For four actions a 1, a 2, a 3 , a 4; A = action
space = {a1, a2, a3} or A = {A1, A2, A3}. In a tree diagram the
acts or actions are given after the initial fork as shown in
Figure 2.2. Acts may be also be represented in the following
matrix form, i.e., either in row or column way.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
40 Research Methodology fo
n ~;
k ; k ‘ kf D r o’
k~ e

Notes
Acts
__________________
A1
__________________ Start A2
A3
__________________ Acts A1 A2 A3 A4 ...... An
Acts Æ
A1
__________________ A2
A3
__________________ OR
A4

........
__________________
An
__________________
__________________
Figure 2.2
__________________
States of Nature (Events)
__________________
If the outcome of some action (or act) is not known in
advance, then that outcome is uncertain, but is vital for the
choice of any act. When there are many possible outcomes
(or states of nature) of an event, one cannot predict what
will happen–it is only in terms of probability that one may
forecast. The various states of nature are outside a firm and,
as such, not under their control, e.g., consumer demand,
change of taste, improvement of technology, etc. These affect
the payoff and, hence, the choice of an act. A set of states of
nature may be represented in any one of the following ways:

S = {S1, S2, S3, ......., Sn}


or E = {E1, E2, E3 ......., En}
or W = {q1, q2, q3 .......}

For example, if a cosmetic product is marketed, it may be


highly liked by customers (outcome q1) or it may not appeal
at all (outcome q2), or it may satisfy only a small fraction, say
20% (outcome q3).
\ W = {q1, q2, q3}

In a tree diagram the places are text to acts. We may also


get another act on the happening of an event in Figure 2.3.

Acts Events Second Act Acts Events


a1
E1
E1 a2
A1 E2
E1
a1
A1 A2
E2 E2
A2 a2 A3 E1

A3 E2

Figure 2.3
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 41
In matrix form they may be represented in either of the Notes
following two ways: __________________

States Acts __________________


of ® __________________
nature S1 S2 ............ OR States A1 A2 ..........
Acts ¯ of __________________
nature ¯ __________________
A1 S1
__________________
A2 S2 __________________
__________________
Outcomes
__________________
There is an outcome (or consequence) of the combination of __________________
each act with each possible state of nature. This outcome is
also known as conditional value. In a tree diagram, outcomes
are placed after events.

Acts Events Outcomes


E1
O11
E2
A1 O12
E1
A2 O21
E2
A3 O22
E1
O31
E2
O32

Figure 2.4

In general, if there are m possible actions and n admissible


states of nature, the consequences will be m × n in
number.

It may be noted here that these consequences may be


evaluated in several ways such as (i) in terms of profit (ii) in
terms of cost (iii) in terms of opportunity loss (iv) utility.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
42 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Consequences of m × n in number are shown in centre cell in


__________________ row and column of a matrix:
__________________
Table 2.1
__________________
__________________ States of
nature S1 S2 ... Sj ... Sn
__________________
®
__________________
Acts ¯
__________________ A1 O11 O12 ... O1j ... O1n
A2 O21 O22 ... O2j ... O2n
__________________ ... ... ... ... ... ...
__________________ ... ... ... ... ... ...
Ai Oi1 Oi2 ... Oij ... Oin
__________________ ... ... ... ... ... ...
... ... ... ... ... ...
Am Om1 Om2 ... Omj ... Omn

Here outcome O 21 means 2 nd row and 1 st column and


Oij means ith row and jth column and so on.

Example 2.1
In case of supply of goods and demand in a market, the
outcome may be unsold goods (+), shortage of goods (-) or no
stock (0). In the following matrix form, outcomes of supply
and demand are shown in Table 2.2.

Table 2.2

States of
nature Demand
®
Acts 100 101 102 103
¯
100 0 -1 -2 -3
101 +1 0 -1 -2
Supply

102 +2 +1 0 -1
103 +3 +2 +1 0

These outcomes will clearly indicate if there is monetary


gain or loss, i.e., payoff for each outcome.

Payoff: The result of combinations of an act with each of the


states of nature is the outcome and monetary gain or loss of
each such outcome is the payoff. This means that the
expression payoff should be in quantitative form.

Payoff may be also in terms of cost saving or time saving.


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 43
Utility: Utility of each outcome is also a quantitative Notes
expression, but takes the preference of the individual __________________
decision-maker. __________________
__________________
Regret (or opportunity loss): The difference between the
highest possible profit for a state of nature and the actual __________________

profit obtained for the particular action taken is known as __________________


opportunity loss. __________________
__________________
Thus, this is the loss incurred due to failure of not adopting
__________________
the best possible course of action. Opportunity losses are
calculated separately for each state of nature. __________________
__________________
Refer the Table 2.3. Let the payoff of the outcomes in the 1st
row be P11, P12, ......, P1n and similarly for the other rows.

Table 2.3: Payoff Table

States of
nature
®
Acts S1 S2 ... Sn
¯
A1 P11 P12 ... P1n
A2 P21 P22 ... P2n
... ... ... ...
... ... ... ...
Am Pm1 Sm2 ... Pmn

If now m1 is the maximum profit (or payoff) corresponding


to S1, then the regret or opportunity loss will be maximum
payoff under state of nature–payoff, i.e.,

Table 2.4: Regret (or Opportunity Loss) Table

States of
nature
®
Acts S1 S2 ... Sn
¯
A1 M1—P11 M2—P12 ... Mn—P1n
A2 M1—P21 M2—P22 ... Mn—P2n

Am M1—Pm1 M2—Pm2 ... Mn—Pmn

Example 2.2
A restaurant prepares a food item at a total average cost of
Rs 3 per plate and sells it at a price of Rs 5. The food is
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
44 Research Methodology fo
n ~;
k ; k ‘ kf D r o’
k~ e

Notes prepared in the morning and is sold during the same day.
__________________ Unsold foods at the end of the day are spoiled and must be
__________________ thrown away. According to past sales number of plates is
__________________ not less than 20 or greater than 23. You are to formulate the
__________________ (i) action space, (ii) state of nature space, (iii) payoff table,
(iv) loss table.
__________________
__________________ Solution
__________________
The shopkeeper will not prepare less than 20 plates or more
__________________
than 23 plates. Thus, the acts or courses of action open to
__________________
him are:
__________________
a1 = prepare 20 plates Thus, action space or

a2 = " 21 " A={a1 a2 a3 a4}

a3 = " 22 "

a4 = " 23 "

The state of nature is the daily demand for food plates. There
are four possible states of nature, i.e.

S1 = demand is 20 plates Hence, the state of nature

S2 = " " 21 " space or W = {S1 S2 S3 S4}

S3 = " " 22 "

S4 = " " 23 "

The uncertainty element in the given problem is the daily


demand. The profit of the shopkeeper is subject to the daily
demand.

Let n = quantity demand


and m = quantity produced
For n ≥ m, Marginal Profit (MP) = 5 – 3=Rs 2
For m > n, Marginal Loss (ML) = Rs 3 due to unsold stock

So, the conditional values for each consequence (or cell) in


the payoff table are:

(MP) × units sold — (MP) × units unsold.

For example, in case a2 = 21, S3 = 22, here m = 21, n = 22


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 45
i.e., n > m, net payoff = 2 × 21 - 0 = 42 Notes
__________________
Again, for a4 = 23, S2 = 21 i.e. m = 23, n = 21, m > n
__________________
net payoff = 2 × 21 - 4 × 2 = 42 – 8 = 34 and so on. __________________
__________________
Table 2.5: Payoff Table
__________________
Demand __________________
(n)
® (S1) (S2) (S3) (S4) __________________
Supply 20 21 22 23 __________________
(m) ¯
__________________
(a1) 40 40 40 40
20 40 __________________

(a2) 36 42 42 42
21
(a3) 32 38 44 44
22
(a4) 28 34 40 46
23

To calculate the opportunity loss we first determine the


maximum payoff in each state of nature:

In first state, maximum payoff = 40


Second, " " = 42
Third, " " = 44
Fourth, " " = 46

Table 2.6: Loss Table Corresponding to the


above Payoff Table

Demand
n
®
Supply 20 21 22 23

20 40 – 40 = 0 42 – 40 = 2 44 – 40 = 4 46 – 40 = 6
21 40 – 36 = 4 42 – 42 = 0 44 – 42 = 2 46 – 42 = 4
22 40 – 32 = 8 42 – 38 = 4 44 – 44 = 0 46 – 44 = 2
23 40 – 28 = 12 42 – 34 = 8 44 – 40 = 4 46 – 46 = 0

Types of Decision Models


The aim of the decision theory is to guide the decision-maker
in choosing the best course of action. Depending upon the
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
46 Research Methodology fo
n ~;
k ; k ‘ kf D r o’
k~ e

Notes available information about the occurrence of various states


__________________ of nature and environment, the decision models may be
__________________ classified into the following types: (i) certainty (ii) risk
__________________ (iii) uncertainty.
__________________
Decision-making Under Certainty
__________________
__________________ In this case the decision-maker can easily select the course
__________________
of action, as he presumes that only one state of nature is
relevant for his purposes. For example, suppose a person
__________________
desires to deposit Rs 10,000 for a certain period. Bank
__________________
deposit gives 12%, Unit Trust of India offers 14% and
__________________
Government bond’s rate is 11% p.a. All the investments are
fully secured. Then surely investment in Unit Trust of India
is the best choice.

Decision-making Under Risk


Here the decision-maker faces many states of nature. As
such, he is supposed to believe authentic information,
knowledge, past experience or happenings to enable him to
allot probability values to the likelihood of occurrence of
each state of nature. Sometimes with reference to past
records, experience or information, probabilities to future
events could be allotted. On the basis of probability
distribution of the states of nature, one may select the best
course of action having the highest expected payoff value.

The expected value (or expected monetary value, i.e., EMV


is widely used to evaluate the alternative course of action
(or act). The expected value is calculated by the following
formula for one row only:

S Pj Oij = P1 Oi1 + P2 O12 + ............ + Pn Oin’

Where P1 to Pn indicate the probabilities of states of nature


(or events) and O ij the payoffs of the outcomes due to
combination of each event and act.

Steps for Calculating Expected Monetary Value


(EMV)
1. Construct the payoff table listing the different courses
of action and states of nature.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 47
2. List the payoffs thus obtained with the corresponding Notes
probability of each state of nature. __________________
__________________
3. Multiply the payoffs with the associated probabilities
__________________
and add up these values for each course of action which
gives the expected value. __________________
__________________
4. On the basis of these expected values, determine the
__________________
best course of action or strategy.
__________________

Example 2.3 __________________


__________________
The payoff table for three courses of action (A) with three
__________________
states of nature (E) (or events) with their respective
probabilities (P) are given. Find the best course of action.

Events E1 E2 E3
(P) .3 .4 .3
Acts
A1 2.5 2 -1
A2 4 2.6 0
A3 3 1.8 1

Solution
Let us calculate the expected value of each act:

A1 : 2.5 (.3) + 2 (.4) + (-1) (.3) = .75 + .8 - .3 = 1.25

A2 : 4 (.3) + 2.6 (.4) + 0 (.3) = 1.2 + 1.04 + 0 = 2.24

A3 : 3 (.3) + 1.8 (.4) + 1 (.3) = .9 + .72 + .3 = 1.92

From the above calculations we find that A2 is the best. So


the decision-maker will select the alternative A2 amongst
the three courses of action.

Example 2.4
You are given the following payoffs of three acts A1, A2, A3
and the events E1, E2, E3.

Acts
States of nature A1 A2 A3
E1 25 -10 -125
E2 400 440 400
E3 650 740 750
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
48 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The probabilities of the states of nature are, respectively,


__________________ 0.1, 0.7 and 0.2. Calculate and tabulate EMV and conclude
__________________ which of the acts can be chosen as the best.
__________________
Solution
__________________
__________________ Calculation and Tabulation of EMV

__________________ Events Prob. Acts


__________________ A1 A2 A3

__________________ E1 .1 25 × .1 = 2.5 -10 × .1 = -1 - 125 × .1 = -12.5


E2 .7 400 × .7 = 280 440 × .7 = 308 400 × .7 = 280
__________________
E3 .2 650 × .2 = 130 740 × .2 = 148 750 × .2 = 150
__________________
EMV 412.5 455 417.5
A2 is best.

Example 2.5
The management is facing the problem of selecting one of
the following two products for manufacturing. After market
research and survey, the probability matrix is as follows:
States of
nature Good Fair Poor
Acts

Table fan .3 .5 .2
Ceiling fan .4 .45 .15

The profits that the management can make for different


levels of market acceptability of the two types of fans are as
follows:
States of
nature Profit in Rs if the market is
Acts Good Fair Poor

Table fan 10,000 15,000 3,000


Ceiling fan 14,000 16,000 Loss 4,000

Calculate the expected value of the choice of alternatives


and accordingly advise the management.

Solution
Let us put the above information in a payoff table with
associated probabilities and the states of nature, i.e., market
conditions.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 49
Notes
Market
conditions® Good Fair Poor __________________
Strategy¯
__________________
Table fan Probability .3 .5 .2 __________________
profits (Rs) 10,000 15,000 3000
__________________
Ceiling fan Probability .4 .45 .15
profits (Rs) 14,000 16,000 -4000 __________________
__________________
Expected value for: __________________
__________________
Table fan = .3 × 10000 + .5 × 15000 + .2 × 3000 = Rs 11, 000
__________________
Ceiling fan = .4 × 14000 + .45 × 16000 + .15 × (-4000) = Rs 12,200 __________________

As the expected payoff for ceiling fans is higher, so this


product should be preferred by the management.

Example 2.6
Management is faced with the problem of choosing one of
the three products for manufacturing. The potential demand
for each product may turn out to be good, moderate or poor.
The probabilities for each of the states of nature were
estimated as follows:

Nature of Demand

Product Good Moderate Poor

X 0.70 0.20 0.10

Y 0.50 0.30 0.20

Z 0.40 0.50 0.10

The estimated profit or loss under the three states may be


taken as:

Rs Rs Rs

X 30,000 20,000 10,000

Y 60,000 30,000 20,000

Z 40,000 10,000 -15,000 (loss)

Prepare the expected value table and advise the management


about the choice of product.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
50 Research Methodology fo
n ~;
k ; k ‘ kf D r o’
k~ e

Notes Solution
__________________
Demand Expected
__________________ Æ Good Moderate Poor value
__________________ ProductØ (in ’000 Rs)

__________________ X Probability 0.70 0.20 0.10 26


Profit (in ’000 Rs) 30 20 10
__________________
Y Probability 0.30 0.30 0.20 43
__________________ Profit (in ’000 Rs) 60 30 20
__________________ Z Probability 0.40 0.50 0.10 19.5
__________________ Profit (in ’000 Rs) 40 10 -15
__________________
Calculation of expected value (in ’000 Rs) for:
__________________
X : .70 × 30 + .20 + 20 + .10 × 10 = 21 + 4 + 1 = 26

Y : .50 × 60 + .30 × 30 + .20 × 20 = 30 + 9 + 4 = 43

Z : .40 × 40 + .50 × 10 + 10 × (-15) = 16 + 5 – 1.5 = 19.5

As the expected value of Y is higher, the management should


choose this product.

Example 2.7
If you make a unit product and it is sold, you gain Rs 5; if you
make a unit and it is not sold you lose Rs 3. Suppose the
probability distribution of the number of units demanded is
as follows:

No. of units

Demanded 0 1 2 3 4 5 or above

Probability .20 .20 .25 .30 05

How many units should you make?

Solution
Let m = no. of units produced or stocked.

n = no. of units demanded.

Now for n ≥ m, profit = 5 m

and for m ≥ n, profit = 5n - 3 (m – n) = 8n – 3m


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 51
Payoff Table Notes
__________________
Demand (n)
0 1 2 3 4 5 __________________
Supply (m) __________________
0 0 0 0 0 0 0
1 -3 5 5 5 5 5 __________________
2 -6 2 -10 10 10 10 __________________
3 -9 -1 7 15 15 15
4 -12 -4 4 12 20 20 __________________
5 -15 -7 1 9 17 25 __________________

Unit-wise expected gain are is follows: __________________


__________________
Supply Expected gain
__________________
(m)

0 0(.2) + 0(.2) + 0(.25) + 0(.3) + 0(.05) + 0(0) = Rs 0


1 -3(.2) + 5(.2) + 5(.25) + 5(.3) + 5(.05) + 5(0) = Rs 2.5
2 -6(.2) + 2(.2) _ 10(.25) + 10(.3) + 10(.05) + 10(0) = Rs 5.3
3 -9(.2) + (-1)(.2) + 7(.25) + 15(.3) + 15(.05) + 15(0) = Rs 5.0
4 -12(.2) + (-4)(.2) + 4(.25) + 12(.3) + 20(.05) + 20(0) = Rs 2.4
5 -15(.2) + (-7)(.2) + 1(.25) + 1(.25) + 9(.3) + 17(.05) + 25(0) = Rs 2.1

Thus, the highest expected gain is Rs 5.20 when 2 units are


produced.

So, 2 units should be produced.

Example 2.8
A retailer has to decide as to the optimum number of units
to be stocked of a certain item under the following conditions:

(a) Cost price in season is Rs 12.

(b) Selling price in season is Rs 18.

(c) Bargain price after the season is Rs 9.

(d) Cost of holding an item beyond the season is Re 1.

The probability distribution of demand based on past data is


as follows:

Units demand: 7 8 9 10 11

Probability : .20 .20 .25 .15 .20

Determine the optimum stock level on the expected monetary


value criterion.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
52 Research Methodology fo
n ~;
k ; k ‘ kf D r o’
k~ e

Notes Solution
__________________
Let m = no. of stock.
__________________
__________________ n = no. of demand.
__________________
For n ≥ m, total profit= 6m, as profit per unit = 18 – 12 = Rs 6
__________________
__________________
For m ≥ n, total profit= 18n – 12m + (m–n) (9 – 1)
= 10n – 4m
__________________
__________________ As the bargain price is Rs 9 and cost of holding is Re 1, so
__________________ net receipt = 9 - 1. Now, according to number of stock and
__________________ demand, the payoff Table is as follows:

Payoff Table

Demand
(n)
Stock 7 8 9 10 11
(m)

7 42 42 42 42 42

8 38 48 48 48 48

9 34 44 54 54 54

10 30 40 50 60 60

11 26 36 46 56 66

Expected Gain Table

Demand 7 8 9 10 11
Probability
.20 .20 .25 .15 .20
Stock

7 42(.20) + 42(.20) + 42(.25) + 42(.15) + 42(.20) = Rs 42

8 38(.20) + 48(.20) + 48(.25) + 48(.15) + 48(.20) = Rs 46

9 34(.20) + 44(.20) + 54(.25) + 54(.15) + 54(.20) = Rs 48

10 30(.20) + 40(.20) + 50(.25) + 60(.15) + 60(.20) = Rs 47.5

11 26(.20) + 36(.20) +46(.25) + 56(.15) + 66(.20) = Rs 45.5

From the above calculation it is clear that the expected gain


is maximum (Rs 48) when the number of stock is 9.

So, the optimum stock level is 9.


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 53
Decision-making under Uncertainty Notes
__________________
In this case the probabilities associated with the different
__________________
states of nature are unknown. Moreover, there is no previous
__________________
data or information available which could allot the
__________________
probability of the occurrence of states of nature. As such,
the decision-maker cannot calculate the expected payoff for __________________

the course of action. The decision-maker faces problems when __________________


he desires to introduce new plant or production. A number __________________
of decision criteria have been provided for decision-making __________________
under such a state of nature. The decision-maker may choose __________________
any one of the following: __________________

(i) Maximin, (ii) Maximax, (iii) Minimax regret, (iv) Hurwitz


criterion, (v) Criterion on rationality.

(i) Maximin: The course of action that maximizes the


minimum possible payoff is selected. The decision-
maker lists down the minimum outcome within each
course of action (or act) and then selects the strategy
with the maximum number.

This is also known as a pessimistic decision criterion as


it locates the strategy having the least loss.

Refer Example 2.2, the row minimums of the different


courses of action are:

a1 : 40 By this criterion the choice


will be a1 which has the
a2 : 36 maximum payoff amongst
all the minimums of four
a3 : 32 acts

a4 : 28

(ii) Maximax: In this case the course of action that


maximizes the maximum payoff is taken. The decision-
maker lists down the maximum payoff associated with
each course of action, then selects that alternative
having maximum number.

This may be called an optimistic decision criterion as


the decision-maker selects the alternative of highest
possible gain.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
54 Research Methodology fo
n ~;
k ; k ‘ kf D r o’
k~ e

Notes In the same example as mentioned above, the maximum


__________________ payoff of each course of action is:
__________________
a1 : 40 Since the course of action a4
__________________
has the maximum payoff,
__________________ a2 : 42 so the decision-maker will
__________________ select this alternative.
__________________
a3 : 44
__________________
__________________ a4 : 46
__________________
(iii) Minimax regret: The regrets (i.e., opportunity loss) for
__________________ each course of action are to be calculated with reference
to the payoff list of various alternative acts. Now obtain
the maximum regret and hence select the course of
action with minimum of the maximum regret values.

Refer the Loss Table 2.6 given in Example 2.2.

State of (20) (21) (22) (23) Maximum


nature S1 S2 S3 S4 Regrets
Acts

(20)a1 40 – 40 = 2 42 – 40 = 2 44 – 40 = 4 46 – 40 = 6 6

(21)a2 40 – 36 = 4 42 – 42 = 0 44 – 42 = 2 46 – 42 = 4 4

(22)a3 40 – 32 = 8 42 – 38 = 4 44 – 44 = 0 46 – 44 = 2 8

(23)a4 40 – 28 = 12 44 – 34 = 8 44 – 40 = 4 46 – 46 = 0 12

Thus the maximum regrets is in a2.

(iv) Hurwitz criterion (or criterion of realism): This


criterion makes a compromise between maximax and
maximin, i.e., and optimistic and pessimistic decision
criterion. At first, a coefficient of optimism a (0 £ a £ 1)
is selected, which is assumed to be degree of optimism.
Now, according to Hurwitz, select that alternative which
maximizes.

a (maximum payoff) + (1 – a) (minimum payoff)

Example 2.9
Choose the list product applying the Hurwitz method with
the coefficient of optimism a=.60 to the following data.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 55
Notes
States of Profit (in Rs) if the market is
nature __________________
Acts Good Fair Poor __________________
Product A 10,000 8,000 2,000 __________________

Product B 12,500 9,000 -1,000 __________________


__________________
Solution __________________

Maximum and minimum payoff are tabulated for each act. __________________

Strategy Max. value Max. in __________________


__________________
Product A 10,000 2,000
__________________
Product B 12,500 -1,000

As a = 0.6, so (1 – a) = 1 – 0.6 = 0.4.

We have for Product A : 10000 × 0.6 + 2000 × 0.4 = 6800


Product B : 12500 + 0.6 + (-1000) × 0.4 = 7100.

As the expected value of Product B is higher, so Product B


should be accepted.

(v) Criterion of rationality: This criterion is based on the


1
principle of equal n likelihood. The decision-maker first
3
calculates the average outcome for each course of action
and then selects the maximum number.
1
Average outcome is (0i1 + 0i2 + ..... + 0in)
n
Example 2.10
Events S1 S2 S3
Acts
A1 400 600 200
A2 650 800 -100
A3 400 500 60

Solution

Average for A1 : × 1(400 + 600 + 200) = 400

A2 : × 1(650 + 800 – 100) = 450

A3 : × 1(400 + 500 + 60 ) = 320

Here the choice is for A2.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
56 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Example 2.11


__________________
Apply (i) Maximin (ii) Maximax (iii) Minimax regret to the
__________________
following payoff matrix:
__________________
__________________ States of
nature S1 S2
__________________ Acts
__________________ A1 -6 2
A2 4 1
__________________ A3 7 -4
__________________
__________________ Solution
__________________ (i) Maximin: Minimum course of action

A1 -6 Now maximum of
minimum is 1, so A2
A2 1 is acceptable.

A3 -4

(ii) Maximax: Maximum of strategies

A1 2 Maximum of the
maximum is 7 for
A2 4 Act A3.

A3 7

(iii) Minimax Regret:

S1 S2 Maximum
regrets
A1 7–(–6) = 13 2–2 = 0 13
A2 7–4 = 3 2–1 = 0 3
A3 7–7 = 0 2–(–4) = 6 6

Minimum of the maximum regrets is 3 in A2.

Example 2.12
A pig breeder can either produce 20 or 30 pigs. The total
production of his competitors can be either 5,000 or 10,000
pigs. If they produce 5,000 pigs, his profit per pig is Rs 60; if
they produce 10,000 pigs, his profit per pig is only Rs 45.
Construct a payoff Table and also state what the pig breeder
should decide.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 57
Solution Notes
__________________
With the given data let us construct the payoff Table:
__________________
Payoff Table
__________________
State of 5,000 10,000
nature __________________
Acts __________________
A1 (20) 20 × 60 = 120 20 × 45 = 900
__________________
A2 (30) 30 × 60 = 1,800 30 × 45 = 1,350
__________________
Assuming the pig breeder to be conservative, we use maximin __________________
approach. As such, the minimums of act A1 is 900 and that of __________________
A2 is 1350, which is, of course, maximum. So the pig breeder
__________________
should accept the course of action A1.

Further, by maximax approach also 1800 is maximum which


also corresponds to A2.

Example 2.13
Given is the following payoff matrix:

Course of action
Do not Expand Expand
States of nature Probability expand 200 units 400 units
Rs Rs Rs Rs
High demand 0.4 2,500 3,500 5,000
Medium demand 0.4 2,500 3,500 2,500
Low demand 0.2 2,500 1,500 1000

What should the decision be if we use: (i) expected monetary


value criterion (ii) the maximax criterion (iii) maximax regret
criterion?

Solution
Let us form the payoff Table as follows:

Payoff Table

States of nature High Medium Low


probability demand demand demand
Course of 0.4 0.4 0.2
action

(A1) Do not expand 2,500 2,500 2,500


(A2) Expand 200 units 3,500 3,500 1,500
(A3) Expand 400 units 5,000 2,500 1,000
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
58 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (i) Expected monetary value:


__________________
forA1 = .4 × 2,500 + .4 × 2,500 – .2 × 2,500 = Rs 2,500
__________________
__________________ A2 = .4 × 3,500 + .4 × 3,500 + .2 × 1,500 = Rs 3,100
__________________
A3 = .4 × 5,000 + .4 × 2,500 + .2 × 1,000 = Rs 3,200
__________________
__________________ (ii) Row minimums of the different courses of action are:
__________________ A1 : 2,500 Maximum of these minimums is 2,500,
__________________ so the choice is A1, i.e., the decision-
__________________ maker.
__________________
A2 : 1,500 Under this criterion the decision-maker
would decide not to expand.

A3 : 1,000

(iii) Row maximums of the different courses of action are:

A1 : 2,500 The maximum of these maximums is


5,000, i.e., choice is A 3 . In other
words the decision-maker under this
criterion would decide to expand 400
A2 : 3,500 units.

A3 : 5,000

(iv) Let us find the regret matrix from the payoff matrix by
subtracting each of the values in a column from the
largest payoff value in the same column.

Regret Table of the Payoff Table

State of
nature High Medium Low Maximum
demand demand demand regrets
Course
of action

Do not expand (A1) 5,000 – 2,500 = 2,500 3,500 – 2,500 = 1,000 2,500 – 2,500 = 0 2,500

Expand 200 units (A2) 5,000 – 3,500 = 1,500 3,500 – 3,500 = 0 2,500 – 1,500 = 1,000 1,500

Expand 400 units (A3) 5,000 – 5,000 = 0 3,500 – 2,500 = 1,000 2,500 – 1,000 = 1,500 1,500

Minimum of maximum regrets is 1,500 that occurs in A2 and


A3. So the decision-maker must choose to expand 200 units
or expand 400 units.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 59
Expected Opportunity Loss (EOL) Notes
__________________
If P(S 1 ), P(S 2 ), ......., P(S n ) be the prior probabilities
__________________
corresponding to the respective states of nature S1, S2, ......,
__________________
Sn, then expected opportunity loss (EOL) of acts:
__________________
A1 = (M1 – P11)P(S1) +(M2 – P12)P(S2)+.............+(Mn – P1n)P(Sn) __________________
A2 = (M1 – P21)P(S1) +(M2 – P22)P(S2)+.............+(Mn – P2n)P(Sn) __________________
__________________
Example 2.14
__________________
Refer the Example 2.13. If 0.3, 0.4, 0.3 be the respective __________________
probabilities of three states of nature (i.e., high demand,
__________________
medium demand and low demand) then EOL of acts:

A1 = 2500 × .3 + 1000 × .4 + 0 × .3 = 1150


A2 = 1500 × .3 + 0 × .4 + 1000 × .3 = 750
A3 = 0 × .3 + 1000 × .4 + 1500 × .3 = 850

In case of Act A2, EOL is minimum.

Expected Value of Perfect Information (EVPI)


The expected value with perfect information is the expected
return in the long run, provided there is perfect information
before a decision is taken.

Calculation
Expected payoff with perfect information (EPPI) = (max. payoff
in 1st state of nature) × (prob. of 1st state of nature) + (max.
payoff in 2nd state) × (prob. of that state) × ..... up to last state.

Now EVPI = EPPI – maximum EMV.

Example 2.15
Payoffs of three acts A, B and C and states of nature X, Y
and Z are given below:
Payoff (in Rs)
Acts
A B C
States of nature

X -20 -50 200


Y 200 -100 -50
Z 400 600 300
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
60 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The probabilities of the states of nature are 0.3, 0.4 and 0.3
__________________
Calculate the EMV for the above data and select the best
__________________
act. Also find the EVPI.
__________________
__________________ Solution
__________________
Let us find the expected monetary value (EMV) of each act.
__________________
__________________ Act A = –20 × .3 + 200 × .4 + 400 × .3 = Rs 194
__________________ Act B = –50 × .3 – 100 × .4 + 600 × .3 = Rs 125
__________________
Act C = 200 × .3 – 50 × .4 + 300 × .3 = Rs 130
__________________
EMV of Act A is highest as seen in the table, so it should be
selected.

Payoff (Rs)

Max. for
States of Prob. A B C state of Max. payoff
nature nature × prob.

X .3 –20 –50 200 200 200 × .3 = 60


Y .4 200 –100 –50 200 200 × .4 = 80
Z .3 400 600 300 600 600 × .3=180

Total 320

EVPI = Expected payoff with perfect information (EPPI) –


EMV = 320 – 194 = Rs 126.

Example 2.16
Mohan Products Ltd. produces a drums which must be sold
within a month of it being produced, if the normal price of
Rs 100 per drum is to be obtained. Anything unsold in that
month is sold in a different market for Rs 20 per drum. The
variable cost is Rs 55 per drum.

During the last five years, monthly demand was recorded


and showed the following frequencies:

Monthly demand (No. of drums) 2000 3000 6000

Frequency 24 12 24
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 61
(i) Prepare an appropriate Payoff Table. Notes
__________________
(ii) Advise the production management on the number of
__________________
drums that should be produced next month.
__________________
Solution __________________
__________________
Gain (on sale) = 100 – 55 = 45, Loss (for unsold) = 55 – 20 = 35
__________________
(in Rs)
__________________
Payoff Table
__________________
Demand 2000 3000 6000 Expected value __________________

Prob. 24 12 24 __________________
— = 0.4 — = 0.2 — = 0.4
Stock 60 60 60

2,000 2,00 × 45 90,000 90,000 90,000 × .4 +


= 90,000 90,000 × .2 +

90,00 × .4 = 90,000

3,000 90,000 – 35,000 3,000 × 35 1,35,000 – 3,000 × 35 55,000 × .4 +


= 55,000 = 13,500 = 30,000 1,35,00 × .2 +
30,000 × .4 = 61,000

6,000 90,000 – 4,000 × 35 1,35,000 – 3,000 × 35 6,000 × 45 –50,000 × .4 +


= -50,000 = 30,000 = 2,70,000 30,000 × .2 +
2,70,000 × .4 = 94,000

Since the expected value is highest (Rs 94,000), in the next


month there should be 6,000 drums in stock.

Decision Tree Diagram


A decision problem may also be represented with the help
of a diagram in the form of a tree, for a clear and systematic
idea. In this case alternative courses of action, states of
nature, likely outcomes, etc., are diagrammatically or
graphically depicted as if they are the branches, sub-branches
of a horizontal tree. Hence, it is known as a tree diagram.

The tree consists of a network of nodes, probability estimates,


payoffs.

Standard Symbol
Nodes: There are two types of nodes (i) decision node
(symbolized as square) and (ii) chance node (marked as circle).

Alternative courses of action originate from the decision


node in main branches. Now, at the terminal point of decision
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
62 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes node, chance nodes exist from where chance nodes emanate
__________________ as sub-branches. Probability and outcomes are shown along
__________________ these sub-branches. Each sub-branch can form a branch again
__________________ so that we can build a tree-like structure representing all
__________________ possible outcomes.
__________________ Types of decision trees: (i) deterministic (ii) probabilistic.
__________________ These again can be divided into single stage and multi-stage
__________________ trees.
__________________
In a single stage decision tree we can find only one decision
__________________ (i.e., no chance event).
__________________
In a multi-stage decision tree, a chain of decisions is to be
made.

Acts Events Outcomes

E1
O1
decision
A1 E2 node
O2

E1
O1
A2 chance
E2 node
O2

Figure 2.5: Decision Tree Diagram

Example 2.17
There is 40% chance that a patient admitted to the hospital
is suffering from cancer. A doctor has to decide whether a
serious operation should be performed or not. If the patient
is suffering from cancer and the serious operation is
performed, the chance that he will recover is 70%, otherwise
it is 35%. On the other hand, if the patient is not suffering
from cancer and the serious operation is performed,
the chance that he will recover is 20%, otherwise it is
100%. Assume that recovery and death are the only possible
results.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 63
Solution Notes
__________________
Construct an appropriate decision tree. What decision should
__________________
the doctor take?
__________________
The decision tree has been constructed as follows: __________________
Probability of recovery on operation = 0.28+0.12 = 0.40 __________________
Probability of recovery for no operation = 0.14+0.60 = 0.74.
__________________
As 0.74 > 0.40, so operation should not be done for recovery.
__________________

recovery 0.7 __________________


(0.28) __________________
operate
death 0.3 __________________
(0.12)
cancer 0.4
recovery 0.35
(0.14)
not operate
death
(0.26)
recovery 0.2
(0.12)
operate
death 0.8
(0.48)
no cancer 0.6
recovery 1.0
(0.60)
not operate
death 0.0
(0.00)

Example 2.18
A manufacturing company has to select one of the two
products, A or B, for manufacturing. Product A requires
an investment of Rs 20,000 and product B Rs 40,000.
Market research survey shows high, medium and
low demands with corresponding probabilities and returns
from sales, in Rs 1000 for the two products, in the following
Table:

Market Probability Return from sales


demand A B A B
High 0.4 0.3 50 80
Medium 0.3 0.5 30 60
Low 0.3 0.2 10 50
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
64 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Construct an appropriate decision tree. What decision the


__________________ company should take?
__________________
Solution
__________________
__________________ Market A B
__________________ demand
x p px x p px
__________________ (’000) (’000)
__________________ High 50 0.4 20 80 0.3 24
__________________ Medium 30 0.3 9 60 0.5 30
__________________ Low 10 0.3 3 50 0.2 10
__________________ Total 32 64

Product Returns (Rs) Investment (Rs) Profit (Rs)


A 32,000 20,000 12,000

B 64,000 40,000 24,000

The company should take decision in favour of B since the


profit is high.

high demand (50)


(20)
0.4
product A medium demand (30)
(9)
0.3
low demand (80) (3)
0.3

high demand (80) (24)


0.3
product B medium demand (60) (30)
0.5
low demand (50) (10)
0.2

Example 2.19
Y.K. Sachdeva has two independent investments A and B
available to him; but he can undertake only one at a time
due to certain constraints. He can choose A first and then
stop, or if A is successful then take B or vice versa. The
probability of success of A is 0.6 while for B it is 0.4. Both
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 65
investments require an initial capital outlay of Rs 10,000 and Notes
both return nothing if the venture is unsuccessful. Successful __________________
completion of A will return Rs 20,000 (over cost) and __________________
successful completion of B will return Rs 24,000 (over cost). __________________
Draw decision tree and determine the best strategy. __________________
__________________
Solution
__________________
The appropriate decision tree is as follows: __________________
(Outcome Rs)
(–10,000) __________________
)
.4

Fail (.6)
il (

(–10,000) __________________
Fa

Suc B __________________
cess ACC
Success (.4)
A

D2 (24,000)
t

(.6)
ep
cc
A

Stop
Do Nothing (0)
D1
(0)
Stop
D3
A

cess (–10,000)
cc

Suc
ep

ACC Fail (.4)


t

A
B

(.4)
(20,000)
Fa

Success (.6)
il (
.6

(–10,000)
)

We find three decision points D 1, D 2, D 3 in the previous


decision tree diagram. For an analysis of tree we start
working backward.

Decision Outcome Probability Conditional Expected


point values values

D3 (i) Accept A success .6 20,000 12,000


failure .4 –10,000 –4,000
8,000
(ii) Stop 0
D2 (i) Accept B success .4 24,000 9,600
failure .6 –10,000 –6,000
3,600
0
D1 (i) Accept A success .6 20,000 + 3,600 14,160
failure .4 –10,000 4,000
10,160
(ii) Accept B success .4 24,000 + 8,000 12,800
failure .6 –10,000 –6,000
(iii) Do nothing 6,800
0
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
66 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes From the column of expected values see Table 2.7, we find
__________________ the best strategy is to accept A and, if it is successful, then
__________________ accept B.
__________________
__________________
Summary
__________________ This unit presented the fundamentals of the Bayesian
__________________ approach to decision-making under uncertainty and
__________________ certainty.
__________________ The unit first emphasized the need for information in
__________________ decision-making. The unit also discussed the concept
__________________ of probability and indicated the basic sources of
assigning probabilities. In the present days, the
probability theory has a vital role, as such, this
probability theory will be used more frequently in the
decision-making theory under uncertainty and risk. The
management of an industry or any concern is to take
decisions about the development or expansion of the
concern. No doubt the concern has to face many problems
from time to time. The management has to overcome or
solve the problems. This unit concluded with brief
descriptions of industry applications of Bayesian
decision analysis and related methods.

Review Questions
1. What is the ‘expected value’ approach to decision-
making?

2. What are the advantages of Bayesian analysis over other


methods?

3. Why is it necessary to estimate the value and cost of


information before conducting research?

4. Carry out a full decision analysis for M/s Excel Ltd, using
the following information:
Events Actions Expert Delay Delay
Now (Rs) 1 Year (Rs) 2 Years (Rs)
Economic conditions remain good (A) 7,00,000 5,60,000 4,10,000
Moderate downturn in economy (B) 5,20,000 4,00,000 3,00,000
Economic recession (C) –2,96,000 26,000 29,000
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 2 Defining Problem 67
Prior probabilities for the various events for the next Notes
twelve months are: __________________
__________________
(A) = 0.3
__________________
(B) = 0.4 __________________

(C) = 0.3 __________________


__________________
Carry out a pre-posterior analysis and find the expected
__________________
value of perfect information.
__________________
5. Describe the process of specifying the information __________________
needed. __________________

6. What do you understand by the term EVPI?

7. Suppose you are given the following Payoff Table:

S1 S2

A1 10 4

A2 2 5

(i) Assuming that the states of nature S 1 and S 2


exhaust the possible environmental conditions,
what probability would have to be associated with
the occurrence of S 1 if the expected payoffs of
alternative A1 and A2 are to be equal?

(ii) Suppose in the above Payoff Table, the payoff


corresponding to the A 2S 2 combination were 4
instead of 5, would the knowledge of the
probabilities assigned to S1 and S2 be required?
69

Unit 3 Notes
__________________
__________________
Relationship between __________________

Management and Research __________________


__________________
Problem __________________
__________________
__________________
__________________
__________________

Objectives
After reading this unit you will be able to:

y Understand the meaning of problem.

y Present the classification of problem.

y Explain the respective role of management and researcher.

Management and Research Problem


An old saying is "A problem well defined is half solved". This
is held true in the area of research. Hence, the research
activities start with defining the problem. Problem is
primarily a situation of dilemma of decision situation and
the simplest of decision situations can be characterized by
the following conditions:

1. There is an individual operating within some


environment to whom the problem can be attributed.

2. The individual has at least two courses of action that


can be pursued, and is able to make a choice from the
available courses of action.

3. There are at least two possible outcomes of that choice,


and one of these outcomes is preferred to the other.

4. There is a chance, but not an equal chance, that each


course of action will lead to the desired outcome.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
70 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Thus, a problem is the dilemma of choosing one alternative


__________________ among various alternatives for making a choice on courses
__________________ of action. Which course of action needs to be selected, gives
__________________
rise to the need of research.
__________________ One of the best ways to approach the research process is to
__________________ state the basic dilemma that arises the need for research
__________________ and than to develop other questions by progressively breaking
down the original question in more specific ones. This original
__________________
question is called the management problem which when is
__________________
broken in specific ones is called the research problem.
__________________
Management problem is the actual problem that the
__________________
management is facing, viz;

u Rising cost

u Sales decline

u Increasing LTO

u Market contraction and expansion decisions; etc.

Breaking down the management problem into specific one


requires an step wise analysis. First, the management
dilemma needs to be defined which is not a difficult task as
all modern organizations keep track of their performance
factors (Like: Sales, Profits, Employees Turnover; etc). For
arriving at final management dilemma the first stage is
exploration where a researcher needs to review published
resources and interview information gatekeepers. By
collecting and recording the exploratory information, the
researcher needs to word the dilemma or the correction of
the symptom in the question form, usually starting with, "How
can the organization…?"

Once the management dilemma is worded in the question


form then further exploratory research is taken to find out
the possible management actions that might be taken to solve
the management dilemma. This requires qualitative research
techniques to be used viz; Brain storming with the experts,
Interviews with information gatekeepers, secondary sources
study, etc. Finally, several research questions is to be framed
which could be used for further investigation by using both
qualitative as well as quantitative study. Each question is an
alternative action that management might take to solve the
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 3 Relationship between Management and Research Problem 71
management dilemma. Among them the most plausible Notes
action, or the one that offers the greatest gain using the fewest __________________
resources should be researched first. __________________
__________________

Discovering Defining __________________


Management Exploration Management __________________
__________________
__________________
__________________
Defining Research Exploration
__________________
__________________

Figure 3.1

Respective Roles
The Management's Role
The manager must move from the management dilemma to
management question to proceed with the research process.
Management problems are too numerous to list, but they can
be categorized in:

u Choice of purpose or objective: "What do we want to


achieve?"

u Generation and evaluation of solution: "How can we


achieve them?"

u Troubleshooting or control situation: "Why are we facing


the present situation?"

Researcher’s Role
Once the researcher has a clear statement of the management
question, they with the help of the manager must translate
them into research question. The research question should
be a fact - oriented, information - gathering question. Only
relevant alternatives should be considered. To have an idea
of relevant alternatives the insight of manager must be
taken. This will also help the researcher having idea of
manager's decision making environment.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
72 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The researchers task is to assist the manager in formulating


__________________ a research question that fits to resolve the management
__________________ dilemma. A Research question is the hypothesis of choice
__________________ that best states the objective of the research study. It is more
__________________ specific management question that needs to be answered and
this can be more than one. After exploration is complete the
__________________
research questions must be fine tuned.
__________________
__________________ After the research problem is formulated it should further
__________________ be broken down in Investigative question and Measurement
__________________
question. Investigative questions are questions the
researcher must answer to arrive at a Conclusion about the
__________________
research question. These questions reveal the specific piece
of information the manager feels to know to answer the
research question. Measurement questions are the questions
that the researcher asks to the respondents. They appear
on the questionnaire and are in line with the information
needed for the research question.

All these steps can be shown as:

Management Dilemma

Management Question

Research Question(s)

Investigative Question

Measurement Question

Management Decision

Figure 3.2
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 3 Relationship between Management and Research Problem 73
Summary Notes
__________________
This unit presented the classification of problems and
__________________
detailed the need of converting management's problem
__________________
into researchers problem.
__________________
The unit first stated and cleared what is a problem and __________________
than has gradually proceeded in explaining the role of __________________
management and researcher to convert managements
__________________
problem into precise detail on which further study can
__________________
be made feasible.
__________________
__________________
Review Questions
1. Define the role of management in the formulation of
research problem.

2. Elucidate the difference between management problem


and research problem.
75

Unit 4 Notes
__________________
__________________
Research Design __________________
__________________
__________________
__________________

Objectives __________________
__________________
After reading this unit you will be able to:
__________________
y Define research design, classify various research design, and
explain the differences between exploratory and conclusive. __________________

y Compare the basic research design: exploratory and descriptive.

y Explain research design and different types of research design.

y Describe the various factors to be considered in the selection of


an experimental design.

y Describe the procedure for reduction of error through design.

Research design or model indicates a plan of action to be


carried out in connection with a proposed research work. It
provides only a guideline for the researcher to enable him
to keep track of his actions and to know that he is moving in
the right direction in order to achieve his goal. The design
may be a specific presentation of the various steps in the
process of research. These steps include the selection of a
research problem, the presentation of a problem, the
formulation of the hypothesis, conceptual clarity,
methodology, survey of literature and documentation, data
collection, testing of the hypothesis, interpretation,
presentation and report writing. When Apple India Ltd.
launched its Apple juice with extraordinary caution no
comprehensive study had ever been done by Apple India Ltd.

In this chapter various types of research designs are


explained with suitable examples.

Introduction
The purpose of marketing research is to provide information
which will aid in management decision-making. A marketing
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
76 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes manager, for example, in a large consumer goods company


__________________ may want to collect information to assess whether or not to
__________________ launch a new product or to determine why the sales of a
__________________ particular product are declining. In collecting this
__________________ information, three major steps may be identified. First and
the foremost is establishing of research objectives. Then a
__________________
research plan needs to be formulated and implemented; and
__________________
finally data analyzed followed by a presentation of research
__________________
findings. A clear and precise definition should be made by
__________________ the management of the decision problem, this should be
__________________ expressed not only in terms of problem symptoms such as a
__________________ decline in market share but also with possible contributing
factors such as changes in competitors’ strategies or in
consumers’ tastes, as well as the actions the management
might take based on research findings. Otherwise, irrelevant
information may be collected.

Development of a research plan requires what data are to be


collected; what research techniques and instruments are to
be used; how a sample is to be selected; and how information
is to be collected from this sample. Sometimes the required
information may already be available in secondary sources
such as the government or trade reports, company records,
or sales force reports. This will not, however, have been
collected with the particular problem in mind. Consequently,
primary data collection may be required, in other words,
collect information specifically for the particular problem.

Where primary data are collected, observational or other


qualitative techniques, experimentation or survey research
may be conducted. Observations and other qualitative
techniques, such as projective techniques (word association,
sentence completion tasks and focus group interviewing) are
most appropriate in the initial stages of research, when little
is known about the problem. The onus of interpretation is,
however, placed on the researcher and consequently, such
techniques are open to criticisms of subjectivity.

Experimental techniques are also potentially applicable, but


they are rarely used except in-store experiments study, for
example, the impact of in-store promotions on sales. Test
marketing can also be viewed as a field experiment. Survey
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 77
research is the technique most commonly used in marketing Notes
research. A standard questionnaire can be administered to __________________
large samples and systematically analyzed using __________________
computerised techniques. __________________

In observational or qualitative research instruments such __________________

as coding scheme, recording sheets and other tests may need __________________
to be designed. The most common instrument used is a __________________
questionnaire. For unstructured interviews and focus __________________
groups, only an interview guide indicating the topics to be __________________
covered may be required. A crucial aspect of survey research __________________
is, however, the design of a carefully worded questionnaire
__________________
to elicit desired information from respondents. Attention to
question form and sequencing is also essential in order to
avoid biased responses.

Three principle methods of data collection may be


considered–telephone, mail or personal interviewing.
Telephone interviewing is quick and can be conducted from
a central location where interviewers are controlled by a
supervisor. However, only with this method the persons
having telephones can be interviewed and a limited number
of questions can be asked. Mail questionnaires are the
cheapest method of survey administration, but suffer from
low response rates and also assumes that the respondent
clearly understands and can respond to questions. Personal
interviewing is the most flexible method since the
interviewer can select the sample by judgement or
convenience sampling and is able to explain questions to the
respondent. It is, however, an expensive method of data
collection and susceptible to interviewer bias. In
implementation of the research plan, this is where major
sources of data inaccuracy and unreliability often arise. In
the case of surveys, for example, respondents may bias
findings by refusing to cooperate, by providing inaccurate
answers, for example on income or by giving socially
undesirable responses. Interviewers may also bias results
by encouraging a specific response, by inaccurate recording
of responses or in extreme cases, by falsifying responses.

Current developments in telecommunications and computer


technology are rapidly changing data collection procedures
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
78 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes and improving their efficiency. The results can also be


__________________ analyzed and updated with each successive response, thus,
__________________ considerably reducing research time and costs. Data analysis
__________________ includes tabulating, classifying and interpreting the
__________________
information collected. The complexity of the analysis
depends to a large extent on management needs. In many
__________________
cases tabulation or cross-tabulation of results with averages
__________________
and other summary statistics may suffice. In other cases,
__________________ more sophisticated multivariate techniques such as factor
__________________ or cluster analysis or multidimensional scaling may be
__________________ required, if more complex interactions in the data are to be
__________________ examined. Finally the presentation of research findings may
be verbal and/or written. In either case, the main focus should
be on clear presentation of key research findings and their
implications on the decisions to be made by the management.
As it is obvious, the most vital element in this process is
research plan or the research framework/design. For the
research to be valid and reliable it is imperative that proper
thought should go into the design of research, otherwise it
may lead to GIGO (Garbage In, Garbage Out) or mere data
collection exercise with no information to the management
for its problem-solving exercise.

A marketing research design may be described as a series of


advance decisions that, when taken together, comprises a
master plan or model for the conduct of the investigations.
It is the blueprint that is followed in completing a study. An
appropriate design requires careful consideration of the
problem and objectives to be met in relation to the time and
resources available for the study. The design or plan of a
research investigation is best put in writing and it ordinarily
covers the following aspects:

(a) Objectives of the research including a statement of


general objectives and other operational objectives in
as specific form as possible.

(b) The relation of these objectives with the problem at


hand.

(c) The form in which the results will be obtained and how
they may be used.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 79
(d) The methods to be used in attaining each of the Notes
objectives of the research. __________________
__________________
(e) A time schedule for the entire research, including
__________________
speculations, deadlines for specific phases of the
research. __________________
__________________
(f) Personnel and administrative set-up, with duties
__________________
specified for each person working on the research
__________________
project.
__________________

Types of Research Design __________________


__________________
On the basis of objectives of the marketing research, the
research can be classified into: (i) Exploratory research (ii)
Conclusive research. On the basis of character of data MR is
often classified into: (A) Quantitative research (B)
Qualitative research. The research design for exploratory
research is best characterized by its lack of structure and
flexibility. It is generally used for the development of
hypothesis regarding potential problems and opportunities.
Exploratory research is further subdivided into (a) Search
of secondary data (b) Case study (c) Survey of experts.
Conclusive research which is used to provide information
for the evaluation of alternative courses of action can be sub-
classified into (a) Descriptive research (b) Causal or
experimental research.

Descriptive market research is used to describe marketing


phenomena while trying to determine the association among
variables. It also seeks to predict future marketing
phenomena.

In cross sectional design, which is typically used in


descriptive research projects, a sample of population
elements is taken at one point of time. In descriptive
research, we can make use of both case study and statistical
study. Causal or experimental research seeks to unearth
cause and affect relationships. A good causal research design
seeks to minimize the interference of external variables
while studying the relationship between variables. There is,
however, a third type of research, also termed as performance
monitoring research, that is an important part of marketing
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
80 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes research activity of a firm. It typically requires a longitudinal


__________________ research design (panel design), in which a fixed sample of
__________________ population elements is measured repeatedly. In the
__________________ traditional panel, the same variables are measured every
__________________ time, as opposed to an omnibus panel where different
__________________
variables are measured each time. In both types of panels,
the sample of respondents remains fixed.
__________________
__________________
Problem
__________________
__________________
Exploratory Research
__________________

Hypothesis New Relationships

Descriptive Research/Causal or Experimental Research

Performance Monitoring Research

Figure 4.1: Research Design in Marketing Research

Exploratory Research Design


Flexibility and creativity characterize exploratory research
study since the aim of its investigation is to find new
hypotheses. Hence, formal design is rare in this type of
research and much is left to the creative imagination of the
market researcher. He could conduct a study of secondary
sources of information or get views from the experts or
resort to case study analysis. Market researchers exercise
their individual initiative in spotting and following leads in
an organized way.

(A) Exploratory research using secondary sources of information


Many companies who regularly conduct market research
studies have maintained a record of research findings over
the years, the access to which is both quick and economical
for a market researcher. Looking for hypotheses, a market
researcher could consult basic research findings reported
in professional journals. Other fruitful sources of secondary
data are public libraries, newspapers, government reports,
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 81
trade journals, general books, etc. Computerization of Notes
database and quick search systems have made this type of __________________
research more organized, exhaustive and quick. At the same __________________
time there comes the problem of information overload and __________________
market research is the only way of shifting the relevant __________________
information by use of creative solutions to get insights on
__________________
which management can act.
__________________

(B) Exploratory research using expert opinions/individuals with ideas __________________


__________________
Experts can give valuable insights into a marketing problem
__________________
because of their experience with the problem at hand, ability
__________________
to act as an observer and an ability to express ideas unlike
any other individual. Despite the desire to find individuals
with ideas, it is important not to concentrate the
investigation only among the better educated or more
articulate persons. A practical solution is to carry out
exploratory research using a mix of cooperative, imaginative
persons as well as randomly selected persons. Respondents
should be given full freedom to respond to get new ideas.
Various techniques called qualitative research techniques
have been developed to unearth consumers’ beliefs, values,
attitudes, feelings that influence consumer behaviour. These
qualitative techniques can be further divided into:

(a) Depth interviews: These are one to one interviews


because most people do not have clear ideas why they
make particular purchase decisions. Individuals
normally do not report decisions. They normally do not
report socially unacceptable motives. So market
researcher in-depth interviews attempt to influence
respondents to talk freely about their subject of interest.
Market researcher leads and probes the topic to get
additional information. As such respondents are not
approached with a fixed list of questions but the
researcher has an outline in mind. No two interviews
will proceed the same way as, for example, in a
structured questionnaire interview. The market
researcher has the flexibility to probe beyond superficial
reasons, this is its strong as well as negative point in
the sense that it makes it difficult to compare results.
There is total reliance on the skill of the market
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
82 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes researcher which makes it expensive to hire such


__________________ knowledgeable persons. Depth interviews were
__________________ discredited after being in limelight in Ditcher’s time
__________________ but have been lately beginning to make a comeback in
__________________ understanding the consumer better for better consumer
service.
__________________
__________________ Direct questions to consumers about their attitudes or
__________________ motives seldom elicit useful information. Direct
__________________ questions do not measure the relative importance of the
__________________
various types of reasons, and many individuals will not
respond to what might be considered socially
__________________
unacceptable.

Instead of approaching respondents with a fixed list of


questions, the interviewer, in depth interviews,
attempts to influence respondents to talk freely about
the subject of interest. This is intended to put
respondents at ease so they express any ideas they have
on the subject. If some idea of interest is passed over
too quickly, the interviewer may seek additional
information by “probing”.

The advantages of depth interviews:

u They bring out information that would not be


obtained in a normal interview.

u The interviewer has a great deal of flexibility and


can use his ingenuity to stimulate respondents to
reveal more of their attitudes and motives.

The disadvantages of depth interviews:

u No two interviewers will proceed in exactly the


same way, thus it is difficult to compare results.

u This creates difficulty in securing respondents


cooperation and increases costs.

u The depth interview is the difficulty and cost of


interpretation.

(b) Focus group interviews: When the management wants


to get a first person feel of consumer reactions on its
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 83
marketing mix variables, a market researcher can Notes
convert a focus group with him/her acting as a __________________
moderator. The discussion is guided from the generic __________________
product category to the product specific category to the __________________
specific brand under consideration. The interviews are __________________
usually recorded so that the market researcher may not
__________________
lose information while keeping the discussion on the
__________________
track. Reports based on focus groups typically may
__________________
contain direct quotations from the interview session.
These can be highly useful and stimulating to marketing __________________

people. The guiding concept in selecting samples for __________________


focus groups is never to use only one focus group but to __________________
keep running groups until the researcher is getting no
more ideas. Individuals for a focus group are selected
on quota basis. Again the disadvantage is that a skilled
modulator is crucial to the success of focus groups. Also
critics of focus groups argue that the results obtained
are little more than creative ideas of the researcher.

Focus groups too, like in-depth interviews, have their


advantages like:

u Bringing out the inner motivations of the


consumers.

u Letting the researcher have a first hand


experience of the consumer reactions.

u Generating new ideas.

Focus groups have their disadvantages too:

u Excessive reliance on the skills of the moderator.

u Wrong combination of consumers in focus groups,


resulting in no tangible output.

u Using qualitative discussions quantitatively


without checking them out.

Focus groups are, however, being used increasingly by


researchers. Usually more than one focus group is used
to get a clear and unbiased understanding of the target
market behaviours and attitudes.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
84 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (c) Projective techniques: In projective techniques


__________________ interviewing, the respondents are asked to project
__________________ themselves into a particular situation. The consumers
__________________ then express different images they may have of
__________________ competitive brands of a given product. A consumer may
be asked to match a brand and a person’s image or asked
__________________
for role reversal. It is another approach to getting
__________________
respondents to reveal more than their surface feelings.
__________________
For example, consumers might be asked such questions
__________________ as:
__________________
If brand A were a motorbike, what would it be?
__________________
If brand B were a TV, which one might it be?

The responses depend on the personality of a consumer.


Every person has certain individual characteristics
called basic personality. This basic personality is very
stable and very difficult to change. These personality
traits are present due to an individual’s biological make-
up. A marketer should remember that it is not possible
to predict an individual’s purchase behaviour based on
his personality only. Nevertheless, by understanding the
product or segment the market aims at the target group
with certain major personality traits.

(d) Exploratory research using case analysis: Although


this method has long since attracted skeptics ire because
of its non-scientific nature, it is making a comeback. It
involves the comprehensive study of one of a few specific
situations and particularly the study of complex
situations in which interrelations of several individuals
are important. A situation is called a ‘case’.

The design of a research project is the plan that ties the


data to the conclusions that will be drawn. This method
is based on finding the differences and the similarities
so as to find the answers. Case research is not a
quantitative exercise but is used to enhance
understanding. It requires a rare combination of
judgement to select cases and insight in interpreting
them. Its strong points are holistic orientation, real life
situation, i.e., in case study method inferences are made
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 85
on the entire picture and not some of its parts; all the Notes
people actually concerned with the problem are __________________
contacted and data is obtained informally as well. __________________
__________________
The disadvantages are mainly owing to its time
consuming procedure; heavy reliance on investigators __________________

and the absence of formal information gathering __________________


techniques resulting in very generalized and random __________________
information which might not be useful at all. __________________
__________________
Conclusive Research
__________________
When a marketing manager has to select one course of action __________________
among a number of alternatives, conclusive research
provides him information that helps him to evaluate various
alternatives and select among them a course of action. This
type of research provides a rational basis for his decisions.
The alternatives may be well or vaguely defined. Conclusive
research design is characterized by formal research
procedures. The research objectives are accurately defined
and so are information needs. Conclusive research studies
can be classified either as descriptive or experimental.

(A) Descriptive research


The research objectives in this type of research are generally
describing the characteristics of consumer segment, viz.,
demographic, socio-economic, geographic, psychographic and
benefits sought. Descriptive studies can also portray buyer
perceptions of brands; audience profiles for media types, viz.,
TV, radio, newspaper, journals and magazines, etc. They can
also portray buying power of consumers, availability of
distributors, product consumption patterns, price sensitivity
of consumers, market share, etc. These are just a few
representative studies out of numerous studies that come
under descriptive research in marketing. Despite the
emphasis on description, it should not be concluded that the
studies should be simply fact-gathering expeditions. They
can also be used to make predictions about the occurrence
of a marketing phenomenon. The data regarding the presence
of an association among variables can only be used for
productive purposes but statements regarding cause and
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
86 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes effect relationships are not possible with descriptive


__________________ research (As may be possible in experimental or causal
__________________ research).
__________________
The purpose and nature of descriptive research is quite
__________________ different from that of exploratory research. Many descriptive
__________________ studies are made with only hazy objectives and with
__________________ inadequate planning. Much of the data collected in such
__________________ studies turns out to be useless. Descriptive studies of this
__________________ type are actually more of exploratory type. Effective
__________________
descriptive research is marked by a clear statement of the
decision problem, specific research objectives and detailed
__________________
information needs. The research design should be fairly
structured. Since the purpose is to provide information
regarding specific questions or hypothesis, the research must
be designed to ensure accuracy of the findings. Since
descriptive studies may cost huge amount of money to carry
out, there is then this necessity of its formal design.
Descriptive research often makes use of survey research
design which consists of a cross-sectional research design,
that is, collecting data on few factors from a number of cases
at one point of time. This is the most popular type of research
design and is useful in describing the characteristics of
consumers and determining the frequency of marketing
phenomenon. As already mentioned it is often expensive and
requires skillful and competent market researchers to
conduct it effectively. It is also termed as statistical method
in contrast to case method which focuses on many factors of
few cases. This method ceases to focus on individual cases
and focuses instead on classes, averages, percentages,
measures of dispersion and more sophisticated statistical
procedures. For example, cluster analysis can be used to
group customers into different classes on the basis of few
customer attributes or characteristics. Factor analysis can
be used to group attributes into few factors which are
important from consumers’ point of view vis-à-vis a product,
etc. In fact descriptive research presupposes that a sound
causal model of marketing system exists in the mind of the
decision-maker in contrast to exploratory research which
seeks to generate hypotheses. A market researcher has to
have a tentative hypothesis for carrying out descriptive
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 87
research for the sole reason that he knows what data to Notes
collect from his respondents in the survey. Descriptive __________________
research designs can also use one or more of the following __________________
sources of data: __________________

(a) Interrogation of respondents __________________


__________________
(b) Secondary data
__________________
(c) Simulation. __________________
__________________
The analysis techniques used in descriptive research are
__________________
those specifically for mass data. Each individual item tends
to lose its identity. This is both an advantage and a __________________

disadvantage. The advantage lies in the objectivity with


which the analysis can be made. Averages or variance can be
computed and compared; two independent researchers will
arrive at similar results which is not the case with the case
method. But the disadvantage lies in its mobility to prove
cause and effect relationships which is the domain of
experimental (causal) research. Even the direction of causal
effect may not be visible through statistical study, e.g. when
advertising and sales co-vary, it is often not clear whether
advertising causes sales or sales cause the expenditure of
more advertising effort because of greater apparent potential
sales results. Simulation of marketing phenomenon consists
in an incomplete representation of the marketing system or
some aspect of this system. It is relatively a new source of
data which is largely computer-oriented. Simulation can be
used to gain insight into the dynamics of the marketing
system by manipulating the independent variables
(marketing mix and situational variables) and observing their
influence on the dependent variables. A marketing
simulation requires data inputs regarding the characteristics
of the phenomenon to be represented and the relationships
present. Simulations should neither over-simplify nor over-
complicate. The limitations are its validity and time/cost of
updating the model as conditions change. Simulation models
can be classified on the basis of the purpose they serve, viz.,
predictive and descriptive.

Very often descriptive research may take the route of


analyzing already existing data on respondents, e.g., census
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
88 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes data. Secondary source of data is much cheaper to access


__________________ than carrying out a full-fledged survey and hence a great
__________________ temptation for a market research. But the validity and
__________________ appropriateness of such a secondary data descriptive
__________________ research should always be borne in mind before embarking
on one.
__________________
__________________ Example 4.1
__________________
The exploratory research on the drinking habit of milk
__________________
yielded a set of three hypotheses to be tested. The
__________________
information requirements were met by collecting data from
__________________
over 100 respondents belonging to the upper-income group.
A statistical frequency analysis and cross-classification
analysis revealed the following:

1. 64 per cent of the respondents preferred milk for


themselves or their children as a breakfast drink. A mere
32 per cent respondents responded positively towards
tea/coffee as breakfast drink. Rest of them opted for juice
in the morning.

2. 76 per cent of the respondents were positive about the


taste of the flavoured milk with the teenagers liking it
much more.

3. 40 per cent elders preferred tea/coffee in the evening,


whereas 35 per cent of the teenagers preferred milk as
a breakfast drink.

4. 80 per cent of the respondents wanted to give their


children milk during the day when they return from
school.

5. 62 per cent of the respondents considered it a status


symbol to offer tea/coffee to their guests or at social
gatherings.

(B) Experimental or causal research


Although, it is the nature of marketing decision-making that
all the conditions allowing the most accurate causal
statements are not usually present but in these
circumstances, causal inference will still be made by
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 89
marketing managers. Because in doing so they would want Notes
to be able to make causal statements about the effects of their __________________
actions. “The new advertising campaign we developed has __________________
resulted in p% increase in sales” or the “sales discount __________________
strategy we followed has resulted in p% increase in sales”,
__________________
etc. In both of these examples, marketing managers are
__________________
making a causal statement.
__________________
However, the scientific concept of causality is complex and __________________
differs substantially from the one held by the common person
__________________
on the street. The common sense view holds that a single
__________________
event (the ‘cause’) always results in another event (the
‘effect’) occurring. In science, we recognize that an event has __________________
a number of determining conditions or causes which act
together to make the effect probable. Note that in the
common sense notion of causality, the effect always follows
the cause. This is deterministic causation in contrast to
scientific notion which specifies the effect only as being
probable. This is termed as probabilistic causation. The
scientific notion holds that we can only infer causality and
never really prove it. That is, the chance of an incorrect
inference is always thought to exist. The world of marketing
fits the scientific view of causality. Marketing effects are
probabilistically caused by multiple factors and we can only
infer a causal relationship. The conditions under which we
can make causal inferences are:

(a) Time and order of occurrence of variables.

(b) Concomitant variation.

(c) Elimination of other possible causal factors.

The fundamental research tool used to identify causal


relationships is the experiment. The objective of an
experiment is to measure the effect of explanatory variables
or independent variables on a dependent variable, while
controlling for other variables that might confuse one’s ability
to make causal inferences. Such research leads to questions
like:

“Is a given print advertisement more effective in colour than


black and white?”

“Which of the several promotional techniques is most


effective in selling a particular product?”
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
90 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes “Can we increase the sales of a product by obtaining


__________________ additional shelf space?”
__________________
As Cox and Enis quote:
__________________
__________________ “The usefulness of experimental design in marketing extends
__________________ across the functional areas of promotion, distribution, pricing
and product policies. Whenever marketing management is
__________________
interested in measuring the effects of alternative courses of
__________________
action, experimentation may be a practical means of reducing
__________________
the risk involved in deciding among the alternatives.”
__________________
__________________
An experiment is executed when one or more independent
variables are consciously manipulated or controlled by the
person running the experiment and their effect on the
dependent variable or variables is measured. In surveys
there is no manipulation of independent variables by the
researchers. This is a fundamental difference between the
experimental and non-experimental research.

Treatments are independent variables that are manipulated


or whose effects are measured. Dependent variables are the
measures taken on test units. Test units are the entities to
whom the treatments are presented and whose response to
the treatments is measured. It is common in marketing for
both people and physical entities, such as stores or
geographic areas, to be used by as test units. For example,
people may be asked to try a product and then have their
attitudes towards it measured. Here people are test units,
product type is the independent variable and attitude is the
dependent variable.

Two concepts of validity are important in experimentation,


internal and external. Internal validity is concerned by with
the question of whether the observed effects on the test units
could have been caused variables other than the treatment.
Without internal validity the experiment is confounded.
External validity is concerned with the ‘generalability’ of
experimental results. A researcher, obviously, would like
an experimental design to be strong in both kinds of validity.
Unfortunately, it is often necessary to trade off one type of
validity for another. There are three true experimental
designs in which a researcher is able to eliminate all
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 91
extraneous variables as competitive hypotheses to the Notes
treatment: __________________
__________________
(1) The Pre-test—Post-test control group design:
__________________
Experimental group R 01 X1 02 __________________

Control group R 03 03 __________________


__________________
(2) The solomon four group design:
__________________
Experimental group 1 R 01 X 02 __________________
__________________
Control group 1 R 03 04
__________________
Experimental group 2 R X 05

Control group 2 R 06

(3) Post-test only control group design:

Experimental group R X 01

Control group R 02

Symbols used:

X Represents exposure of a test group to an experimental


treatment, the effects of which are to be determined.

0 Refers to processes of observation or measurement of


the dependent variables on the test units.

R Indicates that individuals have been assigned at random


to separate treatment groups or that groups themselves
have been allocated at random to separate treatments.

Movement from left to right indicates movement through


time. All symbols in any one row refer to a specific treatment
group. Symbols that are vertical to one another refer to
activities or events that occur simultaneously.

e.g. 0 1 X 1 0 2 indicates that one group received or


measurement of the dependent variable both prior to (0 1)
and after (02) the presentation of treatment (X1). Further the
symbols:
R X1 01
R X2 01
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
92 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes indicate that two groups of subjects were randomly assigned


__________________ to two different treatment groups at the same time. Further,
__________________ the groups received different experimental treatments and
__________________ the dependent variables were measured in the two groups
__________________ at the same time. These purely experimental research
designs control extraneous variables like history,
__________________
maturation, interactive testing effect, instrumentation,
__________________
statistical regression, selection bias and test unit mortality.
__________________
Table 4.1: Types of Experimental Designs
__________________
Type of Experimental Description Situation in which it is used
Design
__________________
1. "After only" design Difference between the experimental group's Measurement of advertising recall, day after
__________________ response before and after the experiment recall, etc. or measurement of increase in sales
after a sales promotion scheme.

2. "Before after" design The experimental group's reaction is first A consumer entering a departmental store,
measured before and after the experimental having various brands of shirts, may be asked
variable is introduced to them to check out the his brand preference. The salesman at the store
difference in behaviour. The disadvantage is would be instructed to influence the choice
that some uncontrollable variable might be towards a particular brand. The brand purchased
responsible for the change in opinion or by the consumer finally would indicate whether
behaviour. the sales people have been successful in
influencing the brand choices of the consumer.
A consumer's intention to purchase earlier is
compared with any changes at the time of final
purchase on account of external stimuli like an
advertisement or influence of the dealer, etc.

3. Before after with control The design consists of having a control group In the experiment described earlier, there could
group which is not subjected to the variable, versus be consumers to whom the salesman does not
the experimental group which is subjected to deliberately push a particular brand. This would
the variable. The difference of their differences constitute the control group. The differences
would give an idea of the extent of uncontrollable between the purchase behaviour of the control
variablespresent. group and that of the experimental group would
indicate whether uncontrollable variables had
caused the change in behaviour.

4. Four group six study To remove the bias which exists on account of The experiments of four groups six study design
design the 'before' measurement four groups are made, are difficult to set up and difficult to interpret and
two as in the before after with control group when have little practical application although they
the before measurement is not made – one are ideal.
control group and the other experimental.
Theoretically, it will be the solution of four
unknown quantities to find their differences.

5. After only with control In this case, no 'before' measurement The experimental group may be sent certain
group is made but only the after measurement is made privileged discount coupons for a sale, whereas
on the experimental group. The difference the control group is not, the results of the sale
between the two would indicate the effect of the would indicate whether the discount coupons
experimental variable. are indeed useful in increasing the sales of the
product.

6. The latin square design If the effect of a single variable is to be studied Suppose the owner of franchised outlets, say
over different time periods and different Bata, would like to know whether there is a
geographic regions a design which takes into difference in sales occurring on account of
account possible combinations is made and differences in three types of display used by
the differences arising on account of either the three similar Bata stores, situated at different
geographic region or the time period are studied. locations in Mumbai, for a period of three months.
The design would be:

Month ® 1 2 3

Bata store location

Crawford market A C B

Linking road C B A

Andheri B A C

where A, B, C are different sales displays.

Latin square designs are extensively used to


establish cause-effect relationships in
marketing situations.

Steps in planning an experimental design:

1. Selection of the problem: Every problem cannot be


studied through experimental method. One example of
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 93
major conditions being capacity to manipulate the Notes
independent variable the effect of which is decided to __________________
be studied. __________________
__________________
2. Proper description of the selected problem: After
selecting the problem, it must be put in proper language, __________________

i.e., the hypothesis must be stated in clear and conceptual __________________


terms. The variables that affect the phenomenon must __________________
be known and conceptualized. __________________
__________________
3. Selecting the setting: The background in which the
experiment relating to phenomenon is to be carried out __________________

is termed as setting. In case of laboratory experiment it __________________


is created artificially and the experimenter decides how
it can be done. In case of a field experiment, natural
setting has to be located where the experiment can be
made.

4. Pilot study: In planning an experiment, a pilot study


may be necessary so that the researcher is brought face
to face with realities and many problems that he had
not thought of. This also will enable him to know more
precisely the various causative factors involved, the
nature and working of the institution, the extent of
cooperation or resistance that he is expected to meet.

5. Research design: The most vital past of the research


is research design as it lays down the manner in which
the researcher will manipulate the situation in order
to study the desired effect. This in itself leads to problem
or control over the phenomenon.

Extensions of True Experimental Designs


There are four designs that permit the researcher to study
more than one level of an independent variable.

Completely Randomized Design


This is the simplest of the designs in which the treatments
are applied to the experimental units by a chance process.
The researcher is making an assumption that there are no
differences in the test units and as a result, all the test units
are treated alike and are randomly assigned to test groups.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
94 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes For instance, to find out the most appropriate price for a
__________________ new product, three different price levels–high, medium and
__________________ low can be tested. The different price levels can be tested in
__________________ three different stores. The analysis would reveal the average
__________________ amount of the product sold at each price level and disclose
whether there is any significant difference in sales between
__________________
the three price levels.
__________________
__________________ The main disadvantage of this design is that it does not take
__________________ into consideration the influence of possible extraneous
__________________
factors. However, this design has the advantages of
simplicity and inexpensiveness.
__________________

Randomized Block Design


This is an extension of the completely randomized design.
In this design, a single extraneous factor is isolated by
blocking out its effects. In the price illustration cited before,
the type of store was not given consideration, but it could be
an important influence on the sales of the product. Thus,
the randomized block design enables the researcher to
consider the effect of one extraneous factor and provides a
clear idea of the treatment’s effect on sales. In our
illustration, we might take six stores of each type like
department stores, speciality stores or grocery stores and
assign them to three price levels. The results will be the
same as that provided by the completely randomized design
but it gives us the effect of one extraneous factor.

Latin Square Design


The Latin square design considers the effect of two
extraneous factors. As the name indicates, the layout of the
design is a square. In this design, the treatments on each
subject are tested one at a time in a sequence and the
treatment is rotated so that one extraneous factor will offset
the other. For instance, if a marketer wants to test the effect
of four window displays in four stores, he may do so by
rotating the displays between the stores for two weeks in
each store with a week’s gap between each test. As the test
would form a square, there would be as many tests as
treatments. With four stores and four displays, we have a
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 95
4 × 4 design. This is a complex and expensive design, but it Notes
is more efficient than the previous designs. __________________
__________________
Factorial Design
__________________
In the Latin square design, only one variable is tested, __________________
whereas in the Factorial design two or more variables can __________________
be determined. For instance, a marketer wants to test the __________________
effect of TV advertising and price reduction on the sales of a __________________
product. Suppose, he considers three price reduction and
__________________
four levels of TV advertising, then it would be 3 × 4 design.
__________________
Hence, 12 markets have to be chosen to test the combinations
__________________
keeping the time period of the test constant. Factorial
designs can be constructed for more than two variables also.

Selecting an Experimental Design


The researcher has to consider three important factors in
the selection of an experimental design. They are:

1. Time: The researcher may be required to provide


results at a speed, as marketing decisions are required
to be fast due to competitive conditions. Hence, the
researchers have to plan for designs which take short
periods to complete.

2. Cost: The cost of executing the research has to


be weighed against the possible outcome and its
value.

3. Secrecy: The major limitation of experimentation is that


it might reveal things to the competitors. Hence, if
secrecy is needed, simulated laboratory experiments are
advisable.

Limitations of Experimentation
The manager should recognize the following limitations of
experimentation:

1. It is not always possible to control the effects


of extraneous variables. Differential effects
among treatment groups can easily occur in field
experiments.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
96 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 2. In field experiments, lack of cooperation from the


__________________ wholesalers and retailers can limit the experimental
__________________ activity.
__________________
3. Lack of knowledge about experimental procedures on
__________________ the part of marketing personnel may limit the usage of
__________________ experimentation and in addition, it may lead to
__________________ experimental conclusions being discarded as not
__________________ meaningful.
__________________
4. Experiments can be costly and time consuming.
__________________
__________________
5. In using people as test units, care must be taken that
the experimenter does not say and do things that bias
test unit responses.

Difficulties in Experimental Designs

1. Problems in experimental setting: Generally it is not


easy to determine the conditions under which
experiment should be setup. In case of scientific
experiments laboratory conditions may be established
but this may not be possible in case of sociological/
psychological experiments. In case of conduct of
experiment in a natural setting, problem arises in
varying various treatments and other conditions.

2. Problems in getting cooperation: In case of business


and social research, to obtain cooperation from people
who form the subject of experimentation is not easy.
Human subjects at times work according to their free
will.

3. Difficulties in establishing control: Control at times


in an experimental situation more so in case of complex
business and socio-economic research is lost since it is
very difficult to get complete knowledge of various
factors influencing the experiments.

4. Problem of consciousness: In case of business


experimental design, experimental subject is rather
fluid and possesses a consciousness which limits the
degree of experimentation.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 97
Designing Consumer Research Notes
__________________
The purpose of the consumer study is determined by research
__________________
design. When descriptive information is required, a
__________________
quantitative study is useful. If new ideas, e.g., positioning of
a new brand is needed, a qualitative study is likely to be __________________

conducted. The consumer research design can be classified __________________


into two categories: __________________
__________________
A–Qualitative research design
__________________
B–Quantitative research design __________________

u Qualitative research design: The purpose of the study __________________


and the type of data needed are considered while
selecting the appropriate research design for a
qualitative study. The data collection techniques
include:

(i) Depth interviews: Lengthy non-structured


interview between a respondent and a highly
experienced interviewer, who minimizes his or her
own participation in the interview.

(ii) Focus groups: Which consist of six to eight


respondents who meet with a moderator or
analyst for a group discussion focussed on a
particular product as product category.

(iii) Projective techniques: Consisting of a variety of


disguised tests that contain ambiguous stimuli,
such as incomplete sentences, untitled pictures
or cartoons, word association tests, etc., which are
designed to tap the underlying motives
of individuals despite their unconscious
rationalizations or efforts at a conscious
concealment.

u Quantitative research design: The design of a


quantitative research study includes the method for
collecting the data, the sample design and construction
of the data collection instrument.

(i) Data collection methods: There are three basic


ways to collect primary data in quantitative
research.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
98 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (a) Observational research–by observing


__________________ consumer behaviour.
__________________
(b) Experimentation–by designing experiments
__________________
to identify cause and effect of the relative
__________________ sales appeal of many types of variables.
__________________
(c) Surveys–by questioning people in person, by
__________________
mail or by telephone about their purchase
__________________
preferences and likings.
__________________
__________________ (ii) Data collection instruments: The primary
__________________
instrument is the questionnaire. Other types of
instruments are inventories and attitude scales
which include Likert scales, semantic differential
scales and rank order scales.

(iii) Sampling–it involves:

(a) The sampling unit–whom to survey:

(b) Sample size–how many to survey; and

(c) Sampling procedure–how to select them.

u Data collection: Highly trained social scientists are


required to collect data in qualitative studies. But, in
quantitative studies a field staff is generally used.

u Analysis: In qualitative researches, the moderator or


test administrator usually analyses the responses
received. In quantitative research, the analysis is
supervised by the researcher.

u Report preparation: It includes a brief executive


summary of the findings. Depending on the assignment,
it may or may not include recommendation for actions.
The report will contain a full description of the
methodology used.

Error Reduction Through Design


The basic purpose of research design is to guide the
researcher in answering the research problem. It is
accomplished through the careful construction of the
research design so that the results obtained are as error free
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 99
as possible. These errors and the researcher strategies for Notes
dealing with them are discussed in the following section: __________________
__________________
1. Planning errors: Planning errors are those which deal
__________________
with the setup of the design to collect information.
Researcher can think of planning errors as being all __________________

those sources of misinformation because of which the __________________


study is inappropriately structured when the research __________________
is undertaken. __________________
__________________
2. Collection errors: They are associated with the
collection of information to answer the research problem. __________________

These errors are sources of misinformation due to the __________________


actual collection of data, given the study is actually well
structured.

3. Analytical errors: They are those which are due to the


inappropriate analysis of the data collected. Researcher
can think of them as the actual manipulation of the data.
To reduce analytical errors, it requires a justification of
the analytical procedures used in manipulating and
summarising the data.

4. Reporting errors: They are those due to the incorrect


interpretation of the results of the study. They put
erroneous meaning into the relationship and numbers
identified from data analyses.

Summary
A research design serves as a bridge between what has
been established (the research objective) and what is to
be done, in the conduct of the study, to realize those
objectives. It anticipates what the client will need in
terms of results and the analytical work on the gathered
data that will convert it to useful findings.

A formal design’s benefits are particularly appreciated


when the researcher is deciding specifically what data
are needed. If data are gathered that prove to be
irrelevant, it is both inefficient and confusing. It is even
more serious to have overlooked some data that are
vitally needed, which may be discovered until too late,
Contd...
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
100 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
at the analytical stage. A design also is appreciated when
__________________
analyzing the data and interpreting and thinking on the
__________________
path to solutions and recommendations.
__________________
__________________
Although, research designs may be classified by many
criteria, the most useful one concerns the major purpose
__________________
of the investigation. On this basis, we may identify the
__________________
broad classes of designs as exploratory, descriptive and
__________________
causal.
__________________
__________________
Review Questions
__________________
1. Discuss the application of exploratory research design
in marketing.

2. Discuss the various types of experimental research


designs.

3. What do you understand by the term research design?


What is its importance? What are the different types of
research design? What relationship exists between
them?

4. What is (A) Exploratory (B) Descriptive (C) Causal


research? Using examples show how they differ from
each other.

5. “The problem definition phase may never come to an


end with one exploratory research leading to another.”
Comment.

6. “In-depth interviews are similar to individual case


studies.” Comment.

7. A coffee marketer uses observational research to study


consumer buying habits. The research design consists
of observing shoppers in supermarkets and recording
the sex and approximate age of consumers who buy
coffee.

(a) Is observation the best research technique to


study this type of consumer behaviour? If not
which method(s) is more appropriate and why?
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 101
(b) In your view is the researcher using a sampling Notes
method? Is this technique appropriate in this __________________
situation? Given the choice, what method would __________________
you use and why? __________________

8. You are the product manager for brand ‘Youth’ shaving __________________
cream, a nationally distributed brand. Brand ‘Youth’ has __________________
been declining in absolute level of sales for the last four __________________
consecutive months. You ask the marketing research __________________
department to do a study to determine why sales have
__________________
declined.
__________________
(a) Indicate whether this is an exploratory, __________________
descriptive or causal study.

(b) What data would be useful for determining the


reasons for decline in sales?

(c) How would you design the study to obtain these


data?

(d) What external secondary data are, in fact, available


that would be useful for this purpose? From what
sources can they be obtained?

9. Describe the various methods for controlling sources of


extraneous variation in experimental designs.

10. Briefly explain the nature and scope of exploratory


research.

11. Describe the sampling design process.

12. A pro-life group wanted to test the effectiveness of an


anti-abortion commercial. Two random samples, each
of 250 respondents, were taken from the city of
Chandigarh. One group was shown the anti-abortion
commercial. Then, attitudes toward abortion were
measured for respondents in both the groups.

(i) Identify the independent and dependent variables


in this experiment.

(ii) What type of design was used?

(iii) What are the potential threats to internal and


external validity in this experiment?
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
102 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 13. What is the relationship between exploratory,


__________________ descriptive and causal research?
__________________
14. What are the requirements for inferring a causal
__________________
relationship between two variables?
__________________
__________________ 15. Differentiate between internal and external validity.
__________________ 16. List any five extraneous variables and give an example
__________________ to show how each can reduce internal validity.
__________________
17. State the type of experiment being conducted in the
__________________
following situations. In each case, identify the potential
__________________ threat to internal and external validity.

(a) A major distributor of office equipment is


considering a new sales presentation programme
for its salespersons. The largest sales territory is
selected, the new programme is implemented and
the effect on sales is measured.

(b) Procter & Gamble wants to determine if the new


package design for “Aerial” is more effective than
the current design. Twelve departmental stores
are randomly selected in Delhi. Further, six
outlets are randomly selected to sell Aerial in a
new packaging. In the other six, the detergent is
sold in the old package. Sales for both groups of
departmental stores are monitored for three
months.

18. You are the product manager for brand S margarine, a


nationally distributed brand. Brand S has been
declining in absolute level of sales for the past four
consecutive months. You ask the marketing research
department to do a study to determine why sales have
declined.

(a) Is this an exploratory, descriptive or causal study?

(b) What data would be useful for determining why


sales have declined?

(c) How would you design the study to obtain these


data?
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 103
19. Explain what research design would you use if you want Notes
to conduct a study regarding the demographic profile __________________
of the users of a particular brand of a product. __________________
__________________
20. List out the uses of consumer research in the field of:
__________________
(a) F.M.C.G. __________________

(b) Durable goods __________________


__________________
(c) Consumer services
__________________
(d) Lubricants. __________________
__________________
21. Every consumer is unique and the study that
concentrates on the ‘average’ consumer is meaningless.
Comment.

22. Compare qualitative and quantitative research.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
104 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes APPENDIX-1
__________________
STRATEGY FOR DESIGNING MARKETING RESEARCH
__________________ PROGRAMME
__________________
__________________
__________________
__________________
__________________
__________________
__________________
__________________
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 105
APPENDIX-2 Notes
__________________
A method of education should rest on some satisfactory theoretical basis.
It should also provide the learner with the knowledge and background __________________
he needs in order to make positive contributions to thought and to society.
__________________
In this context there are three major levels of knowledge concerning __________________
marketing that a student must attain. The first is the acquisition of certain
information, that, in the aggregate, constitutes a description of marketing __________________
institutions and an explanation of marketing processes. This knowledge __________________
may be interesting, but it is arid and purposeless.
__________________
The second level is the development of the principles and theoretical
concepts that are key to understanding the dynamics of the marketing __________________
system. __________________
The third level is the use of the information, principles and theories in the __________________
analysis of marketing problems.

The young person beginning a business career is not expected to


establish the marketing policy for the company. He starts out by performing
more or less routine assignments under the direction of a supervisor. But
as he rises in rank and responsibility, he will have the opportunity to
cope more and more with handling problems–with recommending and
determining what should be changed and with communicating these
judgements to others. If he has the educational experience that accustom
can his mind to problem solution, he will probably be ready for
advancement earlier and will show greater competence in handling the
problems assigned to him.

Although the case method is not a substitute for business experience it is


certainly one of the better types of preparation for making that experience
fruitful. It also provides the participant with the learning method by which
he can enhance not only his knowledge but also develop a sound
diagnostic approach. In the final analysis, it is through the effective use of
one’s own knowledge and insight that he is able to exercise sound
judgement in making decisions.

Importance of Data

The student should attempt to secure general marketing information for


each case. As an example, if there is a question whether or not a company
should use “primary” or “selective” appeals, the student should use some
basic marketing texts as a reference to determine under what conditions
such strategies should be used.

A person having numerous years of marketing experience will certainly


have a great quantity of information. He will know the types of merchandise
that are given counter display space in drug stores. He will know what
items other than food are increasingly being offered in supermarkets. He
will know something of the relative importance of direct selling and selling
through middlemen in department store merchandising. These and many
additional facts will be familiar to him because he has learned them in
the course of his work. As a result, when a marketing problem is presented
to him, he may have many of the facts he needs and can proceed
immediately to analyzing the problem. Often, he may remember similar
problems, and know what solutions have been tried and which ones
have been successful in the past. The student, on the other hand, must
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
106 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes generally search out the information from secondary sources, e.g. trade
association publications, trade publications, texts, industry studies,
__________________
published corporation reports, government data, and primary efforts such
__________________ as interviews with business people and “spot surveys”.

__________________ The marketing student can also rapidly build up a large fund of technical
and trade information by observation and by asking questions in his
__________________
normal contracts with merchants. For instance, many supermarkets and
__________________ many larger drug stores carry hardware store items. In what way does
hardware merchandise in food and drug stores differ from the same
__________________
general type of merchandise carried in hardware stores? This kind of
__________________ knowledge is important in some marketing problems and may be learned
better by observation than any other method.
__________________
__________________ Source: Management Accountant, April, 1998.

__________________
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 4 Research Design 107
CASE STUDY 4.1 Notes
__________________
Shri Ram Consultants
__________________
Shri Ram consultants have been appointed by a leading to prepare a
feasibility report for opening a five-star hotel in Pune. The group had __________________
been the most successful one in the hotel industry and had always kept __________________
its eyes open for new opportunities.
__________________
In view of the very fast industrial growth in the city of Pune, the city had
attracted the attention of the group. Pune, apart from being a modern city, __________________
has the added advantage of pleasant weather and several places of __________________
tourist attraction in the neighbourhood. Moreover, the closeness to
Mumbai, a city of international stature, made it very easily accessible to __________________
international tourists. __________________
For Shri Ram consultants this was the first time that an assignment __________________
concerning the hotel industry had been received. They, however, soon
realized that the assignment was not as simple as it appeared to be in
the first place. The feasibility of such a hotel would depend essentially on
two factors. Businessmen visiting the city for work would constitute one
segment of the market, while tourists would constitute the other. Further,
the tourists could be Indians or foreigners. The success of such a hotel
would also depend upon the relative attraction of other tourist centres in
the vicinity.

Further, it was necessary to estimate fluctuations in demand for hotel


accommodation so that attractive discounts could be offered during the
off-season for business conferences, executive development
programmes, etc.

The consultants realized that they would have to undertake a market


research on a national scale to assess the tourist potential of the city.
They would also have to survey the foreign tourists to estimate one of the
most important segments of the market.

The consultants wondered whether such a survey will have to extend


over a period of one full year to completely take into account the seasonal
variations in tourists traffic. They were also undecided about the manner
in which the survey should be conducted. The company also feared that
in the absence of an accurate definition of the problem, they may land up
surveying the compete tourist market in India rather than studying the
feasibility of a hotel in Pune. They had heard that ITBC had appointed
lady interviewers at several major airports for conducting such surveys
and wondered if such a methodology could be of use to them.

The problem appeared well defined and they were concerned as the
preliminary report explaining methodology of the research and the
questionnaires to be used had to be submitted to the client along with
the estimate of expenses within one month.

Questions

(a) How would you define the problem of assessing the feasibility of
the hotel so as to help design the survey?

(b) How would you plan a survey for actually collecting information on
expected demand for hotel space? How would you estimate the
cost of such a survey?
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
108 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes CASE STUDY 4.2


__________________
Excel Publishing House
__________________
Excel publishing house has been in existence for the past 15 years or so.
__________________ The company was set up by an entrepreneur Shri Anurag Jain, whose
__________________ main emphasis is to publish management books and sell them at
moderate prices so that they may be within the reach of management
__________________ students.
__________________ After a few years of experience in the publishing house, Shri Jain felt that
__________________ the company should concentrate on selected management books,
according to prescribed course contents of the institutes. But he feels
__________________ that the problem of selecting a particular textbook for publication is very
__________________ crucial in view of increasing competition in the textbook market. Mr Jain
wants to develop a proforma to elicit relevant information from institutes/
__________________ authors interested in having their works published.

Questions

1. What sample design is being used by Mr Jain?

2. What type of study is this?


109

Unit 5 Notes
__________________
__________________
Hypothesis Design __________________
__________________
__________________
__________________
__________________

Objectives __________________
__________________
After reading this unit you will be able to understand:
__________________
y Various steps involved while testing the hypothesis.

y The difference between the terms, "standard deviation" and


"standard error".

y The applicability of standard error in testing the hypothesis.

y Conditions required while using the test of significance.

y The concept of X2 test and its features.

y Apply and interpret the results of the ANOVA and n-way ANOVA
statistical methods.

Marketing researchers can use secondary or primary data,


or both, to satisfy their information-gathering needs.
Marketing managers always use secondary data, because
previously collected information can benefit marketing
managers and researchers in a number of ways: (1) Time
saving, (2) Relevance, (3) Availability, (4) Low cost, (5)
Accessibility.

Definition
The distribution of all possible values which can be assumed
by same statistics computed from samples of the same size
randomly drawn from the same population is called the
sampling distribution of the statistic.

Example 5.1
Let the population size =N
Let the sample size =n
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
110 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Then we can draw a total number of NCn samples. Now we


__________________ can calculate any statistic (say mean) of these NCn samples.
__________________ The distribution then obtained is called the sampling
__________________ distribution of the statistics (mean).
__________________
__________________
S1
__________________ Size n Mean m1

__________________
__________________
Population
__________________ Size S2
Size n Mean m2
N
__________________

Sn
Size n Mean mn

Figure 5.1

These NCn samples gives value of mean as m1, m2…m12. The


values form a distribution which is known as sampling
distribution. The sampling distribution has the following
important properties:

1. The arithmetic mean of the sampling distribution is the


same as the mean of the universe from which samples
were taken.

2. The sampling distribution of mean has a standard


deviation (S.E.) equal to the population standard
deviation divided by the square root of the samples size.
i.e.

6
S.E =
n

3. The sampling distribution of the means is normally


distributed.

Steps in Testing the Hypothesis


Steps involved in the process of hypothesis testing are given
by the Figure 5.2.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 111
Notes
__________________
__________________
Problem definition
__________________
__________________
Clearly state the
null and alternate __________________
hypothesis
__________________
Choose the relevant __________________
test and the
appropriate probability __________________
distribution
__________________
Determine the Select the critical Determine the degree __________________
significance level value of freedom

Decide if one or two


tailed test
Compute relevant test Compare test statistic
statistic and critical value

Does the test statistic No Do not reject the null


fall in the critical region hypothesis
Yes
Reject the null
hypothesis

Figure 5.2

Hypothesis Testing and Associated Statistical Tests


Table 5.1
Hypothesis No. of Groups/ Purpose Statistical Test Assumptions/
Testing Samples Comments

Frequency One Goodness of Fit x2

Distributions Two Test of x2


Independence

Proportions One Comparing sample Z If r is known, and for


and populations large samples
proportions

Two do t If r is unknown and for


small samples

Two Comparing two Z If r is known


samples proportions
Contd...
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
112 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes do t If r is unknown

__________________
Means One Comparing sample Z If r is known
__________________ and population
mean
__________________
__________________ do t If r is unknown

__________________ Two Comparing two Z If r is known


samples mean
__________________
__________________ do t If r is unknown

__________________ From independent t If r is unknown


samples
__________________
__________________ Two or more Comparing multiple F Using analysis of
sample means variance

2
Variance One Comparing sample X
and population
variance

Two Comparing sample F


variance

Two conditions are required to use the test of significance:

I st Condition: Researcher will have observed some


difference between two or more groups, i.e. town x versus
town y, educated versus uneducated.

IInd Condition: Researcher will have asked same question


of each group and/or measured the same thing within each
group (e.g., usage of a certain product or brand.)

Standard Error
The standard deviation of a sampling distribution is called
standard error of that statistic. For example,

1. The standard deviation of the means of all possible


samples of the same size drawn from a population is
known as the standard error of the mean.

2. The standard deviation of the proportions of all possible


samples of the same size drawn from a population is
known as standard error of proportion.

Note: The difference between the terms "standard


deviation" and "standard error" lies in the fact the S.D.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 113
deals with original values and S.E. deals with statistics Notes
computed from the samples of original values. __________________
__________________
Importance of the Concept of Standard Error
__________________
1. It can be used in testing hypothesis. For example, let __________________
the level of significance for testing a particular __________________
hypothesis be 5%. We know that corresponding to this __________________
level the value of standard error is 1.96 S.E. Now if the __________________
difference between the observed and expected means
__________________
comes out to be more than 1.96 S.E. then the result of
__________________
the experiment does not support the hypothesis which
__________________
it supports otherwise.

2. Standard error can be used to determine the limits for


parameter values.

3. Standard error provides an idea about the unreliability


of the sample. Because if the standard error has a large
value then there is great difference between the actual
and expected frequency. Hence greater is the
unreliability of the sample. In general, we have,
1
Reliability (R) a
S.E

Difference between Large and Small Samples


It is very difficult to draw a clear cut line of demarcation
between large and small samples, it is normally agreed
amongst statisticians that a sample is to be recorded as large
only if its size exceeds 30. The tests of significance used for
dealing with problems relating to large samples are different
from the ones used for small samples for the reason that the
assumptions that we make in case of large samples do not
hold good for small samples The assumptions made while
dealing with problems relating to large samples are:

1. The random sampling distribution of statistic is


approximately normal, and

2. Values given by the samples are sufficiently close to the


population value and can be used in its place for
calculating the standard error of the estimate.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
114 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Example 5.2


__________________
Calculate standard error of mean from the following data
__________________
showing the amount paid by 100 firms in Delhi on the
__________________
occasion of Durga Pooja.
__________________
Mid Value (Rs.) 39 49 59 69 79 89 99
__________________
__________________ No. of firms 2 3 11 20 32 25 7

__________________ Solution
__________________ σ
S.E. X =
__________________ n
__________________ Table
Mid-value m – 69/10
m f d fd fd2

39 2 -3 -6 18

49 3 -2 -6 12

59 11 -1 -11 11

69 20 0 0 0

79 32 +1 + 32 32

89 25 +2 + 50 100

99 7 +3 + 21 63

N = 100 ∑ fd = 80 ∑ fd 2= 236

2
∑ fd 2  ∑ fd 
s= −  ×i
N  N 

2
236  80 
= −  × 10
100  100 

= 13.11

13.11
S.E. = X = = 1.311
100

Case I: Standard Deviation s Unknown and n ³ 30


In this case, the sample standard deviation is calculated
as:

Σ (x1 − x )
2
δ=
n
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 115
This value of d is used as an estimate of s (provided that n ³ Notes
30). In this case, the values are given by: __________________
__________________
x − µο
= __________________
s n
__________________

Here, __________________
__________________
Critical values:
__________________
δ __________________
X L = µo − α when H1 : µ < µ ο
n __________________
__________________
δ
X U = µυ + α when H1: µ > µ ο
n

Example 5.3
A sample of 64 farm labourers engaged in paddy harvesting
operations shows an average monthly wage rate of Rs 200
with a standard deviation of Rs 9. Using 0.05 level
of significance, verify if the sample result indicates that
their current average monthly wage rate is higher than
Rs 198.

Solution
1. H0 : µ = Rs 198.

H1 : µ > Rs 198.

2. a = 0.05.

3. Critical region: Ζ ≥ Ζ α , where Ζ α = 1.645.

4. Computations:

For x = 200, σ = s = 9, and n = 64,

x − µ0 200 − 198
z= = = 1.78
s n 98

5. Conclusion: since z = 1.78 is greater than za = 1.645, Ho


is rejected and we conclude that the current wage rate
is higher than Rs 198.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
116 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Case II: Population Standard Deviation Unknown


__________________
Many a times situations do arise when neither s is known,
__________________
nor is it possible to draw a sample of size to allow the use of
__________________
S2 as an unbiased estimator of . That is, when n<30 and s
__________________ unknown, testing H0 concerning the population mean m is
__________________ based on the sample statistic:
__________________ X − µ0
T= ∧
__________________
S n
__________________
Which is a random variable having t distribution with
__________________
v = n – 1 df.
__________________
As usual, we draw a sample of size n<30 and compute the
sample mean and the sample standard deviation

Σ (xi − x )
∧ 2

S=
n −1

For a sample of size n<30,

Σ (x i − x )
2
S=
n

Example 5.4
15 workers are selected at random from a large no. of workers
in a factory. The number of items produced by them on a
certain day are found to be:

53 52 54 53 51 53 59 58 56 56 57 54 50 59 60

Would it be appropriate to suggest that the mean of the


number of items produced in the population is 59.

Solution
Let us first calculate sample mean and standard deviation
X X-55 (X-55)2
53 -2 4
52 -3 9
54 -1 1
53 -2 4
Contd...
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 117
51 -4 16 Notes

53 -2 4 __________________

59 4 16 __________________
58 3 9 __________________
56 1 1 __________________
56 1 1 __________________
57 2 4 __________________
54 -1 1 __________________
50 -5 25 __________________
59 24 16 __________________
60 5 25 __________________
------ ------- -------
825 136

Σx 825
X= = = 55
n 15

Σ (x − x ) 136
δ= = = 9.71
n −1 14

Test Involving Two Different Populations


There may be cases when we have to deal with populations
instead of population and sample.

Example 5.5
1. A doctor may be interested in knowing whether a new
medicine A is better than new medicine B in curing a
particular disease.

2. A farmer may be interested in knowing whether a new


variety of wheat seed will result in more yield per
hectare than that of some other variety.

Let the mean of first population = m1

Let the mean of second population= m2

In their case, we have,

Null hypothesis Ho = m1- m2 = do = (say)

Alternative hypothesis H = m1- m2 ¹ do (two in tail test)


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
118 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes = m1- m2 > do (in one tail test)


__________________
= m1- m2 < do (in one tail test)
__________________
__________________ Note: If mean of two populations are equal
__________________
i.e. µ1 = µ 2 ⇒ do = µ1 − µ 2 = 0
__________________
__________________ Then
__________________ H 0 : µ1 = µ 2
__________________
__________________
H α = µ1 ≠ µ 2 , µ 1 > µ 2 , µ1 < µ 2
__________________ Case

1. Population variances σ1 2 and σ2 2 are not known

2. σ1 2 = σ2 2

3. n 1 , n 2 < 30

This test is based on t-distribution null hypothesis and


alternative hypothesis have already stated. In this case,
t-test becomes

t=
(x1 − x2 ) − do
 1   1 
S p   +  
 n1   n 2 

Here,

x1 , x 2 are sample means.

do = µ1 − µ 2 , where µ1 , µ 2 are population means

n 1 + n 2 are size of the two random samples

Sp2 Common variance

n1s12 + n2 s22
SP =
n1 + n2 − 2

Here degree of freedom,

d f = n1 + n 2 − 2
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 119
Example 5.6 Notes
__________________
Consider two groups A and B of 10 and 15 students each
__________________
respectively, Group A is being taught with traditional
__________________
method and group B with some new method. After the
completion of the course a test was conducted with the same __________________

question paper for the two groups. It was found that on an __________________
average group A obtained 65 marks with a standard deviation __________________
of 7 marks and group B obtained 68 marks with a standard __________________
deviation of 4 marks. Assuming the populations to be normal __________________
and having the same variance, test the hypothesis at 0.01
__________________
level of significance that the two methods are equally
__________________
effective.

Solution
Here we have to examine for the two methods to be equally
effective. Let m1 and m2 be the average marks obtained by
the two groups respectively, then

1. Null hypothesis H0= µ1 − µ 2 = 0

= µ1 = µ 2

2. Alternative hypothesis H1 = µ1 ≠ µ 2

3. α = 0.01

4. d.f = n 1 + n 2 − 2 = 10 + 15 − 2 = 23

Now,

t (d.f = 23) = −2.807

α = 0.01

Apply t-test

t=
(x −x2 )
1

1 1
SP +
n1 n2

Here, x 1 = 65 , x 2 = 68

n 1 = 10 , n 2 = 15
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
120 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
n 1s12 + n 2 s22
__________________ SP =
n1 + n2 − 2
__________________
__________________
10 × 72 + 15 × 42
__________________ =
10 + 15 − 2
__________________
__________________ = 31.7
__________________
So,
__________________
165 − 681
__________________ t=
__________________ 31.7 1 10 + 1 15

− 516
=
5.63 × ( 0.1 + 0.06 )
− 516
=
5.63× 0.16

− 516
=
5.63× 0.4

− 516
=
2.252

= − 229.129

Since calculated value is more. So,

Ho is accepted.

Case III: Population Variances σ12 and σ 22 Unknown,


σ12 ≠ σ 22 , and n 1 ≠ n 2 < 30

Some decision problems involving tests concerning the


difference of means may be characterized by the following:

1. The two populations are approximately normal with


σ12 ≠ σ22 , and

2. The sample sizes n1 and n2 are less than 30 n 1 ≠ n 2 .

In such problem situations, testing the null hypothesis


H 0 : µ1 − µ 2 = d o against the alternative hypothesis.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 121
H 1 : µ1 − µ 2 ≠ d 0 , µ1 − µ 2 > d 0 , or µ1 − µ 2 < d 0 Notes
__________________
is based on the test statistic T whose distribution is
__________________
approximately like a t distribution with degrees of freedom
__________________

2 __________________
 ∧ 2   ∧ 2 
 s 1   s 2  __________________
 n  +  n  __________________
 1   2 
v=   __________________
 ∧ 2 
2
  ∧ 2 
2

 s 1 n 1  (n 1 − 1) +  s 2 n 2  (n 2 − 1) __________________
 
     __________________
   
__________________

where a particular t value is defined as:

t=
(x1 − x2 ) − d 0
 ∧2   ∧2 
 s1   s 2 
 + 
 n1   n 2 
   

Example 5.7
A random sample of 10 acres wheat crop selected from
Region I showed an average yield of 20 qtls per acre with a

standard deviation S1 = 5 qtls. Another random sample of 15
acres of wheat crop selected from Region II offered an average

yield of 18 qtls per acre with a standard deviations S2 = 3
qtls. Assuming that the per acre wheat yield in the two states
is normally distributed with different variances, test the
hypothesis at 0.05 level of significance that the average per
acre wheat yield is the same in the two regions against the
alternative that it is higher in Region I than II.

Solution
Let m1 be the average per acre wheat yield in Region I and
m2 in Region II.

1. H0 : m1 = m2

H1: m1 > m2

2. a = 0.05.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
122 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 3. Critical region:


__________________
__________________  ∧2 ∧2 
 s1 s 2 
__________________  + 
 n1 n 2 
__________________  
For v =  2 2
  ∧ 2 2

__________________ ∧  
 s 1 n 1 

(n 1 − 1) +  s 2 n 2 
    (n 2 − 1)
__________________
     
__________________    
__________________
2
__________________  25 9 
 + 
__________________  10 15 
=
 (25 10 )2   (9 15)2 
 + 
 9   14 

= 13,

the tabulated T value for a = 0.05 with v = 13df is ta =


1.771, and the critical region is t ³ ta.

4. Computations: For x1 = 20 , S1 = 5 , n1= 10,

x 2 = 18 , S2 = 3 , n1= 15,

t=
(x1 − x2 ) =
20 − 18
= 1.136
   
∧2 ∧2  25   9 
 S1   S2   + 
 +   10   15 
 n1   n 2 
   

5. Conclusion: Since t = 1.136 is less than ta = 1.771, Ho is


accepted. It means that there is no difference between
the average per acre wheat yield in the two regions.

X2 Test (Chi Square Test)


The X2 test is one of the simplest and most widely used non-
parametric tests in statistical work. The symbol X 2 is the
Greek letter Chi. This test gives the magnitude of
discrepancy between theory and observation. It is defined
as:

Σ
(O − E )2
X2 =
E
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 123
Where O is observed frequency, E is expected frequency. Notes
__________________
In general expected frequency is given by:
__________________

RT × CT __________________
E=
N __________________
__________________
Here, RT= Row Total containing the cell.
__________________
CT= Column Total containing the cell. __________________
__________________
N = Total No. of observations.
__________________
Degree of Freedom (V) __________________

The Chi square test is based on degrees of freedom. This is


obtained as:

(i) In case of one dimension (row or column) there are K-1


degrees of freedom, where K is number of categories of
observed frequencies.

(ii) In case of contingency table

v = (r-1) (c -1)

Here, r = is total no. of rows

c = is total no. of columns

Characteristics of Chi Square Test


1. This test is based on frequencies or events as against
the and the t tests based on parameters like the mean
and standard deviation.

2. This is applied for drawing inference only.

3. This possesses additive properties so that when X 12 and


X 22 are independent and have a Chi square distribution
with n 1 and n 2 degrees of freedom, will also be
distributed as a Chi square distribution with n 1+ n 2
degree of freedom.

4. It is a general purpose test and is very useful in research


work.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
124 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Chi square analysis can be used when the data satisfy four
__________________ conditions:
__________________
1. There must be two observed sets of data or one observed
__________________
set of data and one expected set of data.
__________________
__________________ 2. The two sets of data must be based on the same sample
size.
__________________
__________________ 3. Each cell in the data contains an observed or expected
__________________ count of five or larger.
__________________
4. The different cells in a row or column must represent
__________________ categorical variables.

There are three main encountered applications of Chi square


analysis. Perhaps the most common occurs when researchers
observe that, for a certain product, men show a different
distribution of consumption rates than women. Researchers
often must test whether the difference in such distributions
are statistically significant.

Example 5.8
On the basis of the information given below about the
treatment of 200 patients suffering from a disease, state
whether the new treatment is better or not? (Check at 5%
level of significance).
No. of Patients
Favourable Not Favourable Total

New 60 30 90

Conventional 40 70 110

100 100 200

Solution
1. Null hypothesis (Ho): There is no difference between
new and conventional method.

2. Degrees of freedom V = (r–1) (c–1)


= (2–1) (2–1)
= 1
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 125
Notes
RT × CT
3. Expected frequency for each cell = i.e., __________________
N
expected frequencies are: __________________
__________________
100 × 90 100 × 90 __________________
200 200 __________________

100 × 110 100 × 110 __________________

200 200 __________________


__________________
i.e.
__________________
45 45 90 __________________
55 55 110
100 100 200

We calculate value of X2 as following:


O E O2 O2/E

60 45 3600 80.09
40 55 1600 29.00
30 45 900 20.00
70 55 4900 89.09

218.18

 O2 
X 2 = Σ   − N = 218.18 − 200 = 18.18

 E 

(O − E )2
or we can calculate Σ
E

Here, table value of X2 for V= 1 and p = 0.5 is 3.84

Calculated value is 18.18 So,

X 2col > X 2table

So, null hypothesis is rejected and so the new method is


better.

Example 5.9
Following is the information about the number of clerk
committing errors and not committing errors among trained
and untrained clerks.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
126 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes No. of Clerks No. of Clerks not


__________________ Committing Errors Committing Errors

__________________ Trained 70 530

__________________ Untrained 155 745


__________________
Test the effectiveness of training in preventing the errors.
__________________
__________________ Solution
__________________
H o: Training is not effective in preventing errors. We
__________________ compute expected frequencies as:
__________________
Expected Frequencies
__________________
No. of Clerks No. of Clerks not
Committing Errors Committing Errors

600 × 225 600 × 1275


Trained = 90 = 510
1500 1500

900 × 225 900 × 1275


Untrained = 135 = 765
1500 1500

We calculate the value of X2 as:


Observed frequency Expected frequency
ο–∈
(ο − ∈)2 ∈
ο ∈

70 90 -20 4.444
530 510 20 0.7843
155 135 20 0.29630
745 765 -20 0.5230
2
X = 8.7147

∴ X 2col = 8.7147

Here V = (2–1) (2–l) = 1

X 20.05 (for v=1) = 3.84

X 2Col > X 20.05

So, null hypothesis is rejected and training is effective in


preventing the errors.
Example 5.10
200 digits were selected at random from a set of tables. The frequencies
of the digits were:
Digit 0 1 2 3 4 5 6 7 8 9
Freq. 18 19 23 21 16 25 22 20 21 15
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 127
Use X2 test to test if the digits were distributed in equal nos. Notes
__________________
Solution __________________

Ho: The digits are distributed in equal nos. __________________


__________________
Here expected frequency of each digit
__________________

18 + 19........ + 15 200 __________________


= = = 20
10 10 __________________
__________________
Digit Frequency o-∈ (o − ∈)2
∈ __________________
o ∈ __________________
0 18 20 -2 0.2

1 19 20 -1 0.05

2 23 20 3 0.45

3 21 20 1 0.05

4 16 20 -4 0.8

5 25 20 5 1.25

6 22 20 2 0.2

7 20 20 0 0

8 21 20 1 0.05

9 15 20 -5 1.25

X2 = 4.30

Analysis of Variance
It is an useful technique concerning researches in the fields
of economics, education, business and in other disciplines.
This technique is used when multiple samples cases are
involved. The significance of the difference between the
means of two samples can be found out through Z test or t
test, but the problem arises when we happen to examine the
significance of difference amongst more than two sample
means at the same time.

Basic Principle of Analysis of Variance (ANOVA)


The basic principle is to test for difference among the means
of the populations by examining the amount of variation
within each of these samples, relative to the amount of
variation between the samples. In terms of variation within
the given population, it is assumed that the value of (Xij)
differs from the mean of this population only because of
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
128 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes random effects, i.e., there are influences (X ij) which are
__________________ unexplainable, whereas in examining differences between
__________________ populations we assume that the difference between the mean
__________________ of the jth population and the grand mean is attributable to
__________________ what is called a ‘specific factor’ or what is technically
described as treatment effect. Thus, while using ANOVA we
__________________
assume that each of the samples is drawn from a normal
__________________
population and that each of these populations has the same
__________________
variance. We also assume that all factors other than the one
__________________ or more being tested are effectively controlled.
__________________
__________________ The F Test (Variance Ratio Test)
The F-test is named in honour of the great statistician R A
Fisher. It is used to find out whether the two independent
estimates of population variance differ significantly. For
carrying out the test of significance, we calculate the ratio
F. F is defined as:

S12
F=
S22

(X − X1
1 )
2

Where, S12 =
n2 −1

(
Σ X2 − X2 )
2

and S22 =
n2 −1

Note that S12 is always greater than S22

i.e. S12 > S22

Therefore,

Larger estimate of variance


F=
Smaller estimate of variance

V1 = n1 – 1
V2 = n – 2
V1 = degrees of freedom for sample having larger variance.
V2 = degrees of freedom for sample having smaller variance.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 129
The calculated value of F is compared with the table value Notes
(given at 5% or 1% level of significance). If the calculated __________________
value of F is larger than the table value of F, the null __________________
hypothesis is rejected otherwise it is accepted. __________________
__________________
Example 5.11
__________________
Two samples are drawn from two normal populations. From __________________
the following data test whether the two samples have the __________________
same variance at 5% level.
__________________
Sample 1: 60 65 71 74 76 82 85 87 __________________
__________________
Sample 2: 61 66 67 85 78 63 85 86 88 91

Solution
Null hypothesis Ho : σ12 = σ22 , i.e., samples have been drawn
from two normal populations with the same variance.

Computation of Sample Variances


Sample 1 Samples 2
2 2
X X-X (X-X) Y Y-Y (Y-Y)

60 -15 225 61 -16 256


65 -10 100 66 -11 121
71 -4 16 67 -10 100
74 -1 1 85 8 64
76 1 1 78 1 1
82 7 49 63 -14 196
85 10 100 85 8 64
87 12 144 86 9 81
88 11 121
91 14 196

636 1200

600 770
X= = 75 Y= = 77
8 10

S12 =
1
n1 − 1
Ó X−X ( )
2
=
636
7
= 90.8

S22 =
1
n2 −1
Σ Y−Y ( ) 2
=
1200
9
= 133.33

Clearly, S22 > S12

S22 133 .33


Þ F= = = 1.467
S12 90.8
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
130 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Degrees of freedom F (n2 – 1, n1 – 1) = F (9, 7)


__________________
Now, tabulated value for F at d.f 9 and 7 and 5% level = 3.68
__________________
since
__________________
__________________ Fcal < Ftable
__________________ Þ Null hypothesis is accepted.
__________________
__________________ Example 5.12
__________________ The time taken by workers while performing a job by two
__________________ methods m1 and m2 are given as:
__________________
m1 : 30 15 29 26 21 25

m2 : 26 30 40 31 34 19 27

Check if the data show that the variances of time distribution


in a population from which these samples are drawn do not
differ significantly.

Solution
Null hypothesis Ho : σ12 = σ22 ,

There is no significant difference between the variance of


the time distribution by the workers in performing a job by
m1 and m2.

M1 M2
2
X X-X (X-X) Y Y-Y (Y-Y)2

30 4 16 26 -5 25
15 -11 121 30 -1 1
29 3 9 40 9 81
26 0 0 31 0 0
31 5 25 34 3 9
25 -1 1 29 -2 4
27 -4 16

156 172 217 136

156 217
X= = 26 Y= = 31
6 7

S12 =
1
n1 − 1
(
Σ X−X )
2
=
172
5
= 34.4
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 131
Notes
S22 =
1
n2 −1
Σ Y−Y ( )
2
=
136
6
= 22.66 __________________
__________________
Since, S12 > S22 __________________

S12 34.4 __________________


Þ F= 2
= =
S2 22.66 __________________

Degrees of freedom F (n2 - 1, n1 - 1) = F (6, 5) __________________

Þ Ftable = 4.95 __________________


__________________
Since Fcal < Ftable
__________________
Þ Null hypothesis is accepted.
__________________

Example 5.13
Perform a two way ANOVA on the following data:
T1

I II III

i 30 26 38
ii 24 29 28
T2 iii 33 24 35
iv 26 31 30
v 27 35 33

Use 30 as the code value.

Solution
Since 30 is the code value. So, subtracting 30 from all cells
we have,
T1
Total
I II III

i 0 -4 8 4
ii -6 -1 2 -9
T2 iii 3 -6 5 2
iv 6 1 0 7
v -3 5 3 5

Total 0 -5 18 T=9

T 2 92
Correction Factor (C.F) = = = 5.40
N 15

Total sum of squares (SST) = O2 + (–6)2 + 32 + 62 + (–3)2 + (–4)2


+……..(0)2 + 32 – C.F.
= 271 - 5.4
= 265.63
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
132 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Sum of squares for variance between T1 (i.e. columns) sample


__________________
 2 142 
__________________
i.e. SSC = (O ) + (− 5 )2
+  − 5 .4
__________________  5 
__________________ = 38.8
__________________
Sum of square of variance between T2 samples (i.e. row)

[ ]
__________________
__________________ SSR = 42 + (− 9 )2 + (7)2 + (5)2 3 − 5.4
__________________ = 52.93
__________________
The error sum of squares (SSE)
__________________
SSE = SST - SSR - SSC
= 265.6 - 52.93 - 38.8
= 173.87

ANOVA Table 5.2

Source of Sum of D.f Mean Squares Variance


Variance Square Ratio
Between 38.80(SSC) 2(n – 1) SSC MSC
MSC = FC =
treatment n−1
are samples MSC
= 19.40 21.73
=
19.40
= 1.12
Between 52.93(SSR) 9(k–1) SSR MSR
treatment the MSR = FR =
K−1 MSE
samples
= 13.23 13.23
=
21.13
= 0.61
Error SSE
173.87SSE 8(n–1)(k–1) MSE =
(n − 1)(k − d)
= 21.73

Total 265.60SST 19(nk–1)

Degree of freedom = (10 – 1) = 9

Now X 20.05 (v = a ) = 16.919

Here X 2Cal < X 2Table

So, null hypothesis is accepted.


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 133
Analysis of Co-variance (ANOCOVA) Notes
__________________
The object of experimental design in general happens to be
__________________
to ensure that the result observed may be attributed to the
__________________
treatment variable and to no other casual circumstances. For
__________________
instance, the researcher studying one independent variable
X, may wish to control the influence of some uncontrolled __________________

variable Z, which is known to be correlated with the __________________


dependent variable, y, then he should use the technique of __________________
analysis of co-variance. __________________
__________________
Assumptions in ANOCOVA
__________________
1. Various treatment groups are selected at random from
the population.

2. The groups are homogenous in variability.

3. The regression is linear and is same from group to group.

ANOCOVA Technique
While using ANOCOVA technique, the influence of
uncontrolled variable is usually removed by simple linear
regression method and the residual sums of square are used
to provide variance estimate which in turn are used to make
test of significance. In other words, covariance analysis
consists in subtracting from each individual score (Yi) that
portion of it Y: that is predictable from uncontrolled variable
(Zi) and then computing the usual analysis of variance on
the resulting (Y-Y)s, of course making the due adjustment
to the degree of freedom because of the fact that estimation
using regression method requires loss of degree of freedom.

Example 5.14
The following are paired observations for three experimental
groups.
Group I Group II Group III
x y x y x y
7 2 15 8 30 15
6 5 24 12 35 16
9 7 25 15 32 20
15 9 19 18 38 24
12 10 31 19 40 30
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
134 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Degree of freedom associated with adjusted sums of square


__________________ will be as under:
__________________ Between K-1
Within N-K-1
__________________
Total N-2
__________________
__________________ Y is the co-variate variable, calculate the adjusted total,
__________________ within groups and between groups, sums of squares on X
__________________ and test the significance of difference between adjusted
means on X by using the appropriate F-ratio, also calculate
__________________
the adjusted means on X.
__________________
__________________ Solution
We apply the technique of analysis of covariance and work
out the related measures as under.
Group I Group II Group III

x y x y x y

7 2 15 8 30 15

6 5 24 12 35 16

9 7 25 15 32 20

15 9 19 18 38 24

12 10 31 19 40 30

Total 49 33 114 72 175 105

Mean 9.8 6.6 22.8 14.4 35.0 21.0

SX = 49+114+175 = 338

(ΣX )2 = (338 )2 = 7616 .27


Correction factor for X =
N 5

SY = 33+72+105=210

(ΣY )2 = (210 )2 = 2940


Correction factor for Y
N 5

SX2=9476 SY2=3734 SXY=5838

ΣX.ΣY
Correction factor for XY = = 4732
N
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 135
Hence total SS for X= SX2-correction factor for X Notes
__________________
= 9476-7616.27=18 59.73
__________________

 (49 )2 (114 )2 (175 )2  __________________


SS between for X =  5 + 5 + 5  – (Correction __________________
 
__________________
factor for X)
__________________
= (480.2+2599.2+6125) – (7616.27) __________________
= 1588.13 __________________

SS within for X = (total SS for X) – (SS between for X) __________________

(1859.73)–(1588.13) = 271.6 __________________

Similarly we work out the following values in respect


of Y:
Total SS for Y = SY2 – correction factor for Y
= 3734-2940 = 794

 (33)2 (72 )2 (105 )2 


SS between for Y =  + +  − (Correction factor
5 5 5 
 
for Y)
(217.8+1036.8+2205)-(2940)= 519.6
SS within for Y = (total SS for Y)-(SS between for Y)
794-519.6 = 274.4
We work out the following values in respect of both X and Y.
Total sum of product for XY = SXY - correction factor for XY
5838 – 4732 = 1106

 (49 ) (33) (114 ) (72 ) (175) (105) 


SS between for XY =  + +  –
 5 5 5 
correction factor for XY
= (323.4+1641.6+3675)–(4732) = 908
SS within for XY = (total sum of product) – (SS between for
XY)
1106 - 908 = 198
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
136 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes ANOVA Table for X, Y and XY can now be set up as shown


__________________ below.
__________________ Source d.f. SS for X SS for Y Sum of product XY
__________________ Between group 2 1588.13 519.60 908
__________________
Within group 12 Exx 271.6 Err 274.4 Exy 198
__________________
Total 14 Txx 1859.73 Trr 794.00 Txy 1106
__________________
__________________ (Txy )2 (1106 )2
Adjusted total SS = Txx – or 1859.73– or
__________________ Tyy 794
__________________ 1859.73–1540.6 = 319.13
__________________

Adjusted SS within group = Exx–


(Exy )2 = 271.60 −
(198)2
Eyy 274.4

= 271.6 - 142.87 = 128.73

Adjusted SS between groups = Adjusted total SS – Adjusted


SS within group

(319.13 – 128.73) = 190.4

ANOVA Table for Adjusted X


Source d.f SS MS F-ratio

Between group 2 190.4 95.2 8.14

Within group 11 128.73 11.7

Total 13 319.13

At 5% level the table value of F for V1=2 and V2=11 is 3.98


and at 1% level the table value of F is 7.21. Both these values
are less than the calculated value and according we infer
that F-ratio is significant at both levels, which means the
difference in group means is significant.

Adjusted mean on X will be worked out as follow:

Regression coefficient for X on Y i.e., b =


Sum of product wi thin group
Sum of square within group for Y

198
= .7216
274.40
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 137
Deviation of initial group means from Notes
__________________
general mean (=14) in case of Y
__________________
Final means of group in X (unadjusted) __________________
__________________
Group I –7.40 9.8
__________________
Group II .40 22.8 __________________

Group III 7.00 35.00 __________________


__________________
\ Adjusted means of groups in X = (Final mean) -b (deviation
__________________
of initial mean from general mean in case of Y)
__________________
Hence,

Adjusted mean for group I = 9.8 – .7216 (–7.4) = 15.14

Adjusted mean for group II = 22.80 – .7216 (.40) = 22.51

Adjusted mean for group III = 35.00 – .7216 (7.00) = 29.95

Summary
This chapter presents the fundamental underpinnings
of hypothesis testing methodology. In this chapter
parametric and non-parametric tests are available for
testing the hypothesis related to differences. In the
parametric case T test is used to examine hypothesis
related to the population mean. Different forms of T test
are suitable for testing the hypothesis based on one
sample, two independent samples, or paired samples,
in the non-parametric case popular one-sample test
include chi-square (X2) test. In ANOVA and ANOCOVA,
the dependent variable is metric and independent
variables are all categorical 02 combinations of
categorical and metric variables. One way ANOVA
involves a single independent categorical variables.
Interest lies in testing the null hypothesis that the
category means are equal in the population. The total
variation in the dependent variable is decomposed into
two components – variation related to the independent
variable and variation related to error. The null
hypothesis of equal means is tested by an F statistic,
Contd...
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
138 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
__________________
which is the ratio of the mean square related to the
independent variables to the mean square related to
__________________
error. ANOCOVA includes at least one categorical
__________________
independent variable and at least one interval or metric
__________________
independent variable. The metric independent
__________________ variables, or covariate, is commonly used to remove
__________________ extraneous variation from the dependent variable.
__________________
__________________ Review Questions
__________________
1. Why do we use hypothesis testing?
__________________
2. Why and how would you use t and z test in hypothesis
testing?

3. Why and when would you want to use ANOVA in


marketing research?

4. What will ANOVA test not tell you, and how can you
overcome this problem?

5. Which test is used to check for statistical significance


in an ANOVA?

6. What are the commonly occurring dependent variables


in marketing applications of ANOVA?

7. Discuss the similarities and differences between


analysis of variance and analysis of covariance.

8. What is the relationship between analysis of variance


and the t test?

9. What is the null hypothesis in one way ANOVA? What


basic statistic is used to test the null hypothesis in one
way ANOVA? How is this statistic computed?

10. What is the most common use of the covariate in


ANCOVA?

11. What are the major assumptions of ANOVA?

12. Under what conditions should the t test for the mean
difference in two related populations be selected.

13. What is the difference between two independent and


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 5 Hypothesis Design 139
two related populations? Notes
__________________
14. Under what conditions should the F test be selected to
__________________
examine possible differences in the variances of two
__________________
independent populations?
__________________
15. What are some of the criteria used in the selection of a __________________
particular hypothesis-testing procedure?
__________________
16. Under what conditions should the two tests be used to __________________
examine possible differences in the proportions of two __________________
independent populations? __________________

17. Under what conditions should the X 2 test be used to __________________

examine possible differences in proportions of two


independent populations?

18. What are the similarities and distinctions between the


Z and X2 test for differences in populations proportions?

19. Under what conditions should the X 2 test of


independence be used?
141

Unit 6 Notes
__________________
__________________
Sampling Procedures __________________
__________________
__________________
__________________
__________________
__________________
Objectives __________________
After reading this unit you will be able to understand: __________________

y Meaning of sampling.

y Need of sampling.

y Methods of sampling.

y Features of sampling.

y Uses of sampling.

Introduction
In carrying out a survey relating to the research, we should
first select the problem and study its implications in different
areas. Selection of the research problem, as has already been
stated, should be in line with the researcher’s interest, chain
of thinking and existing research in the same area and should
have some direct utility. What is most important in selecting
a research problem is that the research topic should be within
manageable limits.

Secondly, the topic should have practical feasibility. To study


feasibility, what is important is to prepare a preliminary
abstract on the research topic. Since this chapter is intended
to acquaint the readers with survey procedure, we are not
concentrating on the aspects of research in great detail which
have in fact already been covered in our earlier discussions.

The first and foremost task in carrying out a survey is to


select the sample. The difference between the population
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
142 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes and sample has already been discussed earlier. Sample


__________________ selection is undertaken for practical impossibility to survey
__________________ the population. By applying rationality in selection of
__________________ samples, we generalize the findings of our research. There
__________________ are different types of sampling. We may categorize those in
__________________ three major heads as follows:
__________________ 1. Random sampling
__________________
2. Purposive sampling
__________________
__________________ 3. Stratified sampling
__________________
Random sampling is not a mere chance selection. Instead, it
ensures inclusion of each and every sample of the population.
The conventional way of selection of samples using random
sampling methods are:

u Lottery method

u Tippett’s number

u Selection from a sequential list

u Use of grid system

Under lottery method, numbers or names of various units of


population are noted on chits and put in a container. After
thorough mixing, chits are drawn from the container and
survey of drawn chits is carried out. Since this method of
random sampling has some amount of chance in it, this is
often described as a back-dated one.

Tippett’s number which lists 10,400 four digit numbers


written at random is constructed out of 41,600 digits taken
from census reports by combining them in to fours. The
method of drawing a sample from Tippett’s number is very
easy. If we want to draw a sample of 20 persons from a list of
6000 persons, for this purpose we shall first number each
unit from 0 to 6000 using Tippett’s four digit codes. Then we
open any page of Tippett’s numbers and select the first 20
numbers that are below 6000. Tippett’s numbers are widely
used in sampling techniques and are found to be quite reliable
in regard to accuracy and representativeness.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 143
Selection of sample from sequential list requires Notes
arrangement of names under the intended plan according to __________________
some order which may be alphabetical, geographical or __________________
simply serial. Thereafter, out of the list, every 10 th or __________________
any other number of cases may be taken up. If every 10th __________________
unit is to be selected, the selection might begin from 7th,
__________________
17th, 27th, 37th, 47th, 57th, etc., or from 5th, 15th, 35th, 45th, 55th,
__________________
etc.
__________________
Grid system is applied for selection of sample from a __________________
particular area. Under this method a map of the entire area __________________
is prepared, then a screen of squares is placed on the map.
__________________
The areas falling within selected squares are taken as
samples.

Purposive selection of samples, as the name goes, depends


more on the researcher’s deliberate choice. Thus, such a
selection of samples, in its true spirit defeats the purpose of
research as the samples suffer from the character of
representativeness.

Stratified sampling combines the characteristics of random


sampling and purposive sampling. Initially the population
is defined in different numbers of strata or groups. Then from
each group certain number of items are taken on a random
basis.

Apart from the above sampling procedures, there are other


types of sampling like:

u Quota sampling (a special type of stratified sampling).

u Multi-stage sampling (where samples are selected from


a very large area).

u Convenience sampling (where population is not clearly


defined and complete source of list is not available).

u Self selected sampling, etc.

After deciding over the samples to be surveyed, the next task


is to go ahead with the survey matter.

Survey may be carried out either by directly interviewing


the samples or by sending questionnaire to the samples
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
144 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes or by mere observation of the characteristics of


__________________ samples.
__________________
__________________ The Sampling Process
__________________ There are seven steps involved in this process:
__________________
Step 1 : Define the population: It is the aggregate of all
__________________
elements, usually defined prior to the selection of the
__________________ sample. The population is said to be completely defined if
__________________ at least the following terms are specified:
__________________
i. Elements
__________________
ii. Sampling units

iii. Extent

iv. Time

For example, for monitoring the sales of our product, the


population might be:

u Element—Our product.

u Sampling units—Retail outlets, super markets.

u Extent—New Delhi, Noida, Faridabad, Gurgaon,


Sonepat (Delhi and around).

u Time—May 1-31st, 1999.

Step 2: Identify the sampling frame: The sampling frame


should be so selected which consists of almost all the
sampling units. Though it is not possible to have one-to-one
correspondence between frame units and sampling units,
however, we should choose a sampling frame which yields
unbiased estimates with a variance as low as possible.
Popularly known sampling frames are:

Census reports, electoral registers, lists of member units of


trade and industry associations, lists of members of
professional bodies, lists of dwelling units maintained by
local bodies, returns from an earlier survey and large scale
maps etc.

Step 3: Specify the sampling unit: The sampling unit is


the basic unit containing the elements of the target
population.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 145
Step 4: Specify the sampling method: The sampling Notes
method indicates, how the sample units are selected. The __________________
most important decision in this regard is to determine which __________________
of the two-probability or non-probability samples is to be __________________
chosen. __________________

Step 5: Determine the sample size (n): The decision about __________________
the number of elements to be chosen, i.e., number of __________________
observations in each sample (n) of the target population. __________________
__________________
Step 6: Specify the sampling plan: This means that one
should indicate, how decisions made so far are to be __________________

implemented. All expected pertinent issues in a sampling __________________


survey must be answered by the sampling plan.

Step 7: Select the sample: This is the final step in the


sampling process. A good deal of field work and office work
is introduced in the actual selection of the sample elements.
However, it depends mainly upon the sampling plan and the
sample size required.

Why Sampling?
There are various advantages of sampling, as given below:

i. The ideal solution to know the true or actual values of


the different parameters of the population would be to
take into account the entire population. However, that
is not feasible due to cost, time, labour and other
constraints, hence sampling is more economical.

ii. As the magnitude of operations involved in a sample


survey is small, both the execution of the field work and
the analysis of results can be carried out much faster
and hence is very less time consuming. On the other
hand, if we have to gather information about the whole
population, in certain cases, we may not be able to have
the information collected timely and the entire study
may become redundant.

iii. In case of destructive testing, for the entire


population, we may not be left with anything after
such an enumeration, i.e., testing the life of a tube or
bulb.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
146 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes iv. Relatively very small staff is required for gathering


__________________ information, analyzing the same and preparation of the
__________________ report.
__________________
v. A researcher can collect more detailed information in
__________________ much less time than otherwise be possible in a census
__________________ survey. Moreover, we can afford to have a few specialists
__________________ for collection of specialized type of information which
__________________ otherwise is not possible for a census study because of
__________________
cost and non-availability of specialists.

__________________ vi. As the scale of operations involved in a sample survey


__________________ is small, the quality of the interviewing, supervision and
other related activities can be better than the quality
in a census survey.

vii. In many cases, sampling provides adequate information


needed for the purpose and sufficiently reliable. Also,
the sampling techniques make it possible to quantify
the magnitude of possible error.

Types of Sampling
There are basically two types of sampling methods:

a. Probability sampling method and

b. Non-probability sampling method.

In case of probability sample method:

i. The probability or chance of every unit in the population


being included in the sample is known.

ii. Selection of the specific units in the sample depends


entirely on chance.

In case of non-probability sampling method:

i. The probability of inclusion of any unit (of population)


in a sample is not known.

ii. The selection of units within a sample involves human


judgement rather than pure chance.

iii. The maximum information available “per rupee” which


can be determined from a probability sample is not
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 147
possible in this case and moreover, the degree of Notes
accuracy is not known. __________________
__________________
iv. Although probability sampling is scientific and accurate,
__________________
however, because of convenience and economy, the non-
probability samples are preferred. __________________
__________________
v. Many times, samples are selected by interviewers
__________________
“at random” meaning that the actual sample selection
__________________
is left to the choice of the researcher, such sample
__________________
are non-probability samples and not probability
samples. __________________
__________________
Probability Sampling Methods
The major sampling methods under probability sampling
are:

u Simple random sampling

u Systematic sampling

u Stratified sampling

u Cluster sampling

Simple Random Sampling


Simple Random Sampling is the simplest type of sampling,
in which we draw a sample of size (n) in such a way that each
of the ‘N’ members of the population has the same chance of
being included in the sample. A sample selected in this way
is called a simple random sample.

Selection of Random Samples


One way of drawing a simple random sample is to number
every individual, put the numbers on slips of paper and draw
lots. But such a procedure is impracticable in most situations.
A more convenient method is to use a table of random digits.
Such tables contain lists of digits so chosen that each digit
between 0 to 9 has an equal chance of appearing at a given
spot in a single column and each two digits between 00 to 99
has the same chance of appearing at a given spot in a double
column and so on. Printed tables of random numbers are used
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
148 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes in practice. For example, if we want to select ten items from


__________________ the item produced during a sift from an automatic machine,
__________________ which normally produces 500 items in a shift, we need
__________________ to select 10 random numbers between 001 and 500.
__________________ The selected 10 numbers arranged in ascending order will
give the serial number of the items to be included in the
__________________
sample.
__________________
__________________ We number the individuals from 000 to 499 and proceed to
__________________ read down same three digits column in a random number
__________________
table:
__________________ 135ü 283ü

369ü 476ü
387ü 761û
507û 837û

457ü 116ü
077ü 232û
977û 505û

999û 636û
421ü

We continue to write down three-digit numbers until we


have ten different numbers lying between 000 to 499. Here,
our simple random sample includes: 135, 369, 387, 457, 077,
283, 476, 116, 232, 421

Now put this in ascending order:

077, 116, 135, 232, 283, 369, 387, 421, 457, and 476 (10 Nos.)

0 77

1 16, 35

2 32, 83

3 69, 87

4 21, 57, 76
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 149
Simple Random Sample in Practice Notes
__________________
The use of random sampling is made by researchers for the
__________________
following situations:
__________________
u When a small sample is needed from the list of sample __________________
frame (list of universe items). __________________

u The cost per interview is practically independent of the __________________

location of the sample item. __________________


__________________
u Other than a list of items, no other information is
__________________
available.
__________________
u In simple random sampling, the sample mean
provides an unbiased estimate of the universe mean.
However, the use is severely limited by the following
factors:

– Cost: Sampling frame is required for a simple


random sample. In most situations, it is very
difficult to have a frame, if not impossible and
moreover it is very time consuming and hence is
uneconomical also.

– Statistical efficiency: One sample designing is said


to be statistically more efficient than another
when the frame is of a small size and a smaller
standard error is obtained. Most large populations
are not homogeneous but can be broken down into
more homogeneous units.

In such conditions, one can have sampling design


such as stratified sampling which is statistically
more efficient. Similarly, the use of cluster sampling
whenever we can pick up members from
geographically closer areas, reduces the cost
involved.

– Supervision: The problem of selecting a sample with


the help of random numbers etc., though appears to
be very simple but in practice it is not so. Suppose,
we want to select a sample size of 2,5000 from a list
of 2 million, it's a difficult job and errors are bound
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
150 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes to be there. Moreover, the cost of supervision is also


__________________ high. As compared to this, systematic random
__________________ sampling is much easier.
__________________
Systematic Sampling
__________________
__________________ The systematic sampling also employs the principle of
__________________ random sampling. However, in this method of sample,
__________________
selection of a unit depends upon the selection of a preceding
unit in contrast to simple random sampling, where the
__________________
selection of a unit is independent of each other. Systematic
__________________
random sampling in this sense is called quasi-random-
__________________
sampling.

Systematic Sampling in Practice


The method followed in systematic sampling is simpler as
compared to simple random sampling.

For selecting a sample size of 50 out of 500 the procedure is


as follows:

Step I Find simple fraction K = N/x

When N is total of units of population and x


is the sample size.

500
= 10 in this case
50

Step II Select a number between 1 and 10, at


random, say 9.

Step III Then, the sample comprises of nos.

9, 19, 29, 39, ...............................489, 499

Only the first unit has to be selected and the selection of


remaining numbers is easy.

Limitations
If the sample frame has any periodicity that parallels the
sampling rates, then the possibility of a strong bias is there.
For example, in banking transactions if you choose 1/7, 1/14,
etc., then you will be having same day of the week for your
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 151
study and hence your study will be a biased study. Be it, Notes
Monday or Wednesday or Friday etc. __________________
__________________
Similarly, if the frame is arranged in a way-ascending or
__________________
descending order of some attribute, then the location of the
first sample element may affect the result of the study. In __________________

such cases, one should devise ways to qualify the effect of __________________
bias. __________________
__________________
On the other hand, if the frame is so arranged that similar
__________________
elements are grouped together, then systematic sampling
produces almost a proportional stratified sample. __________________

These samples are statistically more efficient as compared __________________


to simple random sampling.

For estimating the precision of the results, systematic


samples are treated as simple random sample and perhaps
is the most widely used method among the probability
sampling designs.

Stratified Random Sampling


Another useful type of sampling procedure is called
stratified random sampling. In this procedure, the members
of the population are first assigned to strata or groups, on
the basis of some characteristic and a simple random sample
is drawn from each stratum. The individuals in all the
samples taken together constitute the sample from the
population as a whole, viz.
Stratum Number in stratum Number in sample

1 N1 n1

2 N2 n2

3 N3 n3

4 N4 n4

5 N5 n5

6 N6 n6

7 N7 n7

k Nk nk

Total N n
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
152 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes If the component sample sizes n1, n2,.................nk are so chosen


__________________ that,
__________________
n1 n n (n)
__________________ = = 2 =- - - - - - - - - - - - - k
N1 N 2 N (N)
__________________ k
__________________
In this way we do proportional stratified random
__________________ sampling.
__________________
Stratification does not mean absence of randomness. All it
__________________
means is that the population is first divided into certain
__________________
strata that are mutually exclusive and collectively
__________________
exhaustive.

Populations divided into three (3) strata s1, s2, s3.

A stratum, as is clear, is a sub-population which is more


homogenous than the complete population. The members of
a stratum are similar to each other.

Example 6.1
Suppose, the population consists of 45,000 households,
divided into five (5) strata on the basis of monthly income.
This can be illustrating as below:
0 – 1000
1001 – 5000
5001 – 7500
7501 – 10,000
Above 10,000
Then,
a. Find out the number of units from each strata if the
sample constitutes 1% of the population.
b. If selection is for 150 items selecting equally from each
strata, find out the number of sample units from each
strata.

Solution
a. Proportional stratified sampling
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 153
Stratum No. Population (No of Sample Sampling Notes
(i) households) (proportionate) ratio __________________
(ii) (iii) 1%=.01
__________________
0-1000 15,000 150 .01 __________________
1000-5000 12,000 120 .01 __________________
5001-7500 9,000 90 0.01
__________________
7501-10,000 6,000 60 .01
__________________
> 10,000 3,000 30 .01
__________________
45,000 450 .01
__________________
__________________
b. Equal from each strata:
__________________
Total No. of sample units = 150
No. of sample units from each stratum = 150/5
= 30
Sampling ratio = Sample size / Population size

Major Issues
There are three major issues involved in stratified sampling:

i. Bases of stratification: Depends upon the variable


being studied. Quite often it is desired to have more
than one variable. In marketing research, usually
demographic characteristics such as age, sex or income
and geographical distribution such as rural-urban
break-up, break-up by region, state or city form the basis.

ii. Number of strata: Since stratification would enhance


the cost of the survey, one would weigh the benefits
resulting from it vs. the cost involved. As a rule of thumb,
not more than 6 strata should be used for a single overall
estimate.

iii. Sample sizes within strata: Sample size within strata


depends upon the budget and the cost per observation.
Suppose in the above example, if the budget is Rs 50,000
and the cost per observation is Rs 250, then the sampling
size = 50,000/250 = 200 and not 450. However, this
should be allocated among the various strata.
Researchers can use either a proportional or
disproportional sample.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
154 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Proportional Allocation


__________________
Under this method each stratum is related to its weight. In
__________________
the Example 6.1, based on relative weight, if we have to have
__________________
a sample size of 225 then numbers from each stratum would
__________________ be as follows:
__________________ Stratum No of Household For Sample Size n = 225 Sampling Ratio
__________________
0-1000 15,000 75 225/45000
__________________
1001-5000 12,000 60
__________________
5001-7500 9,000 45
__________________
7501-10,000 6,000 30
__________________
10,001 and 3,000 15

above

45,000 225

Disproportional Allocation
There are circumstances where some form of disproportional
allocation should be considered, i.e., sampling different
strata at different rates. As a general principle, when the
variability among observations within a stratum is high,
‘higher rate for higher internal variation strata and lower
rate for less internal variation strata’ is adopted.

If s12 , s22 ,...................sp2 are the variances of the ‘p’ strata


respectively, then the statistical efficiency is highest, when

n1 n2 np
= =
N1s1 N 2s 2 N ps p

Example 6.2
Let us consider a case of 3 strata, of income group with given
stratum variance.
Stratum No. of Households Stratum Variance
0 - 5000 300 4.00
5001-10,000 450 9.00
> 10,000 750 2.25
Total 1500

Find out the nos. from each stratum for a given sample size
of 50?
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 155
Solution Notes
__________________
Disproportional stratified sampling
__________________
Stratum No No of Strata Stratum Sample Sampling
__________________
(i) elements/ variance standard size ratio
households deviation (n i ) (n i /N) __________________
0 - 5000 300 4.00 2.0 10 0.033 __________________
5001-10000 450 9.00 3.0 22 0.049
__________________
> 10,000 750 2.25 1.5 18 0.024
__________________
Total 1500 50
__________________
n1s1 + n2s2+ n3s3 = (300´2.0) + (450´3.0) + (750´1.5) __________________
__________________
= 600 + 1350 + 1125 = 3075
50
\ n1 = × 600
3075
50
n2 = × 1350 = 22
3075
50
n3 = × 1125 = 18
3075

Stratified Sampling in Practice


The main reasons for using stratified sampling for managerial
applications are:

i. It can obtain information about different parts of the


universe, i.e., it allows to draw separate conclusion for
each stratum.

ii. It often provides universe estimates of greater precision


than other methods of random sampling say simple
random sampling.

However, the price paid for these advantages is high because


of the complexity of design and analysis.

Estimation – statistical Inference


As it is not possible to analyze the entire population and
then draw the conclusion because of limited manpower,
money and time factors, we have to draw samples from the
population and draw conclusions about the same based upon
the sample result. We draw a sample at ‘random’ of a given
size (say 'n') and compute a statistic – a characteristic of the
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
156 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes sample ( X , S) and this becomes an estimate of the similar


__________________ characteristic of the population (m, s).
__________________
There are two types of estimates:
__________________
__________________ u Point estimates
__________________ u Interval estimates.
__________________
__________________ Point Estimation
__________________ When we estimate the population parameter (say m or s),
__________________ we take the help of a simple sample value (m in case of m
__________________ and s in case of s ). This is called point estimate as we use
only a single value. The point estimator would be the best
estimator if it has the following form of characterization:

u Unbiasedness

u Consistency

u Efficiency

u Sufficiency

An estimator shall be unbiased, if the expected value of the


sample average equates the population mean : Sx= m

An estimator is consistent estimator, if the statistic


parameter approaches the population as the sample size
increases.

As the stratified random sample is a group of simple random


samples, the estimate of the overall universe mean is simply
a weighted average of the stratified sample mean.

Example 6.3
Suppose, we have the following data on consumption of
sample households.
Income strata (pm) Sample mean No of observation
purchase per in each stratum
households (Rs)
1 – 5,000 1000 5000
5001 – 10,000 2500 1500
>10,000 3500 500
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 157
Estimated Population Mean (Mean Monthly Expenditure) Notes

5000 __________________
‘X = Â Wi x i [I= 1, = 0.714] __________________
7000
__________________
1500 __________________
= .714(1000)+.214(2500)+.072(3500) [I= 2, ]= 0.214
7000 __________________
__________________
= 714 + 535 + 252
__________________
5000 __________________
= Rs 1501 [I= 3, ]= 0.072
7000 __________________
__________________
Estimation of Confidence Interval with Stratified Random Sampling
For finding the confidence interval, we need the following
data on sample: variance, sample size and weight for each
stratum.

Let us discuss the following example.


Stratum (i) Sample variance (si2) Sample size Weight of each
in stratum stratum
(n 1 ) ( Wi2 )

1. (Rich) 3000 60 0.1

2. (Middle class) 2000 100 0.3

3. (Poor) 1000 250 0.6

Estimation of standard error is given by:

2 2
Wi Si
S.E. = ∑ ni

W12S12 W22S22 W12S12


= ∑ n1
+
n2
+
n3

(.01)(3000) (.3)(2000) (.06)(1000)


= + +
60 100 250

= 0.50 + 6.0 + 2.4

= 8.9 = 2.98
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
158 Research Methodology fo
n ~;
k ; k ‘ kf D r o’
k~ e

Notes Let us examine with 95% confidence interval,


__________________
value of Z = ± 1.96
__________________
__________________ At an average monthly consumption of say Rs 1500, the values
__________________ will come out to be between
__________________ ‘ X ± 3s` = Rs 1500 ± 3 (2.98)
__________________
= Rs 1500 ± 8.94
__________________
__________________ = Rs 1508.94 and Rs 1491.06
__________________ say Rs 1509 and Rs 1491
__________________
It may be noted that in the above calculations, differences
among strata means has no where been figured and this is
the reason why stratified sampling gives a more
precise estimate of the population mean than a simple
random sampling for a given sample size.

Cluster Sampling
In the probability sampling methods, we have seen that each
item in the sample is chosen one at a time from the complete
list of universe elements. However, it would be more
expedient to select entire groups or clusters at random. Let
us take, for example, a residential colony comprising 15
Blocks : A to O.

Let us treat each block as a cluster and then select 3 (say)


blocks (cluster) out of 15 blocks at random and then collect
information from all families residing in these 3 blocks
(clusters).

Cluster Sampling in Practice


This method is usually convenient for collection of data as a
cluster is ideally a mini-population and has all the features
of the population. Clusters are heterogenous within
themselves and not like the homogeneous strata. Hence,
collection of data would be far easy as compared to other
method. Moreover, apart from cost considerations, such a
cluster sample would be desirable in the absence of a suitable
sampling frame. Frames, for cluster sampling, are
needed for the selected cluster only and hence reduce the
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 159
cost of developing a frame as compared for simple random Notes
sampling or stratified random sampling. __________________
__________________
Use of Cluster Sampling in a National Survey
__________________
In a national survey, first we’ll select a few districts in the __________________
whole country, (these districts would act as cluster and then __________________
we would take stratified sampling/simple random sampling __________________
so that the survey can be completed with less cost and more __________________
accuracy.
__________________
Area sampling is a special case of cluster sampling. __________________
The question now is how to select a sample in cluster __________________
sampling?

Following are the number of warehouses in 15 cities:


40,22,15, 33, 75, 30, 21, 24, 57, 32, 63, 27, 46, 51,64

If we want to select a sample of 15 blocks using cities as


clusters and then select within clusters proportionate to size,
how many stores from each city should be chosen (use 15 as
the start point)?
City # No of warehouse Cumulative Sample Remarks

1 40 40 15 One sample

2 22 62 55 One sample

3 15 77 Nil

4 33 110 95 One

5 75 185 135 One

6 30 215 175, 215 Two

7 21 236 – Nil

8 24 260 255 One

9 57 317 295 One

10 32 349 335 One

11 63 412 380 One

12 27 439 420 One

13 46 485 460 One

14 51 536 500 One

15 64 600 540, 580 Two

15

600 ¸ 15 = 40
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
160 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Multistage and Multiphase Sampling


__________________
Multistage sampling, as the name implies, means that the
__________________
selection of units is done in more than one stages. The
__________________
number of stages in a multistage sampling is based on
__________________ convenience and the availability of suitable sampling frames
__________________ at different stages. In case of a national survey, it is easier
__________________ to understand that this can involve the following four stages:
__________________
I stage Districts
__________________
__________________ II stage Cities
__________________ III stage Wards or localities

IV stage Households

Example 6.4
Suppose, we want to have 7500 households from all over
the country. In such a case:

I Stage – District: say 30 districts out of 600 are selected


from all over the country;

II Stage – Cities: Suppose 5 cities are selected out of each 30


districts; and

III Stage – Wards/localities: say 10 wards/localities are


selected from each city;

IV Stage – Households: 50 households are selected from


each ward/locality.

In stage I, we can employ stratified sampling.

In stage II, we can use cluster sampling.

In stage III, we can have simple random sampling.

Thus, the use of various methods shall give individually


contribute towards accuracy, cost, time, etc. This leads us to
conclude that multistage sampling leads to saving of time,
labour and money. Apart from this wherever an appropriate
frame is not available, the use of multistage sampling has
universal appeal.

[Exercise : use of fertilizers in national survey]


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 161
Stratified Sampling : Districts as a sampling unit, average Notes
rainfall in the districts as a criterion for stratification, 20 __________________
districts). __________________
__________________
Cluster Sampling: It may be used to select 10 villages in each
of the selected districts. __________________
__________________
Simple Random Sampling: It can be used for selection of 50
__________________
samples from each village, criterion of stratification being
__________________
the holding of land.
__________________
Non-probability Sampling Methods __________________
__________________
Though the probability samples give an unbiased sample and
the parameters used for the study can be tested for a given
confidence internally, still non-probability sampling
finds frequent use in many situations because of difficult
conditions such as frame (list of all sampling units), time
and cost involved. Major non-probability sampling methods
are discussed below:

u Convenience sampling: As the name implies, the


selection of the sample is left to the researcher who is
to select the sample. The researcher normally
interviews persons in groups at some retail outlet,
supermarket or may stand at a prominent point
and interview the persons who happen to be there. This
type of sampling is also called ‘accidental sampling’ as
the respondents in the sample are included merely
because of their presence on the spot. The data
collection and sample cost is minimum in this case.
However, the method suffers greatly from the quality,
i.e., accuracy point of view which can in no way be
determined. However, this type of sampling is more
suitable in ‘exploratory research’ where focus is on
getting new ideas/insights into a given problem.

u Judgement sampling: In judgement sampling, the


judgement or opinion of some experts forms the basis of
the sampling method. It is expected that these samples
would be better as the experts are supposed to know
the population. However, as the use of randomness is
not there and moreover their is no way to find the
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
162 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes accuracy of the samples, hence the method has its


__________________ limitations and is used mainly for situations requiring
__________________ extremely small size of samples, i.e., use of rare events,
__________________ members having extreme positions, etc.
__________________ u Quota sampling: This is the most frequently used non-
__________________ probability sampling method and is used when:
__________________
– Employing stratification: i.e., age, sex, income,
__________________
family size etc. More often compound
__________________
stratification is used in groups with sex.
__________________
Stratum
__________________
Men 21 + 35
36 + 45
Women 21 + 35
36 + 45

– Sample size, (called quotas) are established for


each stratum. This is done to reduce ‘bias’
introduced by sampling.

Let us understand with the help of Example 6.5.

Example 6.5
We have population made up of rich, medium and poor
people in the ratio of 10% , 40%, 50%, respectively. Using
quota sampling, determine the number of representatives
from each category in a sample size of 100.

Category I (rich) = 10% of 100 = 10


Category II (medium) = 40% of 100 = 40
Category III (poor) = 50% of 100 = 50
100

Quota Sample in Practice


u It is economical as travelling costs are reduced.

u It is easy to administer.

u When field work is to be done quickly, this method


provides the biggest advantage.

u No sampling frame is required.


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 163
However, since the method is not based on random selection, Notes
it is not possible to ascertain the accuracy being achieved. __________________
Moreover, the sample may not be a representative sample. __________________
Also, the quality may suffer if skilled interviewers are not __________________
employed. __________________

In many real life situations, the use of cluster sampling __________________


method and that of convenience sampling is being made __________________
mainly because the cost per observation is much use as __________________
compared to other method. Sophisticated research __________________
techniques and methods ensuring greater accuracy are not __________________
finding much use.
__________________

Characteristics of a Good Sample Design


u Goal orientation: Sample design should be oriented
to the research objectives.
u Measurability: The parameters under study should be
some how or the other measurable so that accuracy can
be ensured.
u Usability: The sample size should be convenient to
collect and analyze.
u Cost factor: The total cost of sample design, collection
and analysis of the data should be minimum.

Thus one has to weigh the pros and cons of various sample
designs before selecting the best possible one.

Sampling and Non-sampling Errors


The basic objective of a sample is to draw inferences about
the population from which the sample is drawn. Therefore,
it is necessary that the sampling technique be a reliable
one.

The randomness of the sample is specially important because


of the principle of statistical regularity which states that a
sample taken from a population is likely to possess almost
the same characteristics as those of the population.

In the total process starting from data collection to inferring


results, errors are bound to come in. These errors can be
classified into two groups.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
164 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Sampling Errors


__________________
Sampling errors are those which arise due to drawing of
__________________
faulty inferences about the population based on the results
__________________
obtained from the samples. In other words, it is the
__________________ difference between the results which would have obtained
__________________ if the entire population was taken for such a study and the
__________________ results obtained from the samples drawn from it. The
__________________ sampling error would be smallest if the sample size is large
__________________ in relation to the population and vice versa.
__________________
Non-sampling Errors
__________________
Non-sampling errors are introduced due to technically faulty
observations or calculation during the processing of the data.
The faulty methods could be used at different stages:

u Methods of data collection.


u Incomplete coverage of the population.
u Inaccurate information provided by the participants.

u Errors occurring during editing, tabulating and


mathematical manipulation of data.
u These errors shall arise even if the entire population is
taken understudy.

Both the sampling as well as non-sampling errors must be


reduced to a minimum in order to get a representative
sample of the population as far as possible.

Some Problems on Sampling Fundamentals

Problem 6.1
From a random sample of 49 colleges of Delhi, the average
expenditure and its standard deviation were found to be Rs
2000 and 450 respectively Construct at 95 percent confidence
interval average expenditure of a student in Delhi.

Solution
Here, we have n = 49
x = 2000
Sx = 450
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 165
value of ‘z’ at 95% confidence level = 1.96 (from normal table) Notes
__________________
Sx
Now z = __________________
SE __________________
__________________
Sx 450
SE = = __________________
n 49
__________________
__________________
450
= __________________
7
__________________
ì = x ± Z.(SE) __________________

(450)
= 2000 ± 1.96
7

= 2000 ± 126
= Rs 1874 and 2126

Problem 6.2
What should be the sample size for the following case :

Standard deviation being Rs 200 and the estimate within +


Rs 40 ?

Solution

Here, e = / 40 /

Let the confidence limit be 95%


Then z = 1.96 (from same table)
z(ó ) z(ó )
New = l=
x n
σ σ
Where, =
χ η

200
So, 40 = 1.96
n

n =1.96 × 5 = 9.80

n = 96
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
166 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Standard deviation being Rs 100 and the estimate within +


__________________ Rs 50.
__________________
100
__________________ 50 = 1.96
n
__________________
__________________
1.96 ×100
__________________ n= = 3.82
50
__________________
__________________
n=15
__________________
Problem 6.3
__________________
For the following data, calculate the standard error for mean:

Number of motor cycle accidents is 4.1 in a random


examination of 8 cases out of 2500 with standard deviation
being 0.9.

Solution
ó N −n
Standard error of Mean= .
n N −1
When s = std deviation of population = sx = 0.9

N = finite size of population

n = sample size

0.9 2500 − 81
óx =
81 2500 −1

0.9 2419
=
9 2499
0.9
= × .9679
9
= 0.1 ´ 0.9839

= .09839

Problem 6.4
For problem 6.3, find the value of mean number of accidents
for a 90% confidence interest.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 6 Sampling Procedures 167
Find the standard error of mean if the population size is Notes
“not finite.” __________________
__________________
Solution
__________________
1. x + z (l x ) __________________
__________________
ó
= x +1.645 (z again 90% confidence interval = 1.645) __________________
n
__________________
0.9 __________________
= 4.1 + 1.645 .
81 __________________

.9 __________________
= 4.1 +1.645 ×
9
= 4.1 + 1.645 (0.1)
= 4.1 + .1645
= 4.264 & 3.9355
2. Standard error of mean, when population is “not finite”
ó 0.9 0.9 0.1
lx = = = =
n 81 9 3

Problems 6.5
The average production of Alloy Steel Plant, Durgapur, for
10 shifts taken at random is found to be 100 tones per shift
with sample standard deviation of 3.2 tones per shift.
Construct a 90% confidence interval around the estimate.

Solution
Here, x = 100 tones per shift
sx ¯ = 3.2 tones per shift
n = 10
As, n < 30, we shall apply ‘t’ test (and not z test),
with degrees of freedom = n – 1 = 10 – 1 = 9
Confidence interval for 90%
óx
= x ± t,
n
3.2∂
= 100 ± (2.353)
10
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
168 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
3.2
__________________ = 100 ± (2.353).
3.16
__________________
= 100 + 2.383
__________________
= 97.617 & 102.383
__________________
__________________ Review Questions
__________________
1. Why is sampling used in marketing research?
__________________
__________________ 2. Define the following terms: population, sampling unit,
__________________ sampling procedure in a particular marketing research
__________________ study.

3. What are the major sources of non-sampling errors in a


marketing study?

4. Discuss briefly the following sampling techniques and


the merits/demerits of each:

i. Simple random sampling

ii. Systematic random sampling

iii. Stratified random sampling

iv. Cluster sampling.

5. What is a sample size?

6. What features would you consider in designing a sample


size?

7. What are parametric & non-parametric tests? Where


can each be used more advantageously than the other?
169

Unit 7 Notes
__________________
__________________
Data Collection Methods __________________
__________________
__________________
__________________
__________________
__________________
__________________
Objectives
__________________
After reading this unit you will be able to understand:

y Types of data.

y Procedure of data collection–Primary data.

y Procedure of data collection–Secondary data.

Types of Data
Once the researcher has decided the ‘research design’, the
next job is of data collection. For data to be useful, our
observations need to be organised so that we can get some
patterns and come to logical conclusions.

Statistical investigation requires systematic collection of


data, so that all relevant groups are represented in the data.

To determine the potential market for a new product, for


example, the researcher might study 500 consumers in a
certain geographical area. It must be ascertained that the
group contains people representing variables such as income
level, race, education and neighbourhood. The quality of data
will greatly affect the conclusions and hence, utmost
importance must be given to this process and every possible
precaution should be taken to ensure accuracy, while
gathering and collecting data.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
170 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Depending upon the sources utilised, whether the data has
__________________ come from actual observations or from records that are kept
for normal purposes, statistical data can be classified into
__________________
two categories. primary and secondary:
__________________
__________________ Primary Data
__________________
Primary data is one which is collected by the investigator
__________________
himself for the purpose of a specific inquiry or study. Such
__________________ data is original in character and is generated by surveys
__________________ conducted by individuals or research institutions.
__________________
__________________
Secondary Data
When an investigator uses the data which has already been
collected by others, such data is called secondary data. This
data is primary data for the agency that collects it and
becomes secondary data for someone else who uses this data
for his own purposes. The secondary data can be obtained
from journals, reports, government publications, publication
of professional and research organizations and so on. For
example, if a researcher desires to analyze the weather
conditions of different regions, he can get the required
information or data from the records of the meteorology
department.

Table 7.1: Distinction between Primary Data and


Secondary Data

Description Primary Data Secondary Data

1. Source Original source Secondary source

2. Methods of data collection Observation method Published data of


Questionnaire method government agencies
Trade journal etc.

3. Statistical process Not done Done

4. Originality of data Original first time No data are collected by


collected by user some other agency

5. Use of data For specific purpose Data are taken from


data are compiled other sources and used
for decision making

6. Terms and definitions of units Incorporated Not included

7. Copy of the schedule Included Excluded

Contd...
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 171
8. Methods of data collection Given Not given Notes
__________________
9. Description of sample selection Given Not given
__________________
10. Time More Less
__________________
11. Cost Expensive Cheaper
__________________
12. Efforts More Less
__________________
13. Accuracy More accurate Less accurate __________________

14. Training personnel required Experts/trained required Less trained personnel


__________________
__________________
Data Collection Procedure for Primary Data __________________
__________________
Planning the Study
Since the quality of results gained from statistical data
depends upon the quality of information collected, it is
important that a sound investigative process be established
to ensure that the data is highly representative and unbiased.
This requires a high degree of skill and also certain
precautionary measures may have to be taken.

Modes of Data Collection


There are basically three widely used methods for collection
of primary data:

u Observation

u Experimentation

u Questionnaire

u Interviewing

u Case study method

Observation Process
Information is collected by observing the process at work.
The following are a few examples:

i. Service stations: Pose as a customer, go to a service


station and observe.

ii. To evaluate the effectiveness of display of dunlopillo


cushions in a departmental store, observer notes.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
172 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes a. How many pass by?


__________________
b. How many stopped to look at the display?
__________________
c. How many decide to buy?
__________________
__________________ iii. Super market: What is the best location in the shelf?
__________________ Hidden cameras are used.
__________________
iv. Concealed tape recorder with the investigator helps to
__________________ determine typical sales arguments and find out sales
__________________ enthusiasm shown by various salesmen.
__________________
By this method, response bias is eliminated.
__________________
The method can be used to study sales techniques, customer
movements, customer response, etc. However, the
customer’s/ consumer’s state of mind, their buying motives,
their images are not revealed. Their income and education
is also not known. It also takes time for the investigator to
wait for particular sections to take place.

Experimentation Method
Many of the important decisions facing the marketing
executive cannot be settled by secondary research,
observation or by surveying the opinions of customers or
experts. Experimental method may be used in the following
situations:

i. What is the best method for training salesmen?

ii. What is the best remuneration plan for salesman?

iii. What is the best shelf arrangement for displaying a


product?

iv. What is the effectiveness of a point-of-purchase display?

v. What package design should be used?

vi. Which copy is the most effective?

vii. What media are the most effective?

viii. Which version of a product would consumers like


best?
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 173
In a marketing experiment, the experimental units may be Notes
consumers, stores, sales territories, etc. __________________
__________________
Factors or marketing variables under the control of the
__________________
researcher which can be studied are price, packaging,
display, sales incentive plan, flavour, colour, shape, etc. __________________
__________________
Competitor’s actions, weather changes, in cooperative
__________________
dealers, etc. are environmental factors.
__________________
To study the effect of the marketing variables in the presence __________________
of environmental factors, a sufficiently large sample __________________
should be used. Or sometimes a control group is set up. __________________
A control group is a group equivalent to the experimental
group and differing only in not receiving any
treatment.

The result/response of a marketing experiment will be in


the form of sales, attitudes or behaviour.

Questionnaire Technique
The survey method is the technique of gathering data by
asking questions from people who are thought to have the
desired information.

Advantages
One cannot know by observation, why a buyer makes
particular purchases or what is his opinion about a product.
Compared with either direct observation or experimentation,
surveys yield a broader range of information and are effective
for producing information on socio-economic characteristics,
attitudes, opinions, motives, etc., and to gather information
for planning product features, advertising copy, advertising
media, sales promotions, channels of distribution and other
marketing variables. Questioning is usually faster and
cheaper than observation.

Limitations
a. Unwillingness of respondents to provide information:
This requires salesmanship on the part of the
interviewer. The interviewer may assure that the
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
174 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes information will be kept secret. Motivating respondents


__________________ with some token gifts often yield result.
__________________
b. Inability of the respondents to provide information: This
__________________
may be due to:
__________________
__________________ i. Lack of knowledge.
__________________ ii. Lapse of memory.
__________________
iii. Inability to identify their motives and provide
__________________
‘reasons why’ for their actions.
__________________
__________________ c. Human biases of the respondents: i.e., ego, etc.

d. Semantic difficulties: It is difficult, if not impossible, to


state a given question in such a way that it will mean
exactly the same thing to every respondent. Similarly,
two different wordings of the same question will
frequently generate quite different results. These
limitations can be controlled to some extent by:

i. Careful phrasing of questions.

ii. Careful control of data gathering by employing


specially trained investigators who will observe
carefully and report on the subtle reactions of
persons interviewed.

iii. Cautious interpretation by a clear recognition of


the limitations of the data and an understanding
of what exactly the data represents. This is
especially true of responses to questions like:

What price would you be willing to pay for this


product?

iv. Looking at facts in relative rather than absolute


terms. A survey showed that 60% of families in
the middle income group used toothpaste. Taken
by itself in the absolute sense, the results of the
survey are in some doubt because the question
asked encountered an obvious bias. But if this 60%
is looked at on a relative basis, viz the
corresponding figure of 60% for upper income
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 175
group families, a more meaningful and significant Notes
interpretation can be made, even though the __________________
individual figure for each group may be slightly __________________
inflated. __________________
__________________
Interviewing
__________________
Interview on samples may be carried out either with a __________________
structured framework or with an undirected approach. The __________________
structured framework involves use of some pre-
__________________
determined questions. Such pre-determination enables the
__________________
researcher to standardize the responses with some fixed
__________________
alternatives. The samples here are merely directed to
choose answers/responses from different pre-determined
alternatives. Thus the researcher can or may quantify the
responses in line with his research object. Standardizing the
responses with pre-determination involves great amount of
risk unless the researcher acquaints himself with the
intricacies of the research matter in much greater details.
However, this approach is more scientific in nature for its
feasibility of quantifications with least trouble and
application of scientific techniques with more rationality.

Unstructured or undirected interview approach enables the


respondents or the samples to answer the researcher’s
queries with greater amount of flexibility. Since no pre-
determined responses here are advised, the researcher may
proceed, keeping in tune with the research matter, with
greater amount of flexibility too. However, quantification of
the responses from unstructured interviews are difficult
unless the researcher fixes the standard of all responses with
some amount of control. If sample characteristics go on rising,
enumeration become difficult. Thus unstructured approach
may defeat the purpose and object of research. This approach
is resorted to usually in cases where the selected samples
need to be interviewed in a more intensive way.

There are different types of unstructured interviews:

1. Focussed interview, which is directed to focus the


attention of respondents to give some experience and
its effects.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
176 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 2. Clinical interview, which is somewhat similar


__________________ to focussed interview but enables the samples
__________________ to underlie their feelings or motivations in much
__________________ broader perspectives. This method is usually
__________________ administered in psychiatric clinics and in prison
administration.
__________________
__________________ 3. The third method of unstructured interview is non-
__________________
directive approach. Under this approach the initiative
is left completely in the hands of the respondents.
__________________ Psychoanalytic research is usually done with a non-
__________________ directive approach.
__________________
Interviewing the subjects or the samples is more
advantageous than sending questionnaires through mail.
Interview method enables the researcher to personally feel
the problems of samples. Moreover interviewer/researcher,
being present on the spot, case study certain qualitative
variables like facial expressions and gestures of me samples.
For high reliability and feasibility of scoring using test
devices, interview approach is more scientific than mailing
questionnaire.

Structured Study
If a radio manufacturer wanted to find out how many people
own a radio, what type it is, when they bought it, the
respondents could be asked a set of questions in the following
given sequence.
Does your family own a radio? Yes/No
(If yes, ask)
What brand is it? Number____________
How many valves? Number____________
When did you purchase this radio? Date________________
This is an example of structured and non disguised study.

Non-structured Studies
More than anything else marketing men want to know why
people buy or don’t buy their products.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 177
Reasons for why can be classified as: Notes
__________________
1. Those reasons which are a part of the individual own
__________________
purposes or attitudes.
__________________
2. Those reasons which are the result of outside influences __________________
such as advertising. __________________

3. Those reasons which are based on characteristics of the __________________


product itself. __________________
__________________
But questions will have to be aimed at these three categories
__________________
separately, which makes the approach satisfactory.
__________________
Many people will not report motives which might be
considered biased or socially unacceptable.

To overcome these difficulties, techniques have been


developed by psychoanalysts.

Depth Interview (Non-disguised)


Instead of approaching the respondent with a fixed list of
questions, the interviewer attempts to get the respondent
to talk freely about the subject of interest. By doing so the
interviewer hopes to get the respondent at ease and then
encourage him to express any ideas which he has on the
subject. If some idea of interest is passed over too quickly,
the interviewer may seek more information by “probing”. For
example, he may comment “that is interesting. Why do you
feel that way?” This encourages further discussion or the
point. Various probes can be used to get the respondent to
expand on any particular ideas. Although no formal
questionnaire is used in interviewing of this type, the
interviewer has an outline in mind. If the respondent does
not get into areas of special interest, the interviewer will
insert questions opening up these topics. The objective of
these interviews is to get below the respondent’s surface
reasons for particular marketing decisions, and to find the
underlying or basic motives.

Interviewer should have background of social psychology and


field experience of 500 or more interviews.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
178 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Sometimes, a group of 6 to 8 people are called for a discussion


__________________ with the interviewer acting as a moderator.
__________________
Projective Technique (Disguised study)
__________________
__________________ Respondent is given an ambiguous situation and asked to
__________________ describe it. The description given contains a projection of
__________________ the respondents’ personality and attitudes to the situation
__________________
described.
__________________ Various projective techniques are used but the most
__________________ commonly used are word association, sentence completion
__________________ and story telling.

In word association, a series of words is read one at a time


to the respondent. After each word, the respondent says the
first thing that comes into his mind. Sentence completion
requires the respondent to complete partial sentences. In
story telling the respondent is shown a picture or given a
description and asked to tell a story about it.

Major Steps in Conducting a Survey


Deciding on the Research Objectives
Every effort should be made to state the objectives in specific
terms. Surveys, in particular, can proceed in an almost
unlimited number of directions. To prevent all kinds of
questions from being asked, clear informational objectives
should be developed and put in writing, if possible.

Methods of Collection of Data


Following methods are in use for collection of data for
questionnaire technique:
a. Telephone enquiries.
b. Postal or mail questionnaire.
c. Personal interviewing.
d. Panel research.
e. Group interview technique.
f. Special survey techniques.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 179
Each of this method has its own advantages and Notes
disadvantages. Telephone interviewing stands out as the best __________________
method for gathering quickly needed information. It has the __________________
advantage over a mailed questionnaire as it permits the __________________
interviewer to talk to one or more persons and to clarify his __________________
questions, if they are not understood. The response rate for
__________________
telephone interviewing seems to be a little better than for
__________________
mailing questionnaires. The two main drawbacks of
__________________
telephone interviewing are that only people with telephones
can be interviewed and only short, not too personal __________________

interviews can be carried out. __________________


__________________
The questionnaire mailing may be the best way to reach
persons who would not give personal interviews or who
might be biased by interviewers. It is typically the
least expensive than other major methods. On the other hand,
mailing questionnaires require simple and clearly worded
questions. The response rate to mailed questionnaires is
typically low.

Personal interviewing is the most versatile of the three


methods. The personal interviewer can ask more questions
and can supplement the interview with personal
observations. These advantages come at a high cost, however,
personal interviewing is the most expensive method and
requires much more technical and administrative planning
and supervision. In a real sense, companies turn to telephone
interviewing or questionnaire mailing as a second choice out
of cost consideration.

Construction of a Questionnaire
When information is to be collected by asking questions to
people who may have the desired data, a standardized form
called questionnaire is prepared. The questionnaire has a
list of questions to be asked and spaces in which the
respondents record the answers. Each question is worded
exactly as it is to be asked. Also, the questions are listed in a
established sequence.

Questionnaire construction is discussed the coming in nine


steps. These steps may vary in importance in individual
projects, but each one must be thought through.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
180 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The nine steps are:


__________________
i. Decide what information is wanted.
__________________
__________________ ii. Decide what type of questionnaire (personal interview,
__________________ mail, telephone) to use.
__________________ iii. Decide on the content of individual questions.
__________________
iv. Decide on the type of question (open, multiple choice,
__________________
dichotomous) to use.
__________________
__________________ v. Decide on the wording of the questions.
__________________
vi. Decide on question sequence.

vii. Decide on lay out and method of reproduction of


questionnaire.

viii. Make a preliminary draft and pretest it.

ix. Revise and prepare the final questionnaire.

Let us discuss each of these steps in detail.

– Basically, a questionnaire must serve two functions;


translate research objectives into specific questions and
motivate the respondent to cooperate with the survey
and furnish the information correctly. Therefore, before
a questionnaire can be formulated, a specific statement
of the information which is needed must be made. The
complete analysis must be anticipated. For example, it
is not enough to know that the objective is to find what
type of persons form the market for the product. The
specific characteristics that are thought to be important
must be stated. Thus a deodorant soap company wishing
to define its market might specify that the survey
determine the proportion of individuals using the soap
within the groups shown in Table 7.1. With such a
listing the questionnaire framer could develop a series
of questions that would elicit the information needed
for the proposed analysis. This statement of objectives
is part of the problem definition step. At the risk of
being repetitive, it is important here to emphasize that
a questionnaire cannot be started until the precise
information sought is known.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 181
Table 7.2 Notes
__________________
Age Sex Income Geographical Area City Size
__________________
Under 20 Male Under $3,000 North-east Over 1,000,00 __________________

20-29 Female $3,000-$4,999 North-central 50,000-1,000,000 __________________


__________________
30-45 $5,000-$7,499 South 2,500-49,999
__________________
Over 45 $7,500 and over West Rural
__________________
– Determine the manner questionnaire to be used: __________________
Questionnaire can be used by personal interview, mail, __________________
or telephone. The choice among these alternatives is __________________
largely determined by the type of information to be
obtained and by the type of respondents from whom it
is to be obtained. It is necessary to decide on the type of
questionnaire at this point since the questions asked,
the way, in which they are asked and the sequence in
which they are asked will all be influenced by this
decision.

– Determine the content of individual questions: Once the


specific information needed is known and the method
of communication is decided, the researcher is ready to
begin formulating his questionnaire. A first problem is
to decide what to include in individual questions. The
following points are in the nature of standards against
which to check possible questions; obviously, they leave
much to the originality of the researcher.

l Is the question necessary?

l Are several questions needed instead of one?

l Does the respondent have the information


requested?

l Is the point within the respondent’s experience?

l Can the respondent remember the information?

l Will the respondent have to do a lot of work to


get the information?

l Will respondents give the information?


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
182 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Some respondents do not want to answer particular


__________________ questions. It goes without saying that such questions hurt
__________________ cooperation for the rest of the interview and, therefore,
__________________ should be eliminated. It is often possible, however, to change
__________________ such questions so as to secure the desired information.
__________________ General motors developed a classic questionnaire to get
__________________ consumer opinion on car styling. Few consumers could
__________________ describe the type of car that they liked best, but, if they were
__________________ shown pictures of ten cars they could point out the car they
__________________
liked best. The pictures made it possible for the respondent
to report his answer.
__________________
The interviewing situation, as seen by a respondent, is quite
different than as seen by the interviewer. The respondent
has no particular interest in the accuracy of the survey as
such. His sees a situation in which a person is asking him
questions; his answers will, in his opinion, influence the
interviewer’s opinion about him. Thus, he may be hesitant
to answer accurately if he sees the question as one which
relates to her social status. Such questions, therefore, will
elicit incorrect answers which are useless. A questionnaire
on magazine readership, for example, showed relatively little
readership of pulp magazines in the upper-income groups.
On the theory that these people might feel it was degrading
to admit reading such “trash,” the study was revised. People
were asked to contribute their old magazines to charity. A
count of these magazines showed that the number of women
contributing certain magazines exceeded the number who
reported reading them by 50 per cent. During the redemption
of war bonds, display of government posters and absenteeism
were points of personal prestige.

Questions of this type should be omitted unless a way can


be found to ask for the information that will obtain correct
responses. In many research studies it is desirable
to know the income of respondents. It is commonly believed
that direct questions on this subject may elicit incorrect
replies. The amount spent for housing, however, is
usually correlated with income. Furthermore, houses (and
incomes) within a block tend to be similar. It is possible in
larger cities, therefore, to obtain estimates of income from
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 183
census data by checking the average rental level of the blocks Notes
in which respondents live. In other cases respondents can __________________
be asked how much rent should be for “houses in this block”. __________________
Such a question relates less directly to the respondent and, __________________
hence, is more apt to obtain an unbiased reply. __________________

Another procedure which seems to reduce the resistance to __________________


reporting income data is to use an income card. On the card __________________
various income classes are listed. For example: __________________

Class A ...............................under $3,000 __________________


__________________
Class B ...............................$3,000-$5,000
__________________
Class C ...............................over $ 5,000

Interviewers hand over the card to respondents and ask into


what class their income falls. Respondents seem to find it
easier to say “class A” than to report income of $2,7000.

– Determine the type of question to use: Once the content


of individual questions is decided, the researcher is
ready to begin framing the actual questions. Before he
can work on the wording of each question, he must decide
on the type of questions to use. There are three major
types from which he may choose: (1) Open (2) Multiple
choice, and (3) Dichotomous.

Multiple choice questions overcome some of the


disadvantages of open questions, but incur some new
ones. Open questions are subject to interviewer bias in
recording of answers. This is not the problem with
multiple choice questions where answers are in one or
more of the stated alternatives. All that the interviewer
has to do is to check the applicable reply. Thus, the
multiple choice questions are faster and less subject to
bias.

Also, multiple choice questions simplify the tabulating


process. The difficult and time-consuming editing
process is reduced to a rapid check for mechanical
accuracy.

Multiple choice questions give a list of alternative


answers. It is important to point out that this list must
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
184 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes include all alternatives or there will be a bias against


__________________ those options which are omitted. The alternatives
__________________ mentioned in a multiple choice questions will be
__________________ reported by more respondents compared to a situation
__________________ when open questions are used.
__________________ Multiple choice questions also tend to bias results by
__________________ the order in which the alternative answers are given.
__________________
This bias of position can be overcommed by alternating
__________________
the order in which alternatives are listed.
__________________
__________________
Dichotomous question or two-way question, is an
extreme of the multiple choice question. The idea is to
offer only two choices – yes or no, did or didn’t, cash or
credit, railroad or airline, etc. Such questions are the
most widely used of the three basic types. The following
are examples of dichotomous questions:

u Would the service proposed by X lines make


motor freight service more useful to you?
u Is any of this discount normally passed on to
other?
u Did you buy it or was it a gift?
u Was it new or used when you got it?

To understand the weight of responses, such type of


questions are used. We have used such questions in our
sample questionnaire appended at the end.

u Decide on question sequence: Once the wording of the


individual questions has been determined, it is
necessary to set them up in some order. The order
chosen can change the results obtained. There are three
major sections in a questionnaire (1) The basic
information sought, (2) Classification information, and
(3) Identification information. Since questions
pertaining to these sections tend to be of declining
interest to the respondent, the sections are usually put
in the order shown. Questions relating to the basic
information are place first. To help in analyzing this
information, it is usually necessary to be able to classify
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 185
respondents on such bases as age, sex, income, education Notes
and nationality. Questions on these points form the __________________
classification section. The identification section __________________
identifies all parties involved. This includes the name __________________
and address of the respondent, and the names of such
__________________
individuals as the interviewer, editor, and card puncher.
These are used to permit checking for cheating among __________________
interviewers and to assign responsibility for the tasks __________________
done. __________________

u Decide on layout and reproduction: The physical __________________


layout and reproduction of a questionnaire can influence __________________
its success. Three major points should be considered in __________________
planning the layout and reproduction of the
questionnaire:

l Securing acceptance of the questionnaire by


respondents,

l Making it easy to control the questionnaire, and

l Making it easy to handle the questionnaire.

u Pretest : Before a questionnaire is ready for the field it


needs to be pretested under field conditions. No
researcher can prepare a questionnaire perfectly in
the first attempt, improvements can hence be suggested
in field tests. Pretests are best done by personal
interview even if the survey is to be handled by mail or
telephone. Interviewers can note the respondents
reactions and attitudes which cannot otherwise be
obtained. After any pertinent changes in the
questionnaire have been made, another pretest can be
done by mail or telephone, if those methods are to be
used in the survey. This latter pretest should uncover
any weakness peculiar to the method of communication.

u Revisions and final draft: After each significant


revision of the questionnaire, another pretest should
be done. When the last pretest suggests no new
revisions, the researcher is ready to print the actual
questionnaire to be used in the survey.

u Selection of a sample: An elementary unit is an


element or group of elements on which observations can
be made or from which the required statistical
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
186 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes information can be ascertained according to a well


__________________ defined procedure. Elementary units or groups of such
__________________ units which are convenient for purposes of sampling are
__________________ called sampling units.
__________________ The totality of all sampling units belonging to the population
__________________ to be studied with their proper identification particulars is
__________________ termed as the sampling frame. Editing would also help to
__________________ eliminate inconsistencies or obvious errors due to
__________________ arithmetical treatment.
__________________ When the data is to be processed by computers, then it must
__________________ be coded and converted into the computer language. This
coding job should be done while editing the data.

Methods for Secondary Data Collection


Secondary data may either be published data or unpublished
data such as:

u Various publications of the central, state and local


governments.

u Various publications of foreign governments or of


international bodies and their subsidiary organizations.

u Technical and trade journals.

u Books, magazines and newspapers.

u Reports and publications of various associations


connected with business and industry, banks,
universities, economists, etc., in different fields.

u Reports prepared by research scholars, universities


economists, etc., in different fields and

u Public records and statistics historical documents, and


other sources of published information.

The sources of unpublished data are many; they may be found


in diaries, letters, unpublished biographies and
autobiographies and may also be available with scholars and
research workers, trade associations, labour bureaus and
other public/private individuals and organizations.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 187
Before using secondary data, the researcher must ensure Notes
the reliability, suitability and adequacy of data. __________________
__________________
Secondary Data – Internal
__________________
Internal records or published records are often capable of __________________
giving remarkably useful information. Sometimes, the __________________
information may be sufficient enough to give the desired __________________
result. However, this preliminary information shall most of __________________
the time help in developing the overall research strategy
__________________
and hence must be undertaken before any further research
__________________
is contemplated. For a manufacturing industry, for example,
__________________
the internal production and sales records, if designed and
maintained properly, can help in a big way even for
formulating the companies strategies.

Secondary Data – External


External sources of data include statistics and reports issued
by governments, trade associations and other reputable
organizations such as advertising agencies and research
companies and trade directories.

In India, some of the major sources of secondary data are:

Indian Council of Agriculture, Central Statistical


Organization , Army Statistical Organizations, National
Accounts Statistics, Bulletin on Food Statistics, Handbook
of Statistics on Small Scale Industries, RBI Bulletin, Annual
Survey of Industries, Indian Labour Yon Book, etc.

Schedule Method
Schedule Defined
Schedule may be defined as a proforma that contains a set
of questions which are asked and filled by an interviewer
in a face to face situation with another. It is a standardized
device or tool of observation to collect the data in an
objective manner. In this method of data collection the
interviewer puts certain questions and the respondent
furnishes certain answers.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
188 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The interviewer records as they are given:


__________________
__________________ Purpose/Objectives of the Schedule
__________________ The main objectives of the schedule are as follows:
__________________
u Delimitation of the topic: A schedule is always about
__________________
a definite item of enquiry. It's subject is a single and
__________________
isolated item rather than the research subject in general.
__________________
The schedule therefore delimits and specifies the
__________________
subject of enquiry.
__________________
__________________ u Aids to memorize: It is not possible for the interviewer
to keep in mind or memorize all the information that
he collects from different respondents. With our a
standardized tool, he might ask different questions to
different respondents and thereby get confused when
he requires to analyze and tabulate the data. Therefore
schedule acts as an "aide memoire".

u Aid to classification and analysis: Another objective


of the schedule is to tabulate and analyze the data
collected in a scientific and homogeneous manner.

Types of Schedules
These are as follows:

Observation Schedule
The schedules which are used for observation are known as
observation schedules. Using this schedule, observer records
the activities and responses of an individual respondent or
a group of respondents under specific conditions. The
main purpose of the observation schedule is to verify
information.

Rating Schedule

Rating schedules are used to assess the attitudes, opinions,


preferences, inhibitions, perceptions and other similar
elements or attributes of respondent. As evident from the
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 189
term rating in these, schedules the value and the trend of Notes

the mentioned qualities is measured. __________________


__________________
Document Schedule __________________
__________________
These schedules are used in exploratory research to obtain
data regarding written evidence and case histories from __________________

autobiography, diary, or records of government etc. It is an __________________

important method for collecting preliminary data or for __________________

preparing a source list. __________________


__________________
Institution Survey Schedules __________________

This type of schedule is used for studying different problems


of institutions.

Interview Schedule

Using his schedule, an interviewer presents the questions


to the interviewee and records his responses in the given
space of the questionnaire.

Merits of Schedule Method

The schedule method has the following merits:

u Higher response: In the schedule, since a research


worker is present and he can explain and persuade the
respondent, response rate is high. In case of any mistake
in the schedule, the researcher can rectify it.

u Saving of time: While filling the schedule, the


researcher may use abbreviation or short forms for
answers, he may also generate a template. All these
steps help in saving of time in data collection.

u Personal contact: In the schedule method there is a


personal contact between the respondent and the field
worker. The behaviour, and character of respondent
obviously facilitates the research work.

u Human touch: Sometimes reading something does not


impress as much as when the same is heard or spoken
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
190 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes by experts as they are able to lay the right emphasis.


__________________ This greatly improves the response.
__________________
u Deeper probe: Through this method it is possible to
__________________
probe deeper into the personality, living conditions,
__________________
values, etc., of the respondents.
__________________
__________________ u Defects in sampling are detected: If there are
__________________ some defects in schedule during sampling it easily
__________________
come to the notice and can be rectified by the
researcher.
__________________
__________________ u Removal of doubts: Presence of enumerator removes
the doubts in the minds of respondent on the one hand
and avoid from the respondent artificial replies owing
to fear of cross checking on the other hand.

u Human elements make the study more reliable and


dependable: The presence of human elements makes
the situation more attractive and interesting which helps
in making interview useful and reliable.

Limitations of the Schedule Method


Following are the main limitation of the schedule method:

u Costly and time-consuming: This method is costly and


time consuming due to its basic requirement of
interviewing the respondents. This becomes a
serious limitation when respondents are not found in
a particular region but are scattered over a wide
area.

u Need of trained field workers: The schedule method


requires involvement of well trained and experienced
field workers. This involves great cost and sometimes
workers are not easily available forcing engagement of
inexperienced hands, which defeats the purpose trained
of research.

u Adverse effect of personal presence: Sometimes


personal presence of enumerator becomes an inhibiting
factor. Many people despite certain facts knowing
cannot say them in the presence of others.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 191
u Organizational difficulties: If the field of research Notes
is dispersed, it becomes difficult to organize it. __________________
Getting trained manpower, assigning them duties __________________
and then administrating the research is a very difficult __________________
task. __________________
__________________
Characteristics of a Good Schedule
__________________
The following are the essentials or characteristics of a good __________________
schedule:
__________________
u Accurate communication: It means that the questions __________________
given in the schedule should enable the respondent to __________________
understand the context in which they are asked.

u Accurate response: The schedule should structured in


such a manner so that the required information are
accurate and secured. For this, following steps should
be taken.

l The size of the schedule should be precise and


attractive.

l The questions should be clearly worded and should


be unambiguous.

l The questions should be free from any subjective


evaluation.

l Questions should be inter-linked.

l Information sought should be capable of tabulation


and subsequent statistical analysis.

Suitability of Schedule Method


This method is generally applied in the following situations:

u The field of investigation is wide and dispersed.

u Where the researcher requires quick result at low cost.

u Where the respondents are educated.

u Where trained and educated investigators are available.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
192 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Table 7.3: Distinction between Schedule and Questionnaire


__________________ Description Schedule Questionnaire
__________________
1. Methodology Direct method of primary Indirect method of data
__________________ data collection. collection.

__________________ 2. Contact with Direct contact exists in this Direct contact may not exist
respondent method between respondent in this method. Response may be
__________________ and researcher. through post only.
__________________ 3. Coverage of area Limited geographical area. Useful for very largely dispersed
__________________ geographical area.

__________________ 4. Reliability of data High degree of reliability Less reliable as personal contact
collection. may not be there.
__________________
5. Types of questions Short and to the point answers Lengthy and elaborative
__________________ type questions which are of questions: answers required may
yes or no nature. be are detailed and lengthy.

6. Response to the Very high. Low response.


question

7. Clarification It is done during Not possible when questionnaire


questions. direct contact and discussion are
mailed.

8. Distribution Full questionnaire or Full text of question have to be


part there of can be distributed. distributed to the respondents.

9. Persuasion for It is feasible as respondents It is not always feasible to


response motivate can be as respondents
motivated. respondents.

10. Use in sampling Very successful method May cannot be used in sampling
method of research in sampling method of method of research.
research.

11. Instrument design Questionnaire are framed The questions are framed
keeping in view the difficulties keeping in view the educational
of tabulators and field workers. and economic standard of the
respondents.

12. Bias in the data There is great degree of bias in Probability of biasness in data
the collection of data. collection does not exist.

13. Cost and time Very large cost and time Less costly and less time
requirement is required. consuming.

14. Trained staff Trained and qualified staff. Not so trained staff is required.

15. Organization Difficult to organize. Simple to organize.

Questions to be Included in the Schedule


u The questions should be short, simple and easy to
understand by the respondent.
u The questions should have some direct relation to the
problem.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 193
u The information collected through schedule should be Notes
amendable to statistical manipulation and tabulation. __________________
__________________
u The questions must be interrelated and such that cross
__________________
checking is feasible.
__________________
u Questions should be free from personal bias and should __________________
not hurt the feelings of the respondents.
__________________
u Questions should be standardized. __________________
__________________
u Precise terms should be used in framing the questions.
__________________
u They should be subject to pretesting. __________________

Organization of Schedule
1. Selection of respondents: Generally sampling method
is used for selection of respondents. The sample must
be representative and should contain the biographical
details and other relevant information about the
respondents.

2. Selection and training of the field workers: In the


schedule method, field workers carry out the interview
and collect data. Since there is a shortage of trained
field workers, they have to be selected carefully.

3. Interviews and correct replies: For a successful


interview and correct result following steps are to be
taken:

u Correct approach: The field worker should


approach the respondent in a manner so that
the respondent understands correctly the
purpose. Generally, the respondent should be
approached when he is not busy and willing to
corporate.

u Proper response: For proper responses the


researcher should be able to convince the
respondents.

u Correct reply: For proper response and correct


reply the respondent should not be hurt. He should
be understood in right perspective and context.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
194 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Testing the Validity of the Results


__________________
After receipt of the filled-in schedule it should be subject to
__________________
certain tests and checking so that the correctness of the
__________________
collected data can be established. It can be done in several
__________________ ways. The investigator himself may select the respondent
__________________ and administer the schedule and find out the variation, if
__________________ any. If variations are of a large degree then it should be either
__________________ rejected or modified.
__________________
Research through Case Study Method
__________________
__________________ Case study method basically a qualitative research analysis
pertaining to an organization. In some cases, however, where
quantitative details about the organization are available, case
study may also provide certain insights to operational details
of research questions scientifically. Normally case study
method of research focuses on organization or at best two or
three homogenous or competing organization (for
benchmarking) unless it is a cross-organizational study to
authenticate any management practices or approve or
disapprove any hypotheses framed by a researcher to
establish certain theories. Unlike other methods of research,
case study method requires in-depth analysis of some
situations and study on their inter relationship in an
organization. Important characteristics of case study method
of research can be listed as under:

i. In case study method, a researcher can choose to study


a single organization or social unit or more. Even be
may confine his focus only on certain areas or situations
of the sample organization. To illustrate if the purpose
of research is to review an organization's strategy a
researcher may critically examine, the vision, mission,
goals, objectives, strategies, action plan and policies and
document his observations. Why Reliance can retain
manpower or GE capital not or for that matter even
TCS not, despite the fact that both GE and TCS have
pro-active HR policies and Reliance is more traditional
in their approach is something which can be answered
analyzing their strategies.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 195
Why HLL is successful in repositioning their 'Lifebouy' Notes
brand and why Colgate not again is embedded in their __________________
marketing strategy? __________________
__________________
ii. Case study method research is done very intensively
focussing on some limited identified areas rather than __________________

exhaustively. Such study may be longitudinal in nature __________________


for drawing correct inferences. __________________
__________________
iii. Even though perennially case study method is
__________________
considered a qualitative analysis, modern researchers
can also do quantitative analysis and interpret the __________________

organizational issues, when such detail are available. __________________


For example, a researcher can analyze the resource
allocation strategy of an organization in terms of cost
benefit analysis. Similarly efficacy of a compensation
strategy can be studies comparing outsourcing vis-a-vis
employment on permanent payroll.

iv. In case study method, a researcher always attempts to


establish a causal relationship between various factors
couching qualitative and quantitative information. To
illustrate effective Customer Relationship Management
(CRM) may be co-related with increased market share.
Employee motivation may be related to productivity.

v. Case study method of research is direct approach and,


therefore, requires analytic rigour and experience.

vi. Case study methods unleash the best management


practices and therefore significantly contribute in
enriching the theories. For example benefit of 'direct
participation', which is a strategy to use collective
wisdom of the people to develop a response capacity
toward some targeted goals of the organization would
not have been known to the world, unless the practices
of 'Chevron', the petroleum giant, brought into the
limelight by the researchers. Despite its advantages,
research through case study may often suffer from the
problem of generalization and unless properly
structured may also vitiate the purpose of research.
Proper structuring requires initial identification of the
research problem, collection of data, establishing casual
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
196 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes relationship and also finally developing remedial


__________________ measures. Authors suggest unless researcher has
__________________ proper knowledge about business process, one should
__________________ avoid using this method, as strong business and analytic
__________________ skills are essential prerequisite for the success of such
research.
__________________
__________________
Review Questions
__________________
__________________ 1. What are the various methods of collecting statistical
data? Which of these are reliable and why?
__________________
__________________ 2. Distinguish between primary and secondary data.

3. What is primary data? What are the advantages and


limitations of the primary data? Describe briefly various
methods of collecting primary data.

4. Describe the importance of interview method in business


research.

5. Define the term schedule. Explain the objectives of the


schedule and describe briefly the types of schedule.

6. Explain the case study method of data collection. What


are the characteristics of case study method? Explain
its advantages and limitations.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 197
APPENDIX-1 Notes
__________________
INTRODUCTION AND GUIDELINES TO CASE STUDY
(with a sample Case Study and its solution) __________________

What is a Case? __________________


__________________
A case is a short description, in words and numbers, of an actual
management situation. A typical case consists of number of pages of __________________
prose description with some quantitative materials (figures), exhibits,
organization charts, graphs or tables. __________________
__________________
The use of cases for advanced management education started at Harvard
University in 1908. __________________

What is a Case Study? __________________


__________________
Case study is a systematic way of helping students to learn from
experience. Such studies are intended to serve as the basis for class
discussion and not as an illustration of either good or bad management
practice. They are not intended to be a comprehensive collection of
teaching material. They provide a useful backbone to a programme of
study. In research also we use this method now-a-days.

Steps for a Case Study

1. Read the case carefully and note down all the material facts and
identify the problem(s).

2. Mention the area of management to which the main problem


relates. It is possible that the problem may overlap other area as
well, which may demand simultaneous solution. Total view of the
situation should not be lost.

3. Place yourself in the position of a manager and view the case


situation and the case problem.

4. The analysis should be from a managerial angle – unbiased and


objective.

5. List out various possible alternatives and then prioritize on the


basis of comparative superiority and feasibility.

6. Select the best alternative and state the same as your decision.

7. Indicate an estimate of outcome including the possible side effects,


if any.

Important Guidelines

– Note that there is no single correct solution even for a specific


management problem. Logical thinking in line with the accepted
basic principles and practices (processes) of management is what
is judged and assessed by the examiner.

– Questions which are given at the end of each case study suggest
the line of enquiry, hence study the questions carefully.

– Mere restating the facts of a case report without giving any meaning
should be avoided.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
198 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Incident Process/Incident Study


__________________ Incidents are a short sketch (within 50 to 100 words) of something which
__________________ has actually happened. This provides students with greater opportunity
to learn by doing things for themselves from situational experiences.
__________________ Often the incident seems trivial but its implications are important. Thorough
investigation of the facts of incident is necessary for two-part decision;
__________________
__________________ – To settle the immediate issue,

__________________ – To suggest long-term preventive and remedial action.


__________________ Example of a Case Study
__________________
The Case Problem
__________________
Vinod was the works manager of Laxhmi Engg. Rahul was with the
__________________ company as finance manager and head of the accounts & finance division
reporting to the works manager.

Rahul’s general attitude was to be tight fisted with financial matters. He


was always conscious of the need to conform to company policy and
procedures and he firmly opposed any deviation from policy, but was
often willing to explain the reasons for his view. He prided himself as a
man of principles. Vinod, works manager, was the type that wanted to
take action regardless of past practice or policy. He considered himself a
result oriented manager.

The differing attitudes of the works manager and the finance manager
had led to conflicts on past occasion and Vinod had warned Rahul on
two occasions that if Rahul could not carry out his instructions he was
free to search for a job elsewhere. Rahul had argued his case with a
measure of success stating that his approach was proper and that in
financial matters policy guidelines had to be observed.

One afternoon Vinod approached Rahul and said, “Here is an incentive


plan for the maintenance group. I am notifying it today and introducing it
from tomorrow. Have a look at it,” Rahul appeared to be surprised. He
had not known that an incentive plan was being contemplated for
maintenance workers as the company’s policy till then was to pay
incentives only to direct production workers. Nevertheless he took the
plan with him and returned to Vinod a couple of hours later. “You can’t
introduce this plan straightaway”, said Rahul to Vinod, and added, “we
have to give this some more thought. Our maintenance costs are too
high mainly due to lack of adherence to norms on consumption of spare
parts. This incentive plan gives no weightage to consumption of spare
parts. Further, it will only add to maintenance costs without any real
benefit to the company.”

Vinod’s reaction was one of anger. “You understand nothing of incentive


plans”, he retorted, and added, “spare-parts consumption is high due to
poor quality of spares bought by the materials department. Anyway, I
don’t intend to waste time on this with you. I am notifying this incentive
plan today”.

An altercation followed and the arguments of both Rahul and Vinod


became so loud that they attracted the attention of others in the hall,
outside Vinod’s office. “I am the boss here,” screamed Vinod and added,
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 199
“if you can’t work with me and obey my instructions, you are free to leave Notes
your job and go elsewhere.”
__________________
A few moments later Rahul was back in his office, tired and sullen. He __________________
called his secretary and said, “now, I have no option, please write down”.
He dictated his letter of resignation. __________________

Questions __________________
__________________
1. Discuss the leadership style used by Vinod. How effective is his
style? What changes would you advise? __________________

2. To what extent were Rahul’s needs being considered? What kind __________________
of organizational climate was being created? __________________
Read the case carefully and analyze the content, i.e., the situation and __________________
the personalities involved before answering the questions. The analysis
could be done mentally after jotting down the salient points arising in the __________________
case:

Analysis of the Case

This is basically a case of restraining leadership but to some extent it


also brings out issues relating to motivation of subordinates.

The facts of the case reveal:

i. Difference in the basic approach to leadership between Vinod and


Rahul. Rahul is conservative in outlook, opposed to deviations
from policy and a man of principles whereas Vinod tends to ignore
company policy and practices in an effort to be result-oriented.

ii. Rahul expects a role whereby he can state his differences with his
superior. He is also capable of convincing his superior by
expressing his point of view successfully.

iii. Vinod is autocratic in his approach. He appears to be a person


who concentrates authority and decision-making in himself and
views leadership as a matter of issuing orders and believes that
subordinates must obey them unquestioningly.

iv. The situation as revealed by the facts of the case is one of conflict
between the subordinate and superior mainly due to the leadership
style adopted by the superior.

v. Vinod does not consider it necessary to involve his subordinate in


decision-making or explain the rationale of the reasons for which
he wishes to make changes in respect of the incentive scheme,
deviating from company policy.

vi. There is no effort on the part of Vinod to answer the doubts raised
by Rahul on the demerits of the proposed incentive scheme clearly
and cogently.

vii. When Rahul explains his reasons for the need to give some more
thought before introducing the change in the incentive scheme,
the reaction of Vinod is one of anger. Anger and admonition are
expressed by the superior openly in front of others, thus leaving
the subordinate humiliated and frustrated.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
200 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes viii. The conflict leads to a decision by the subordinate to quit the
organization.
__________________
__________________ Leadership Style Used by Vinod

__________________ Vinod’s leadership style is autocratic and authoritarian. An autocratic


leader concentrates all authority and decision-making powers on himself.
__________________ There is no participation by his subordinates in the decision-making
__________________ process, they simply do what they are told to do.

__________________ An autocratic leader believes that his leadership is based on the authority
conferred on him by his position and that his subordinates should follow
__________________
instructions unquestioningly.
__________________
Autocratic leadership leads to lack of confidence and insecurity in the
__________________ subordinate and stifles creativity and innovative ability of the subordinate.
The autocratic style stifles subordinate development and organizational
__________________
effectiveness.

How Effective is Vinod’s Style?

Vinod’s style is not effective, viewed from the point of view of the
leadership that a works manager should provide. In fact it is Vinod’s style
that has led to a major conflicting situation within the organization between
him and the finance manager which has also attracted the attention of
other persons in the organization. This is bound to lead to a sense of fear
among all the persons in the organization and also frustration due to
lack of opportunity and fear on the part of subordinates to express a view
different from that of their superior.

Vinod’s style of leadership will lead to resentment and, over a period,


resistance to his orders. Further, the style adopted by Vinod will lead to
his subordinates showing no active interest in their job, becoming hesitant
in taking initiatives and dependent upon the boss’s, instructions. All these
factors can defeat the purpose for which Vinod is attempting to modify
the incentive scheme.

What Changes would You Advise?

The facts of the case reveal that Rahul was always cognizant of the
company’s policies and procedures. He had the ability to argue his view
point with a measure of success and where he differs from his superior,
he was willing to state the reasons about his differences. In this situation
Vinod should have adopted a participative style of leadership with trust
and confidence in Rahul so that communication with him is open and
teamwork is built up.

To be an effective leader, Vinod should have a greater measure of


confidence and trust in his subordinates, help them in the development
of methods to achieve organizational objectives and also encourage
participation.

The facts of the case show that Rahul is willing to explain to his superior
his points of view on matters where he differs from his superior. He seeks
the freedom to discuss matters. His objective is to interact with his superior
without fear bearing in mind organizational goals.

Lack of communication that underlines the leadership style of Vinod and


the sense of insecurity that stems from this leadership style have an
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 7 Data Collection Methods 201
adverse impact on the feeling of pride that Rahul has. Pride is a powerful Notes
and valuable motivator and when this is stifled, the subordinate feels
__________________
frustrated.
__________________
Rahul expects a measure of involvement in decision making. Involvement
satisfies one’s ego and self-esteem. But in the present situation Rahul is __________________
denied the opportunity for involvement in decision-making.
__________________
Rahul's needs for motivation are not satisfied to any extent. This is the __________________
reason why he decides to leave the organization.
__________________
Vinod’s style will lead to a climate where the subordinate refuses to
__________________
participate in the decision-making process. Subordinates will in this
situation become ‘yes men’ and refuse to confidently accept authority __________________
delegated to them. In this environment, group effectiveness will be lost. In
the final analysis this adverse climate will not provide a conducive __________________
atmosphere for Vinod to be an effective leader who can function __________________
comfortably and achieve results competently. Vinod’s claim that he is a
‘result-oriented’ manager will be self-defeating if he does not direct the
organization to a participative style of leadership.
203

Unit 8 Notes
__________________
__________________
Data Analysis-Methods __________________
__________________
__________________
__________________
__________________
__________________
__________________
__________________

Objectives
After reading this unit you will be able to understand:

y Statistical methods of analysis of data.

y Statistical methods of interpretation of data.

Presentation of Data
All business decisions are based upon evaluation of some
data. Availability of right information is very important for
making right decisions. Once the data begins to flow in,
attention turns to data analysis.

It is almost impossible for the management to deal with all


this data in its raw form. Such data must be presented in a
suitable and summarized form without any loss of relevant
information so that it can be efficiently used for decision
making.

Whenever there is a need to present numerous figures or to


describe a technical process or procedures, graphic aids can
help to communicate this information to your audience more
quickly. The two graphic aids mostly used in research reports
are tables and graphs. Besides making the report easier to
read and understand, graphic aids improve its physical
appearance.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
204 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes General Guidelines for Presenting Graphic Aids


__________________
Normally, it is best to place an illustration within the text if
__________________
the reader needs to refer to it while reading the report. It
__________________
should be placed as close to the discussion as possible. If
__________________ information is supplemental or extremely lengthy, it can be
__________________ placed in the appendix.
__________________
All graphic aids should contain the following elements:
__________________
__________________ 1. Table or figure number: This permits easy location in
__________________ the report.
__________________ 2. Title: The title should clearly indicate the contents of
the table or figure.

3. Boxhead and subhead: The boxhead contains labels


to the column in a table, while, the subhead contains
the labels for the rows.

4. Footnotes: Footnotes explain or qualify a particular


section or item in the table or figure.

Data can be presented in tabular or graphic form. The tabular


form (tables) implies numerical presentation of data. The
graphical form (figures) involves the presentation of data in
terms of structures which can be visually interpreted, e.g.,
bar charts, pie charts, line charts.

Oral Presentation
Many companies require oral presentation of research
reports. Prior to the presentation follow this checklist:

1. Check all equipment (e.g., light, microphones, projectors


etc.) thoroughly before the presentation.

2. Have a contingency plan for equipment failure.

3. Analyze your audience. How will they react to the


research findings?

4. Practice the presentation several times. If possible, have


someone comment on how to improve its effectiveness.

5. Start the presentation with an overview – tell the


audience what you are going to tell them.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 205
6. Face the audience directly at all times. Notes
__________________
7. Talk to the audience or decision makers, rather than
__________________
read from a script or a projection screen.
__________________
8. Use visual aids effectively – charts and tables should be __________________
simple and easy to read. __________________

9. Avoid distracting mannerisms while speaking. __________________


__________________
10. Remember to ask the audience if they have any questions
__________________
after your report is concluded.
__________________
Editing __________________

The first step in analysis is to edit the raw data. Editing


detects errors and omissions correct them as far as
possible.

Purpose of Editing
u To ensure accuracy,

u To bring about consistency with other information,

u Make sure that data is uniformly entered,

u It is complete, and

u Arranged to simplify coding and tabulation.

The editor’s responsibility is to decide which of the responses


is consistent with the intent of the question or other
information in the survey and is most accurate for its
individual respondent.

Classification of Data
When the raw data has been collected and edited it should
be put into an ordered form (ascending or descending order)
so that it can be looked at more objectively. The next
important step towards processing the data is classification.
Classification means separating items according to similar
characteristics and grouping them into various classes. The
items in different classes will differ from each other on the
basis of some characteristics or attributes.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
206 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The classification of data into limited categories specifies


__________________ some data detail but is necessary for efficient analysis.
__________________
Coding
__________________
__________________ Coding is the process whereby the data is assigned a
__________________ numerical code and value so that it can be more easily fitted
__________________ into appropriate categories. Coding is necessary where
__________________
machine tabulation is used.
__________________ There are four rules that guide coding:
__________________
u Appropriate to the research problem and purpose
__________________
u Exhaustive

u Mutually exclusive

u Derived from one classification principle.

Data Entry
Data entry convert information gathered by a researcher
through primary and secondary method to a medium for
viewing and manipulation.

Tabulation
Tabulation consists of counting the number of cases which
fall into the established categories. Before tabulation can
take place the plan of analysis must be thought through. A
tabulation plan which specifies the precise counts to be
obtained needs to be prepared. Tabulation may be
accomplished using manual or machine methods. Both
processes have unique advantages and disadvantages.

Machine tabulation is preferable where a large number of


questionnaires is involved and where cross tabulation is
planned.

After the tabulation step has been completed, it is necessary


to determine whether any differences exist and whether
these differences are statistically significant. If the results
from a marketing study are to be useful they must be
organised and summarized in such a way as to enable the
objectives of the study to be achieved. This requires the use
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 207
of those summary statistical measures which typify the group Notes
such as arithmetic mean and percentages, and the use of __________________
those measures which indicate dispersion such as standard __________________
deviation. All these measures have their advantages and __________________
disadvantages and the analyst must be competent in __________________
selecting the most appropriate measure to use.
__________________
Further, since most marketing data are collected using a __________________
sample, the reliability of the summary estimates must be __________________
determined. Standard error and confidence limits are two __________________
terms and concepts which are fundamental to the general __________________
subject of reliability.
__________________
To determine the significance of any difference existing
between a sample estimate and a similar estimate from
another sample requires the use of tests of significance. Such
tests indicate the probability of the difference having a cure
due to random sampling variations. All these tests make use
of the standard error and confidence limit concepts.

The next step in the analysis procedure consists in trying to


explain ‘why’ differences exist. This is accomplished by
attempting to identify a casual relationship between two or
more variables. To accomplish this, the survey data are
arranged to approximate and control the experiment. Casual
analysis requires cross tabulations. This procedure
introduces one or more additional variables into the analysis
by dividing the sample into subgroups. This helps to confirm
or reject the earlier relationship. In introducing additional
variables, the analyst must consider the size of the sample
in the various subgroups and also be on guard against
spurious correlations.

Statistical Analysis and Interpretation of Data:


Non-parametric Tests
Introduction
Theoretical aspects of research have been adequately
discussed to enable the readers to understand norms in their
true spirit. This chapter is intended to focus on basic
statistical techniques and their applicability in research.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
208 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes As has already been pointed out, social science research, or


__________________ to be more particular, research on commerce and
__________________ management, mostly deals with qualitative variables. Unless
__________________ qualitative impressions are duly quantified, research
__________________ findings may not be clearly understood and may even be
__________________
questioned by others as to their authenticity. Statistical
techniques help the researcher to counter this problem. In
__________________
fact, it is being increasingly felt that understanding basic
__________________
statistical norms is a prerequisite for the success of a
__________________ researcher.
__________________
Stuart A Rice put forth the following arguments to justify
__________________
quantitative expressions of research findings:

u It reduces individual bias to a minimum.

u Permits verification by other investigators.

u Reduces and at the same time makes evident the margin


of error.

u Replaces the less exact meaning of descriptive words


with the precision of mathematical notation.

Data collected from secondary sources may be compiled with


great ease in the form of tables. Data from secondary sources
enables us to define the characteristics of phenomena we
which again try to investigate in the light of our own findings
based on data collected from primary sources.

In some researches, however, where we do not collect or


compile data from primary sources, we generally base our
arguments by mere analysis of data from secondary sources.
For example, an investigation on the ‘structural change of
occupational pattern in India’ may be made based on analysis
of census data over dacades. Similarly, changing pattern of
sectorial distribution of Net Domestic Product (NDP) in
India may be analyzed in the light of secondary data alone.
Use of secondary data in direct form may not always help a
researcher to relate such quantitative trends to his research
problem. For example, an absolute analysis of wage cost and
its shares in total cost of production in comparison with other
countries may certainly lead a researcher to conclude that
labour wage costs in India are relatively poor. But if a
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 209
researcher, with some amount of inquisitiveness, goes into Notes
the depth of problem and tries to compare the per capita __________________
share of wages as percentage to the per capita share of NDP, __________________
the picture will be totally different. In fact, Indian workers, __________________
in this respect, are more privileged to gain a higher stake __________________
than workers in UK, USA and Japan. Readers here are
__________________
requested to refer the world development reports published
__________________
by the world bank.
__________________
Again, while using data from secondary sources, the __________________
researcher should thoroughly check the background of such __________________
data collection, its methodology and limitations, etc. The
__________________
methodology of data collection varies even for same data
provided by different governmental agencies in India. For
example, data related to labour matters as made available
in Indian labour year books and annual survey of industries
widely varies for obvious differences in methodology
followed in collection and compilation. Even the same source
may use different methodologies in different years of
publications. The census data is a glaring example in this
respect.

Collection of data from primary sources involves survey which


may be carried out either by interviewing the samples in
person or by mailing questionnaires to samples. Since mailing
response is alarmingly poor, the researchers here are advised
to follow the interview methods for carrying out the survey.
In selecting samples, thus, he should be cautious. If he is
unable to visit areas widely dispersed, sample should be
chosen from those areas where he can afford personal visits.
Similarly, he should also consider other constraining factors
like time and money while deciding the number of samples.

Usually in case of structured questionnaire, samples are


interviewed. Structured questionnaire ensures easy
quantification.

After necessary collection of data from primary sources, with


the help of tabulation, the researcher arranges these in some
concise and logical order. And for such purpose he or she
needs to classify the data. The basis of classification depends
on the characteristics of data collected. In most cases such
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
210 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes classification scheme is pre-determined. However, such


__________________ scheme may get altered or modified after data collection.
__________________
Following criteria are generally considered for statistical
__________________
classification:
__________________
__________________ u Geographical
__________________ u Chronological
__________________
u Qualitative
__________________
__________________ u Quantitative
__________________
Under geographical classification, data are organised in
terms of geographical division as country, state, region, city,
village, block, ward etc. Similarly, century, decade, year,
month, week, day, etc., is considered while arranging data
chronologically. Qualitative attributes such as sex, religion,
literacy, occupation, marital status, etc., are also considered
in data classification. Distribution of data by size and
magnitude (frequency distribution as for example) is done
through quantitative classification. After determining the
classification, the collected data are sorted and counted as
per various categories. For handling large data, hand
tabulation may not always be possible. When mechanical
tabulation is used, data are first transferred to punch cards
by punching holes for each item. Then with the help of
sorting machine and tabulation machine, data are finally
arranged in the form of tables as per the above classification.

But in our country, it is not always possible for a researcher


to avail such mechanical facilities and in most cases
tabulation part is completed manually.

Assigning numerical values to each variable is a difficult task


for a researcher. Quantitative variables, i.e., variables which
have a certain number of defined units may either be
cardinally (explicit) or ordinally (implicit) manipulated by
simple mathematics. But qualitative variables may be
manipulated only by defined characteristics. Similarly, the
researcher should also understand the basic difference
between continuous and discrete variables while arranging
data.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 211
With the help of frequency distribution, variables are Notes

classified according to magnitude or size. After necessary __________________


tabulation, the next task of the researcher is to arrange them __________________
in statistical tables. There are certain standard norms which __________________
usually are followed in constructing tables. __________________
__________________
For simple understanding it should be remembered that
such tables should be able to reveal significant relations __________________

among the data or should be able to emphasize certain __________________

facts. __________________
__________________
The frequency distributions may also be presented in the
__________________
form of graphs. Histogram, frequency polygon, smoothed
frequency curve are the commonly used curves.

After necessary frequency distributions to represent the


characteristics of entire series by few significant figures, we
take the help of averages. Mean, median and mode are three
of the most commonly used averages.

Here we have discussed various statistical models for


research.

Hypothesis Testing

Inferences on population characteristics (or parameters)


are often made on the basis of sample observations,
especially when the population is large and it may not
be possible to enumerate all the sampling units belonging
to the population. In doing so, one has to take the help
of certain assumptions (or hypothetical values) about
the characteristics of the population if some such information
is available. Such hypothesis about the population is
termed as statistical hypothesis and the hypothesis is
tested on the basis of sample values. The procedure
enables one to decide on a certain hypothesis and test its
significance. “A claim or hypothesis about the population
parameters or values is known as null hypothesis and is
written as, Ho.”

This hypothesis is then tested with available evidence and a


decision is made whether to accept this hypothesis or reject
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
212 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes it. If this hypothesis is rejected, then we accept the alternate


__________________ hypothesis.
__________________
This hypothesis is written as H1.
__________________
__________________ For testing hypothesis or test of significance we use both
__________________ parametric or standard tests and nonparametric or
distribution free tests. Parametric tests assume within
__________________
properties of the population, from which we draw samples.
__________________
Such assumptions may be about population parameters,
__________________
sample size, etc. In case of non-parametric tests, we do not
__________________ make such assumptions. Here we assume only nominal or
__________________ ordinal data.

Important parametric tests used for testing or hypothesis


are:

i. z – test

ii. t – test

iii. x2 test; and

iv. f – test

When x2 test is used as a test of goodness of fit and also as a


test of independence, we use non-parametric tests.

As has been stated earlier all parametric tests used for


testing of hypothesis are based on the Assumption of
normality, i.e., population is considered to be normally
distributed.

Procedure for Testing of Hypothesis


1. State the null hypothesis as well as the alternate
hypothesis: For example, let us assume the population
mean = 50 and set up the hypothesis m = 50. This is
called the null hypothesis and is denoted as;

Null hypothesis, Ho : m = 50

Alternative H1 : m ≠ 50

hypothesis
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 213
2. Establish a level of significance (prior to sampling): Notes

The level of significance signifies the probability __________________

of committing Type 1 error µ and is generally taken __________________


as equal to 0.05. Sometimes, the value µ is established __________________
as 0.01, but it is at the discretion of the investigator to __________________
select its value, depending upon the sensitivity of the __________________
study. To illustrates per cent level of significance __________________
indicates that a researcher is willing to take 5 per cent __________________
risk of rejecting the null hypothesis when it happens to __________________
be true. __________________

3. Choosing a suitable test statistic: For the purpose of __________________

rejecting or accepting the null hypothesis, a suitable


statistics called “test statistics” is chosen. This
means that Ho is assumed to be really true. Obviously
due to sampling fluctuations, the observed value of
the statistic based on random sample will differ
from the expected value. If the difference is
large enough, one suspects the validity of the
assumption and rejects the null hypothesis (Ho). On the
other hand, if the difference may be assumed due to
sampling (random) fluctuation, the null hypothesis (Ho)
is accepted.

Fixing up the Level of Significance and Defining the Critical rejection)


Regions

Based on the sampling distribution of the statistic, the


probability of obtaining a difference equal to or greater than
the observed difference is calculated on the assumption that
the null hypothesis is true.

If we select the value of µ = Level of significance = 0.05,


and use the standard normal distribution as our test
statistic for testing the population parameter u, then
the value of the difference between the assumption of
null hypothesis (assumed value of the population parameter)
and the value obtained by the analysis of the sample results
is not expected to be more than 1.96 s at u = 0.05.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
214 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes This relationship can be shown by the Figure 8.1 given


__________________ below:
__________________
__________________ Acceptance Zone
__________________
__________________
__________________ Rejection
__________________ Zone
__________________
__________________ Rejection
Zone
__________________

1234567890
1234567890 12345678
1234567890
1234567890 12345678
1234567890 12345678
12345678
1234567890 12345678
1234567890
1234567890 12345678
1234567890 12345678

Figure 8.1
z = Observed value of the statistic (x) – Expected value (u)
Standard error of the estimate
Calculated value of z as given by the formula above,
= |Absolute value|

Now find out the critical (or tabulated) value of the statistic
(z or t or f, etc).

z-test is used to measure the difference between any variable


value (x) and the mean of all variable values or
'x' values, which is indicated by 'm', divided by the standard
deviation(s). It is based on the normal probability
distribution. In following cases, we find the usage of
2-test:

i. To judge the significance of statistical measures,


particularly the mean. This is done by comparing the
observed value (test statistic) with the probable value
(table value) at a specified level of significance.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 215
ii. It is used to compare the mean of a sample with some Notes
hypothesized mean of the population. __________________
__________________
iii. It is also used to judge the significance of difference
__________________
between means of two independent samples.
__________________
iv. It can also be used for judging the significance of __________________
different between sample and population proportion or
__________________
proportions of two independent samples.
__________________
v. Finally this test can also be used for measuring the __________________
significance of medium, mode coefficient of correlation __________________
and other measures. __________________

We have illustrated usage of z-tests in testing hypothesis with


some real life examples. In case of small sample and in cases
when sample variance is constructed as population variance,
population variance, being not know, we use t-test based on t-
distribution to judge the significance of sample mean or
significance of difference between two sample means.
Normally when sample size is less than 30 we use t-test.

Let us explain the usage of z-test and t-test within flow chart,
as Figure 8.2 given below:

Start

Yes Use z test in


Is s Stop
computations

No

Is n>30? Yes Use z test in


computations Stop

No

Use t-test in
computations

Stop

Figure 8.2: Flow Chart for 'z' and 't' Tests


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
216 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Questions
__________________
1. A company manufacturing automobile tyres finds that
__________________
tyre life is normally distributed with a mean of 40,000
__________________
kms and standard deviation of 3,000 kms. It is believed
__________________ that a change in the production process will result in a
__________________ better product and the company has developed a new
__________________ tyre. A sample of 64 new tyres has been selected. The
__________________ company has found that the mean life of these new tyres
__________________ is 41,200 kms. Can it be concluded that the new tyre is
significantly better than the old one?
__________________
__________________ In a problem of this type, we are interested in testing
whether or not there has been an increase in the mean
life of the tyre. In other words, we would like to test
whether the mean life of new tyres has increased beyond
40,000 kms.

The various steps in testing the hypothesis are:

i. Null hypothesis and alternative hypothesis are:

H0 : m = 40,000 kms
H1 : m ≠ 40,000 kms

ii. The significance level is taken as 0.05. That is, in


5 out of every 100 occasions, there is a risk of being
wrong in accepting or rejecting the hypothesis.

iii. The test criterion is the z-test.

iv. Computations : Substituting the value of standard


deviation, s = 3000 kms in the formula:
σ
σx =
n
3000
=
64

= 375

x−µ
z=
σx
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 217

41200 − 40000 Notes

375 __________________
__________________
= 3.2 __________________

v. Decision: At 0.05 level of significance, the critical __________________


(table) value of Z = ± 1.64. __________________
__________________
As the computed value of Z = 3.2 falls in the
__________________
rejection region we reject the null hypothesis that
__________________
m = 40,000 kms. That is, the alternative hypothesis
__________________
that m > 40,000 kms is accepted. We, therefore,
__________________
conclude that the new tyre is significantly better
than the old one.

2. Assuming that the average annual income of low level


group is Rs 18,750 at national level with standard,
deviation of Rs 2610. A random sample of 100
employees of this group was taken and it was found that
their average salary was Rs 19,240.

At a level of significance a = 0.05 (95% confidence


limit), can we conclude that the average salary of
this group is representative of the national (salary)
average?

Solution

i. Null Hyp. Ho : m = Rs 18,750 n = 100,

Alternative Hyp H1 : m # Rs 18,750 x = Rs 19,240

ii. Level of Significance as given a = 0.05 s = Rs 2,610


x−µ
iii. Test Statistic Z =
SE
σ
SE = Where, x = Sample mean
n
m = Population mean

s = Standard Deviation of population mean

Z Calculated
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
218 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes iv. Since it is a two tailed test,


__________________
u The rejection region will be on both sides, i.e., 2.5%
__________________
on RH and 2.5% on LH.
__________________
__________________ u Region of acceptance = 95% z = ± 1.96 (From
__________________ Table)
19240 − 18750
__________________ z = calculated = = 1.877
2610 / 100
__________________
Since the calculated value of z is lower than the
__________________
tabulated value, the null-hypothesis stands
__________________
accepted.
__________________
\ The sample average income is a representative of
the national average income.

Second Method
What are the limits within which the average salary lies?

m = 18750

z = ± 1.96 at a = 0.05

x−µ
z=
σ/ n

x−µ
±1.96 =
2610 / 100

FG 2610 IJ
x = µ ± 1.96
H 10 K
Hence x1 = 18238

x2=19261

As the value Rs 19,240 lies between x1 and x2, hence the


sample average income is representative of the national
average income.

If the sample mean lies within these values, then we


cannot reject the null hypothesis.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 219
Notes
__________________
__________________
__________________
__________________
__________________
__________________
__________________
18750 19240 19261 __________________
__________________
__________________
One Tailed Test
3. A manufacturer of light bulbs claims that a light bulb
lasts on an average 1600 hrs. We want to test his claim.
Here, we can reject the hypotheses if the average life of
bulb is much less but we cannot reject it if it is greater
than 1600 hrs. (Whatever big value it may be). A sample
of 100 light bulbs was taken at random and the average
life was computed to be 1570 hrs with a standard
deviation of 120 hrs. At a = 1%, test the validity of the
claim of the manufacturer.

Solution
As no of observations per sample is very large , we can take
s=s

Null hypothesis, Ho : m > 1600


Alt hypothesis, H1 : m <1600
at 99% confidence limit (a = 0.01)

Z (Tabulated) = – 2.33 on the left tail of the standardized


normal curve
Acceptance Region

Rejected
1570 1600
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
220 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
s 120
__________________ z=- 2.33 and SE = =12
n 100
__________________
__________________ ' x -m 1570 - 1600 -30
__________________ z= = = =- 2.5|Z|= 2.5
σ/ n 12 12
__________________
__________________ As calculated value of |z| > Tabulated value of |z|
__________________ Hence, Ho stands rejected, i.e., management claim is not
__________________ valid.
__________________
4. The supporters of a party claim that their party has 60%
__________________
following in Delhi. To test this claim a city paper
undertakes a test sample of 400 persons on the Internet,
220 people support this claim. Test at 1% level of
significance whether the supporters claim is correct or
not.

Solution
It is a one tailed test

H0 : p = 0.6 (or H0 : p ³ 0.6)


Acceptance
H1 : p < 0.6
220
Sample proportion Px = = 0.55
400

Rejected Region
z = 2.33

SE =
b
p1− p
=
g
pq
n n

0.6( 0.4)
= = 0.0245
400

Px - P 0.55 - 0.6
z= = =- 2.04
σ/ n .0245
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 221
|z| tabulated = -2.33 at 1% level of significance. Notes
__________________
As, calculated < tabulated, hence null hypothesis is
accepted. __________________
__________________
\ Claim of the party is valid.
__________________
5. An educator claims that the average IQ of city college __________________
students is no more than 110. To test his claim, a sample __________________
of 150 students is taken and their average IQ was found
__________________
to be 111.2 with a standard deviation of 7.2 at a = 0.01.
__________________
Test his claim.
__________________
Solution __________________

Here n = 150 H0 : m £ 110

x = 11.2 H1 : m > 110

s = 7.2

Clearly it is a one tailed test, as we shall reject his claim of


Av value of IQ if found to be significantly higher than 110.

Rejected Region

m = 110

x -µ σ s 7.2
z= SE = = = = 0.59
SE n n 150

111.2 - 110
Z calculated = = 2.03
0.59

Z tabulated = 2.33 ( at a = 0.01)

As Z calculated < Z tabulated .


Hence, null hypothesis is accepted.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
222 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 6. An airline claims that at most 8% of its lost luggage is


__________________ never found. A consumer advocacy agency wants to test
__________________ the claim at 99% confidence level. Study of 200 random
__________________ cases was taken and found that for 22 cases, the luggage
__________________ was never found?
__________________
Solution
__________________
__________________
Clearly, it is one tailed test;
__________________ H0 : p £ 0.08
__________________
H1 : p > 0.08
__________________

22
Px = 0.11 (Sample proportion)
200

pq 0.08².92
SE = =
n 200

= 0.019

px - p 011
. - 0.08
z= = = 158
.
SE 0.019

Z tabulated =2.33 at a 0.01

Z calculated = 1.58

As calculated value of Z< tabulated value of Z,

Ho stands accepted.

Examples of t-test
Tensile strength of carbon steel samples in kg weight as
recorded by a operator using UTS matrix is given below:
15, 20, 18, 16, 17, 21, 20, 19, 17, 15
At 5% level of significance find out whether the mean
breaking strength of the lot can be considered 18 kg weight.

Solution
Null hypothesis that the population mean is equal to
hypothesized mean of 18 kg.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 223
Weight can be written as under: Notes
__________________
Ho:m=mHo=18 kg.
__________________
H1:m#mHo __________________
__________________
Here sample size being small, we can use t-test assuming
normal distribution, population standard deviation being not __________________

know. Test statistic (computed valve) of 't' is obtained using __________________


following formula. __________________
x − µ Ho __________________
t=
σs / n __________________
To find out x¯ and ss. We make following calculations: __________________
Sl. No xi (xi – x¯) (xi – x¯)2
1 15 –2.8 7.84
2 20 2.2 4.84
3 18 0.2 0.04
4 16 –1.8 3.24
5 17 –0.8 0.64
6 21 3.2 10.24
7 20 2.2 4.84
8 19 1.2 1.44
9 17 –0.8 0.64
10 15 –2.8 7.84
n=10 å xi =178 å (xi–x¯) 2=41.6

. x=
∑x i
=
178
= 17.8kg
n 10

∑ dx i − xi
2
41.6
ii. σs = = = 2.15 kg
n−1 10 − 1

17.8 − 18
iii. t=
2.15 / 10

= –0.294

iv. Degree of freedom = (n-1)

=(10–1)=9

v. At 5 per cent level of significance with 9 d.f. table value


of t-distribution = 2.262.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
224 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes vi. As the observed value of t (i.e., – 0.294) is in the


__________________ acceptance region, we acceptance region, we accept Ho
__________________ at 5 per cent level and conclude that the mean breaking
__________________ strength of carbon steel may be taken as 18 kg. weight.
__________________
Examples of t-test
__________________
__________________ A company has drawn sample of sales (in units) in two shops,
__________________
located in two different cities for their new launched product.
Data presented as as under:
__________________
City Mean of sales Variance Sample size
__________________
Delhi 89 6.5 7
__________________
Kolkata 82 5.6 4

The company would like to assess the difference in sales in


two cities. Find out at 5 per cent level of significance
difference between the means of two samples.

Solution
Considering null hypothesis that means of two populations
are not different, we can write as under:

Ho: m1=m2

Ho: m1#m2

We can write the given information as under:

Sample from x1 = 89 σ s21 = 6.5 n1 = 7


City Delhi

Sample from x 2 = 82 σ s22 = 5.6 n2 = 9


City Kolkata

As in this case variances of the populations are not known


and sample sizes are small, to measure the difference
between the means of two samples, we can use t-test to work
out the test statistic to as under:

x − x2 1 1
t= × +
b n − 1g σ s21 b
+ n2 − 1 g σ s22 n1 n 2

with d.f. = (n1+n2-2)=9

= 89–82
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 225

a f a f×
Notes
7 6.5 + 4 5.6 1 1
+ __________________
7+4−2 7 4 __________________

=1.5973559 __________________
__________________
observed 't' value, i.e. 1.5973559 is less than table value of 't'
__________________
at 5 per cent level of significance with d.f. 9, which is 2.262.
__________________
Therefore difference in sales of two towns is significant at
5% level, hence our null hypothesis is rejected. __________________
__________________
Chi-square Test __________________

A chi square (x2) test can be used when the date satisfies __________________

four conditions:

i. There must be two observed sets of data or one observed


set of data and one expected set of data (generally, there
are n-rows and c-columns of data).

ii. The two sets of data must be based on the same sample
size.

iii. Each cell in the data contains the observed or expected


count of five or large?

iv. The different cells in a row or column must have


categorical variables (Gr male, female or younger than
25 years of age, 25 year of age, older than 40 years of age
etc).

Examples of situations where chi-square tests can be used


are:

u For a certain product men have a different distribution


of comparative rates than women.

u Users of a certain brand in city ‘A’ show a different


income distribution than users of the same brand in city
‘B’.

Application Areas
When tests are undertaken to examine whether the sample
data supports the hypothetical distribution, such problems
are called “test of goodness of fit.”, i.e.,
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
226 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes u To test whether the sample differences among various


__________________ sample proportions are significant or can they be
__________________ attributed to chance.
__________________
u To test the independence of two variables in a
__________________ contingency table.
__________________
u To use it as a test of goodness of fit.
__________________
__________________ The c2 distribution typically looks like a normal distribution
__________________ which is skewed to the right with a long tail to the right. It
__________________ is a continuous distribution with only positive values.
__________________ c2 test has the following steps:

i. State the null hypothesis and calculate the numbers in


each category.

ii. Determine the level of significance (i.e., how much risk


of type I error) the researcher is prepared to take.

iii. Calculate c2, as follows:

X 2
=∑
dO − E i
i j
2

Ej

Where
Oi = Observed frequency
Ej = Expected frequency in the category.

iv. Find the critical value of c2 against the no of degrees of


freedom for the specified level of significance.

v. Compare the calculated value of c2 with the tabulated


(critical) value and determine the region of reflection.

NB: Care should be taken to see that the expected


frequency in the contingency table is not less than 5.

Solved Questions on c2 Test


1. A Company has been engaged in the manufacturing of
ice-cream which is sold in three sizes – large, normal
and small. It has observed that it sells the ice-cream in
large, normal and small size in the ratio of 3:5:2.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 227
The company has been thinking of introducing a Notes
different flavour and has conducted a test market for __________________
two weeks in a certain territory. The results of the test __________________
market show that it has sold 250 large packs, 400 normal __________________
packs and 350 small packs of the newly developed ice __________________
cream.
__________________
The company is now interested in knowing whether the __________________
consumption pattern of new ice-cream is different from __________________
that of the earlier one. If so, it may have to change the __________________
percentage of three different packs for the new __________________
ice-cream.
__________________

Solution
Ho : Null hypothesis, “consumption is not different from
earlier one”.
Packing size Obs freq Expected freq O-E (O-E) 2 (O-E)2
E

Large 250 300 -50 2500 8.33

Normal 400 500 -100 10000 20.00

Small 350 200 150 22500 112.50

1000 x2 140.83

Degrees of freedom = (3-1) = 2

At 5% level of significance, value of c2 for df (2) = 5.991

As c2 calculated > c2 tabulated,

We, therefore, reject the null hypothesis, “the


consumption pattern of new ice cream is not different”
from the earlier one.

2. In winter season, a study has been made for 10 days


to see whether the distribution of daily reservation
is uniform or not, and the results are tabulated
hereunder.

Find whether the distribution is uniform or not.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
228 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Distribution of actual nos of reservations:


__________________
S No 1 2 3 4 5 6 7 8 9 10
__________________
__________________ No of Reservations 65 80 100 98 75 80 82 70 60 90

__________________
Solution
__________________
__________________
Ho: Distribution of daily reservation is uniform in winter
season.
__________________
__________________ Test the validity of the assumption.
__________________ Ho : p1 = p2 = p3 = .................. = p10
__________________
H1 : p1 # p2 # p3 #.................... # p10

Average value = a65 + 80 + 100 + 98 + 75 + 80 + 82 + 70 + 60 + 90f


10

800
= = 80
10

x 2
=∑
do − E i = a65 − 80f + a80 − 80f
i j
2
2 2
+−−−−−
a90 − 80f 2

Ej 80 80 80

= 19.72

df = (c-1)(r-1) = (2-1) (1-1) = 9

value at 5% level of significance = 16.919

As calculated c2 > tabulated c2

Hence, we reject the null hypothesis, i.e., the reservation in


winter season is uniform.

Calculation of Expected Frequencies

No. of Farmers No.of Farmers who Total


Owned farms Rented farms
Using Fertilizers 100 x 80 60 100
= 40
200
Not Using Fertilizers 40 60 100
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 229
Calculation of c2 : Notes
__________________
Row Column Frequency O-E (O -E) 2 (O - E)2
__________________
O E E
__________________
1 1 50 40 10 100 2.50
__________________
1 2 50 60 -10 100 1.62
__________________
2 1 30 40 -10 100 2.50
2 2 70 60 10 100 1.67 __________________
Total 8.34 __________________
__________________
c2 = 8.34, calculated;
__________________
2´2 table D f = (2-1) (2-1) = 1 ´ 1 = 1 __________________

c2 tabulated at 5% level of significance = 3.84

As c2 calculated >c2 tabulated

Therefore, Ho stands rejected

Hence, there is an association between the ownership of


farms and the use of fertilizers.

Test for Independence of Attributes

(contingency table 2´2)

[ ac b
d ] R1
R2
®
®

C1 C2

R1= a+b C1= a+c

R2= c+d C2= b+d

N= a+b+c+d= R1+R2= C1+C2

X 2

bobserved frequency − Expected frequencyg 2

Expected frequency
b
with deg rees of freedom = R − 1 C − 1 gb g
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
230 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Yate’s Correction


__________________
To improve the approximation result obtained before it is
__________________
necessary to apply a correction known as Yate’s correction.
__________________
The formulae for calculation for c2 becomes.
__________________
__________________
X2 =
n bad − bcg − n 2s 2

b gb g
with d. f . = 2 − 1 2 − 1 = 1
__________________ R1 R2 C1 C2
__________________
__________________
The application of this formula is illustrated by an example
given below.
__________________
__________________ Example 8.1
In testing for any relationship between the opinion on
nuclear disarmament and the sex of the person, let us assume
that 100 persons including 60 males were asked about their
opinion and their responses were classified into two
categories, yes and no, as follows:
Male Female Total

Favour 35 25 60

Against 25 15 40

Total 60 40 100

Solution
i. Let null hypothesis, H0: there ‘s no difference of opinion
between male and female.

Then alt hypothesis, H1: there is significant difference


of opinion between male and female opinion.

Method 1. Approximation method with Yates correction.

ii. Calculate c = 2
a f
N ad − bc − n / 2
2

R 1 R 2 C1 C 2
Here, N=100
a=35 R1= a+b=60
b=25 R2= c+d=40
c=25 C1= a+c=60
d=15 C2= b+d=40
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 231

X2 =
100 n b35 × 15 − 25 × 25g − 100 2s2 Notes
__________________
60 × 40 × 60 × 40

l q2
__________________
100 525 − 625 − 50
= __________________
60 × 40 × 60 × 40
__________________
100 × 22500
= __________________
60 × 40 × 60 × 40
225 __________________
=
576 __________________
= 0.390 __________________
__________________
iii. Tabulated X2 at 5% level of significance __________________

= 30841

df = (r-1) (c-1)

= (2-1) (2-1)

=1

iv. As X2 calculated < X2 tabulated

Null hypothesis stands accepted.

i.e., there’s is no difference of opinion between male and


female

Method II

∑=
b fo − f e g2
fe

where, fo= observed frequency

fe= expected frequency

R t × Ct
i. Calculate frequencies by this formula; Gt

Where Rt = Total of row

Ct = Total of column

Gt = Grand total
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
232 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Male Female

__________________ Favour 35 25 60
Against 25 15 40
__________________
Total 60 40 100
__________________
__________________ 60 × 60
Expected frequency for cell (1,1) = = 36
__________________ 100
__________________
60 × 60
__________________ Expected frequency for cell (1,2)= = 24
100
__________________
__________________ 40 × 60
Expected frequency for cell (2,1) = = 24
__________________ 100

40 × 40
Expected frequency for cell (2,2) = = 16
100

ii. Let null hypothesis H0: there is no difference of opinion


between male and female.

Then alt hypothesis H1: There is significant difference


of opinion between male and female.

iii. Calculate

X2 = ∑
bfo − fe g 2

fe

=
a35 − 36f + a25 − 24f + a25 − 24f + a15 − 16f
2 2 2 2

36 24 24 16

= 0.028 + 0.041 + 0.041 + 0.062

= 0.172

d.f = (c-1) (r-1) = (2-1) (2-1)=1

Tabulated x2 at 5% level of significance

= 30841

df = (r-1) (c-1)

= (2-1) (2-1)

=1
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 233
iv. As, X2 cal < X2 tabulated Notes
__________________
Null hypothesis stands accepted.
__________________
i.e., there is no difference of opinion between male and __________________
female. __________________
__________________
Measures of Associations
__________________
Research questions in business frequently revolve around __________________
the study of relationship between two or more variables. __________________
Various objectives may be served by such an analysis. The
__________________
strength, direction, shape and other features of the
__________________
relationship may be discovered. Or tactical and strategic
questions may be answered by predicting the values of one
variable from those of another.

With correlation, one calculates an index to measure the


nature of the relationship between variables calculated. With
regression, an equation is developed to predict the values of
the dependent variable. Both are affected by the assumptions
of measurement level and the distributions that underline
the data.

Bivariate Correlation Analysis


Bivariate correlation analysis differs from non-parametric
measures of association and regression analysis in two
important ways. First, parametric correlation requires two
continuous variables measured on an interval or ratio scale.
Second, the coefficient does not distinguish between
independent and dependent variables. It treats the variables
symmetrically since the coefficient r xz has the same
interpretation as ryx.

Linear Regression Equation


The pattern of the scatter diagram, in most cases, is lineally
related. The line which runs through most of the points, (if
not all), is known as the “line of regression”. This line
should be the most representative of the data, i.e.,
most suitable and is thus popularly known as the “line of
best fit”.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
234 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The best line would be the one which passes through all the
__________________ points. But this is not possible in most situations. The next
__________________ step is to find the line which is closest to all the points, i.e.,
__________________ the distance between the line and points is minimum. As
__________________ some points shall be above the line and some below the line,
the distance above line ‘time’ and below ‘we’, therefore, we
__________________
take the square of donations (which will be time both cases)
__________________
and hence the line of best fit would be one which gives the
__________________
minimum differences
__________________
Y=a+bx
__________________
__________________ where Y = Dependent variables

X = Independent variables

a and b are costants which determine the completed line.

Example 8.2
A researcher wants to find if there is a relationship between
the heights of sons and fathers. In other words, do tall father
have tall sons?

Their heights in inches is given below in an ordered


arrangement.
Father (X) Son (Y)
63 66
65 68
66 65
67 67
67 69
67 69
68 70

a. For this data complete the regression line.

b. Based upon the relationship between the heights, what


would be the estimated height of the son if the father
height is 70”.

Solution
We can solve the problems under:

Let the line be Y = a + b x


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 235

b=
b g b gb g
n Σxy − Σx Σy
Notes

n e Σx j − b Σx g
__________________
2 2
__________________
= Y − bX __________________
__________________
__________________
X Y X2 XY Y2
__________________
63 66 3969 4158 4356
__________________
65 68 4225 4420 4629
__________________
66 65 4356 4290 4225
__________________
67 67 4489 4489 4489
__________________
67 69 4489 4623 4761

68 70 4624 4760 4900

å X = 396 å Y = 405 å X2 = 26152 å XY = 26740 å Y2 = 27355

b g
6 26740 − 396 405 b g = 160440 − 16038
b=
b g
6 26740 − 396 396 b g 156912 − 156816
60
= = 0.625
96
`
a=
405
6
− 0.625 396 6 b g
= 67.5 − 41.25 = 26.25

Hence, the regression line would be

Y =a+bx

= 26.25 + 0.625 x

If fathers height is 70”, height of son will be (X – 70)


Y = 26.25 + .625 (70)
= 26.25 + 43.75 = 70

Expected height of son is also 70".

Standard Error of the Estimate

How good is this fit?


The closer these values are to each other, the better the fit.
Accordingly, a measure of the variability of the scatter
around the regression line would determine the reliability
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
236 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes of this estimate. The smaller this estimate, the more


__________________ dependable the prediction will be.
__________________
For our problem stand and error of the estimate would be:
__________________

∑ b yg c∑ yh − bc∑ xyh
__________________ 2
−a
__________________ Syx =
n−2
__________________
__________________
__________________
a f
27355 − 26.25 405 − 0.625 26740 a f
4
__________________
__________________ 11.25
= 2.8125 = 1.678
4

The Coefficient of Determination (r2)


The square of the coefficient of correlation (r) is called the
coefficient of determination (r2).

It is a more precise measure of the straight relationship


between the two variables and lends itself to more precise
interpretation because it can be represented as a proportion
or as a percentage. The coefficient of determination can be
defined as a proportion of the variation in the dependent
variable y, that is explained by the variation in independent
variable x, in the regression model.

Explained var iation


r2 =
Total var iation

=
a ∑ Y + b ∑ XY − b∑ Y g / n 2

∑ aY f − b ∑ Y g / n
2 2

For the above example under discussion

2
r =
a f a f a f
26.25 405 + 625 26340 − 405 / 6
2

27355 − a 405f / 6
2

10631.25 + 16712.5 − 27331.5


=
27355 − 27377.5
6.25
=
17.5
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 237
= 0.357 ( 35.7% of Variation in y) Notes
__________________
and
__________________
2
r = r = 0.357 __________________
__________________
= 0.597
__________________
Rank Correlation __________________
__________________
For the two variables x and y whose distribution is unknown,
the degree of association between x and y is ascertained by __________________

the Spareman’s rank correlation. __________________


__________________
As the name suggests, this method is based on the rank (or
order) of the observations rather than on a specific
distribution of x and y.

In marketing research, the ranking or ordering of alternative


preferences is quite common.

The method is very handy and involves simple calculations


only.

The rank correlation coefficient is a measure of correlation


that exists between two sets of ranks. In other words, it is a
measure of association that is based on the ranks of the
observations and not on the numerical values of the data. It
was developed by the famous statistician, Charles Spearman,
in the early 1900s and as such, it is also known as Spearman’s
rank correlation coefficient.

For calculating Spearman’s rank correlation coefficient, first


of all the actual observation has to be replaced by their ranks,
giving rank 1 to the highest value, rank 2 to the next highest
value and following this very order, ranks are assigned for
all values.

The second step is to record the difference between ranks


(or ‘d’) for each pair of observations. Spearman’s rank
correlation coefficient, rs, is worked out as under:

6∑ d2
rs = 1 −
d
n n2 − 1 i
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
238 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Where n = number of paired observation.


__________________
The value of Spearman’s rank correlation will always vary
__________________
between + 1 indicating a perfect positive correlation and –
__________________
1, indicating perfect negative correlation between two
__________________ variables.
__________________
This is a measure of correlation that exists between two sets
__________________
of ranks, a measure of degree of association between
__________________
variables that we would not have been able to calculate
__________________
otherwise.
__________________
__________________
This method helps to simplify the process of computing
correlation coefficient from a very large set of data for each
of the two variables. We can compute measure of association
that is based on ranks of observations, not the numerical
values of the data.

These values are rather quick and easy to use and can be
applied to ordinal or nominal scale data.

They do not presuppose any particular distribution and


consequential assumptions.

There are many situations in which the various assumptions


required for standard tests of significance cannot be met.
We can use this method under such situations as it is easier
to explain and understand.

Example 8.3
Suppose 10 salesman employed by a company were given a
month’s training. At the end of the specified training they
took a test and were ranked on the basis of their
performance. They were then posted to their respective
areas. At the end of six months they were rated in respect of
their sales performance. The data are given in the table
below:
Salesman Rank obtd Rank on the Difference Difference
in training basis of sales squared
performance

# X Y d d2

1 4 5 -1 1
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 239
2 6 8 -2 4 Notes
__________________
3 1 3 -2 4
__________________
4 3 1 +2 4
__________________
5 9 7 +2 4
__________________
6 7 6 +1 1
__________________
7 10 9 +1 1
__________________
8 2 2 0 0 __________________
9 8 10 -2 4 __________________
10 5 4 +1 1 __________________
__________________

Here åd2 = 24. Using the formula, we get

rs = 1 −
6∑ d2
=
a f 1 − 144
1 − 6 24
nd n − 1i 10d10 − 1i
2 2
990

= 1–.145

= 0.855

This shows that there is a very high correlation between


performance in training and the sales performance of the
ten salesmen.

Analysis of Variance (ANOVA)


Analysis of variance (ANOVA) enables us to infer whether
population from which we have drawn more than two
samples are having the same mean values. By testing the
significance of difference between more than two sample
means we make such inferences. To test the equality of
variances, we use ‘F’ test. By comparing the observed ‘F’ value
with table value of ‘F’ and analyzing the significance of
differences (if any) we test our hypothesis. Before elaborating
on the ‘F’ test, it is important for us to understand the
principles and techniques of analysis of variance (ANOVA).
Only after analysis of variance, we can submit it to ‘F’ test to
make necessary inferences.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
240 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes To test the equality of variances in such cases too,we use ‘F’
__________________ test, the formula for which is:
__________________ σ 2si
F= =with d.f. (n1–1) for sample 1.
__________________ σ 2s 2
__________________ (n2 – 1) for sample 2.
__________________
__________________
σ 2s 1 = variance of sample 1

b g
__________________ 2
∑ X li − X 1
bn − 1g
__________________
calculated as
__________________ 1

__________________
We can also use n1 when sample size is 30 or more.

σ 2x 2 =variance of sample 2

b g
∑ X 2i − X 2
2

calculated as =
bn − 1g
2

X1i = is the observed frequencies of sample 1.

X2i = observed frequencies of sample 2. and X 1 and X 2 are


the respective mean values of samples 1 and 2.

But ANOVA technique is important to compute variances


when we want to compare more than two samples to infer
whether the said samples have been drawn from population
having the same mean values. Without going into the trouble
of considering all possible combinations of populations,
ANOVA technique enables the researcher to investigate the
mean differences of all the populations simultaneously.

We have introduced the term dependent variable and


independent variable with an example earlier. ANOVA
enables a researcher to investigate factors which influence
a dependent variable. Let us assume productivity rate as a
dependent variable. We know that productivity usually
depends on multiple factors like level of technology used in
production, workers’ skill, incentives and gain sharing,
training, supervision and industrial relations. For us all
these factors are independent variables. Now a researcher
may be interested in knowing the differences amongst these
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 241
factors, i.e. which one of the factors is more effective in raising Notes
productivity (by studying their respective differences at __________________
various level of significances)or he may be interested in __________________
investigating the differences amongst various categories of __________________
same factor (for example various methods on incentive and __________________
gain sharing) which influence productivity. When we
__________________
investigate various categories of one factor only, we carry
__________________
out one-way ANOVA.
__________________
But when we are interested in investigating two factors at a __________________
time, we carry out two-way ANOVA. Likewise with the help __________________
of ANOVA, we can investigate, in general any number of
__________________
factors and also various categories within each factor to
study their respective influences. Thus we find following
steps are usually involved in analysis of variance:

i. Estimation of population variance based on variance


between the sample means.

ii. Estimation of population variance based on variance


within the sample means.

iii. The third step involve is to compare these variances to


get F value.

iv. The third step is to compare the obtained F value with


the respective F table value at given degrees of freedom
and to judge the differences at desired level of
significances.

The formula for computing ‘F’ value takes the following shape
MS between
F=
MS within
Now unless the above steps are explained with a practical
example, it would be difficult for readers without knowledge
of statistics to understand F test in its true spirit. Let us
first have an example of one-way ANOVA.

In a company there are four shop floors. Productivity rate


for three methods of incentives and gain sharing in each
shop floor is presented in the given table. Analyze
whether various methods of incentives and gain sharing
differ significantly at 5% and 1% F-limits.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
242 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
Shop floor Productivity rate data for three methods of
__________________ incentives and gain sharing

__________________ X1 X2 X3

__________________ 1 5 4 4

__________________ 2 6 4 3
3 2 2 2
__________________
4 7 6 3
__________________
__________________
Solution
__________________
__________________ Step 1: Calculate mean of each of the three samples (i.e., x1,
__________________
x2 and x3, i.e. different methods of incentive gain sharing).

5+6+2+7
X1 = =5
4

4+3+2+3
X2 = =3
4

4+3+2+3
X3 = =3
4

X1 + X 2 + X 3
Step 2: Calculate mean of sample means i.e., X =
K

where, K denotes No. of samples

=
5+3+3
3
b
= 4 approximated g
Step 3: Calculate sum of squares (s.s.) for variance between
and within the samples.

ss between = n1 x 2 − x b g 2
b
+ n2 x2 − x g 2
d
+ n3 x − x i 2

b
ss within = ∑ x1i − x1 g 2
b
+ ∑ x 2i − x 2 g 2
b
+ ∑ x 3i − x 3 g 2

Sum of squares (ss) for variance between samples is obtained


by taking the deviations of the sample means from the mean
of sample means ( x ) and by calculating the squares of such
deviation, which are multiplied by the respective number of
items or categories in the samples and then by obtaining
their total. Sum of squares (ss) for variance within samples
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 243
is obtained by taking deviations of the values of all sample Notes
items from corresponding sample means and by squaring __________________
such deviations and then totalling the. For our illustration __________________
then __________________

ss between = 4(5 – 4)2 + 4 (4 – 4)2 + 4 (3 – 4)2 __________________


__________________
= 4+0+4= 8
__________________
__________________

{1b54444444
− 5g + b6 − 5g + b2 − 5g + b7 − 5g }
2 2

4244444444 3
2 2 __________________

b g 2
__________________
∑ x1 i − x1
__________________
{b g b
2
+ 4−4 + 4−4 + 2−4 + 6−4
1444444442444444443
g b
2
g b
2
g}
2

ss within = b
∑ x 2 i − x2 g 2

{b g2 b
+ 4−3 + 3−3 + 2−3 + 3−3
1444444442444444443
g2 b g2 b g2 }
b
∑ x 3 i − x3 g 2

= (0 + 1 + 9 + 4) + (0 + 0 + 4 + 4) + (1 + 0 + 1 + 0)

= 14 + 8 + 2

= 24

Step 4: ss. of total variance which is equal to total of s.s.


between & ss within and is denoted by formula as follows:

e
∑ X ij − X j 2

where

i = 1.23

j = 1.23

for our example, total ss will thus be:

{b5 − 4g + b6 − 4 g + b2 − 4g + b7 − 4g } 2 2 2 2

+ {b 4 − 4 g + b 4 − 4 g + (2 − 4) + (6 − 4) }
2 2 2 2

+ {b 4 − 4 g + b 3 − 4 g + b 2 − 4 g + b 3 − 4 g }OP
2 2 2 2
Q
= nb1 + 4 + 4 + 9g + b0 + 0 + 4 + 4 g + (0 + 1 + 4 + 1s
= 18 + 8 + 6 = 32
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
244 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes We will, however, get the same value if we simply total


__________________ respective values of ss between and ss within. For our
__________________ example ss between is 8 and ss within is 24, thus ss of total
__________________ variance is 32 (8+24).
__________________ Step 5: Ascertain degrees of freedom and mean square (MS)
__________________ between and within the samples. Degrees of freedom (df)
__________________ for between samples and within samples are computed
__________________ differently as follows:
__________________
For between samples, df, is (k–1), where ‘k’ represents
__________________ number of samples (for us it is 3). For within samples df is
__________________ (n–k), Where ‘n’ represents total number of items in all the
samples (for us it is 12).

Mean squares (MS) between and within samples are


computed by dividing the ss between and ss within by
respective degrees of freedom. Thus for our example:

ss between
i. MS between =
(K − 1)

8
=4
2

where (K–1) is the df.

ss within 24
ii. MS within = =
(n − k) 9

= 2.67

where (n–k) is the df.

Step 6: Now we will have to compute F ratio by analyzing


our samples. The formula for computing ‘F’ratio is:
MS between
MS within

Thus for our example F ratio

4.00
= = 1.5
2.67
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 245
Step 7: Now we will have to analyze whether various Notes
methods of incentives and gain sharing differ significantly __________________
at 5% and 1% ‘F’ limits. For this, we need to compare __________________
observed ‘F’ ratio with ‘F’ table values. When observed ‘F’ __________________
value at given degrees of freedom is either equal to or less __________________
than the table value, difference is considered insignificant.
__________________
In reverse cases, i.e., when calculated 'F' value is higher than
__________________
table-F value, the difference is considered significant and
__________________
accordingly we draw our conclusion.
__________________
For example, our observed ‘F’ ratio at degrees of freedom __________________
(V1* & V2**, i.e., and 9) is 1.5. The table value of F at 5% level
__________________
with df 2 and 9 (v1 = 2, v2 = 9) is 4.26. Since the table value is
higher than the observed value, difference in rate of
productivity due to various methods of incentives and gain
sharing is considered insignificant. At 1% level with df 2
and 9, we get the table value of F as 8.02 and we draw the
same conclusion.

We can now draw an ANOVA table as follows to show our


entire observation.

Variation SS df MS F-ratio Table value


of F

5% 1%

Between 8 (k–1)= ss between MS between F (v1, v2) F (v1, v2)

sample (3–1)=2 (k–1) MS within =F (2,9) =F(2,9)

=8/2 = 4 =4/2.67 = 4.26 8.02

Within

simple 24 (n–k)= ss.within =1.5

(12–3) (n–k)

=9 = 24/9

= 2.67

Two way ANOVA technique is used when our data are


divided to both columns and rows and we way to study the
effects. Let us assume that our dependent variable
productivity rate is classified on the basis of incentives and

* v1=d.f. for greater variance (between samples).


** v2=d.f. for smaller variance (within samples).
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
246 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes gain sharing and on level of technology used in production


__________________ (independent variables). To study the effects of independent
__________________ variables on dependent variable, under the given
__________________ assumptions, we use two way classification of ANOVA.
__________________ Again there may be repeated measurements of all categories
__________________ where we need to study even the interaction variation or
__________________ there may not be any such repeated measurement. Based on
__________________ such differences in data collection and presentation, our
__________________ computational method differs.
__________________ When we do not have repeated measurement, we follow the
__________________ following steps for two-way classification of ANOVA:

Step 1: Calculate total values of individual items in all


samples and indicate it by T. P So T stands for total value of
all samples.

Step 2: Calculate correction factor using following formula:

T2
where n denotes number of samples.
n

Step 3: Calculate total ss and subtract the correction factor


to obtain sum of squares of deviation for total variance. Total
T2
ss is denoted by ∑ χ ij −
2
. Thus our formula for step 3 takes
n
the following shape.

T2
∑χ 2
ij

n
Step 4: Compute total ss between columns and subtract the
result from the correction factor. Total ss is computed by
obtaining the square of each column total and subsequently
dividing such squared values by number of items and totalling
the result. Thus we denote step 4 by following formula
F Tj I

H K−T 2

, ( where j=1,2,3,.....i.e. individual column values


nj n
nj denotes total number of items in the column).

= sum of squares of deviations of variance between columns.


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 247
Step 5: Compute total ss between rows and subtract the Notes
result from correction factor to obtain sum of squares of __________________
deviation for variance between rows. Total ss between rows __________________
is calculated by squaring each row total by dividing squared __________________
values by respective number of items in the row and finally __________________
by taking the total of the results. Thus we can denote step-5
__________________
by the following formula:
F Tj I
__________________


H K−T , where i = 1,2,3 (individual row values) n j
2 __________________
__________________
nj n
__________________
denotes total number of items in the row.
__________________
Step 6: Compute sum of squares of deviations for residual
or error variance by subtracting the sum of ss between
columns and ss rows variance from total variance (s.s total).

SS for residual variance= Total ss – (ss. between columns+ss.


between rows)

Step 7: Next task for us is to to ascertain the degrees of


freedom and mean square. Mean Square (MS) as we know is
obtained by dividing ss values by respective degrees of
freedom. In two way ANOVA, we need to compute three MS
as under:
i. MS between column variance

ss between columns
=
(c − 1)
where (c–1)is the df

C= total number of columns.

ii. MS between row variance

ss between rows
=
(r − 1)
where (r–1) is the d.f.

r = total number of rows.

iii. MS of residual or error variance

ssresidual
=
(c − 1)(r − 1)
where (c–1) (r–1) is the df.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
248 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Step 8: This step involves computation of F ratio. Here also


__________________ for two-way ANOVA, we will have to compute two F-ratios
__________________ as under for having two independent variables to measure
__________________ the significance of differences for each independent variable
__________________ on dependent variable.
__________________ F ratio for column independent variation
__________________
MS between columns
__________________ =
__________________ MS residual
__________________
F-ratio for row independent variation
__________________

MS between rows
=
MS residual

To compare table value of F with our observed F ratios for


drawing necessary conclusion at defined level of significance,
we consider respective df of SS between columns and df of
MS residual to find out table value of F for column
independent variation and df of SS between rows and df of
MS residual for row independent variation.
Let us now frame a problem to study the effects of incentive
and gain sharing and level of technology (independent
variables) on productivity rate (dependent variable).
Productivity Rate Data of Workers of M/s. XYZ & Co.

Incentives and
gain sharing A B C
Level of
technology
W 4 3 3
X 5 3 2
Y 1 1 1
Z 6 5 2

Solution
1. Total values (T) of individual item = 36, n = 12

(T) 2 36 × 36
2. Correction factor = =
n 12

= 108
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 249
3. Total ss = (16+9+9+25+9+4+1+1+1+36+25+4) Notes
__________________
= 140 – 108 =32
__________________
4. ss between columns: __________________

L 16 × 16 + 12 × 12 + 8 × 8 OP − 108
__________________

=M
N 4 4 Q
__________________
4 __________________

5. ss between rows: __________________


__________________

LM 10 × 10 + 10 × 10 + 3 × 3 + 13 × 13 OP − 108 __________________
=
N 3 3 3 3 Q __________________
LM + + + OP − 108
100 100 9 169
=
N3 3 3 3Q
= [33.33 + 33.33 + 3 + 56.33] – 108

= 126 – 108

= 18 (after adjusting fraction)

6. ss residual:

= Total ss – (ss between column + ss between rows)

= 32 – (8+18) = 6.

Now we need to set up ANOVA table

Variation source SS d.f M.S F ratio 5% 1%

8 4
Between columns 8 (c–1) =4 =4 F (2, 6) F (2, 6)
2 1
=2 = 5.14 = 10.92

18 6
Between rows 18 (r–1) =6 =6 F (3, 6) F (3, 6)
3 1
=3 = 4.76 = 9.78

6
Residual 6 (c–1) x =1
6
(r–1)
=6
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
250 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes From the ANOVA table, we find that differences related to


__________________ varieties of incentives and gain sharing are insignificant at
__________________ 5% level as the calculated F-ratio, i.e., 4 is less than table
__________________ value of F, which is 5.14. However differences are significant
__________________ for different levels of technology at 5% level as the observed
F ratio is higher than table value of F. At 1% level, however,
__________________
differences are insignificant.
__________________
__________________ Now coming to the problem of two-way classification of
__________________ ANOVA in cases of repeated measurements apart from the
__________________
earlier steps, we need to calculate interaction variation also
as it enables us to measure the interrelationship among the
__________________
two different classifications. For example, we may classify
our productivity rate data in two groups of workers, assuming
our company is having two separate production units.

We follow upto Step-5, as elaborated earlier for two-way


ANOVA without repeated measurements. Rest of the steps
are as follows:

Step 6: For having two different classifications of data (for


us productivity rate), apart from computing total SS, SS
between columns and between rows, we need to work out ss
within samples, i.e. sum of squares of deviations between
two samples classifications.

Step 7: Our next task is to compute ss for interaction


variation (within sample error). For this we subtract total
ss from the sum of ss between columns, ss between rows and
ss within samples.

Step 8: In this step, we need to calculate degrees of freedom


and Mean Squares (MS) For this ANOVA too we should
compute three mean squares as under:
ss between columns
i. MS between column variation =
(c − 1)
Where c represents total number of column and (c–1)
represent degree of freedom.
ss between rows
ii. MS between rows =
(r − 1)
where ‘r’ represents total number of rows and (r–1)
represents degree of freedom.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 251
iii. MS for interaction variation Notes
__________________
ss interaction variation
= __________________
(c - 1)(r - 1)
__________________
where (c–1) and (r–1) are the degrees of freedom.
__________________
iv. MS for within samples variation __________________

ss within samples __________________


=
(n - k) __________________
__________________
where ‘n’ is the total number of samples (all observations
are included ) and ‘k’ represented total number of samples __________________

group. __________________

Step 9: This final step involves calculations of ‘F’ ratio. For


having repeated values, in addition to measurement of
significance of differences for each independent variables
on dependent variable, we need to measure significance of
such differences between groups, i.e., how different groups
of people (based on our sample observations) are affected
differently (it terms of variation in productivity rate for our
example) for independent variables. Thus we compute
following F ratios :

i. ‘F’ ratio for column independent variation

MS between rows
=
MS within samples

ii. ‘F’ ratio for row independence variation

MS between rows
=
MS within samples

iii. F ratio for interaction variation

MS for interaction variation


=
MS within samples

We then compare the observed ‘F’ ratios with table values


of ‘F’ at given df and draw our conclusions. Degree of
freedom is determined by same principles followed
earlier.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
252 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Multivariate Analysis


__________________
__________________ Definition of Multivariate Analysis
__________________ Multivariate analysis or multivariate techniques may be
__________________ defined as the collection of methods for analysing data in
__________________ which a dependent variable is represented in terms of
__________________ several independent number of observations which are
__________________ available to define such relationship. In brief, techniques that
__________________
take account of the various relationships among variables
are termed multivariate analysis or multivariate techniques.
__________________
Mathematically, multivariate analysis was defined by
__________________
Takeuchi, Yanai and Mukherjee as, “forming a linear
composite vector in vector space, which can be represented
in terms of projection of a vector on to certain specified
subspaces.”

A multivariate method

Are some of the variables


dependent on others?

Yes No

Dependence Interdependence
methods methods

How many variables Are inputs metric?


are dependent?

Several Yes No
One

Is it metric Are they


metric?

Yes No Yes No

Multivariate Metric multi-


Multiple Factor Cluster
analysis of dimensional
regression analysis analysis
variance scaling

Multiple discriminant
analysis Canonical analysis

Non-metric Conjoint Latent structure


scaling measurement analysis

Figure 8.3
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 253
Objectives of Multivariate Analysis Notes
__________________
The basic objectives of multivariate analysis are:
__________________
(i) To represent the collection of large set of data in a __________________
simplified way, by transforming large number of __________________
observations into smaller composite scores. __________________
__________________
(ii) To predict the variability of the dependent variable
__________________
based on its covariance with all the independent
__________________
variables.
__________________
(iii) To classify individuals or objects into one of the two or __________________
more mutually exclusive and exhaustive groups on the
basis of a set of independent variables.

Advantages of Multivariate Analysis

The main advantage of multivariate analysis is that since it


takes into account more than one factor/element of
independent variables which affect the variability of
dependent variable, the conclusions drawn are more
accurate. The conclusions are more realistic and nearer to
the real life situation.

Disadvantages of Multivariate Analysis

(i) It requires rather complex computations to arrive at a


satisfactory conclusion.

(ii) Due to the above fact that a large number of


observations for large number of variables need to be
collected and tabulated, it is a rather time consuming
process.

(iii) Obviously due to (i) and (ii) above, at times multivariate


analysis proves to be an expensive proposition, in terms
of cost.

(iv) Specialized trained staff is required to process and


analyze the complex data utilizing multivariate
techniques.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
254 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Applications of Multivariate Analysis


__________________
These techniques are successfully employed in the following
__________________
areas:
__________________
__________________ (a) Econometrics or decision-making in economics:
__________________
Such as impact of inflation, money circulation, lowering
of tariffs, etc., on price rise.
__________________
__________________ (b) Sociological decision-making: Such as divorce rates,
__________________ and their cause-effect relationships, with marriage,
__________________ social demographics and income levels.
__________________ (c) Agrarian predictions: Such as impacts of rain,
fertilizers and mechanization on agricultural yields per
acre.

(d) Drug testing: Such as impact of new drugs on the main


disease and other side effects.

(e) State fiscal decision-making: Such as impact of tax


structure, duties, penalty rates, etc., on the government
revenues.

(f) Industrial decision-making: Such as plant location


which depends on infrastructure, availability of raw
materials, distribution channel, etc.

Factor Analysis
It is one of the more popular "analysis of interdependence"
techniques. In studies of interdependence, all the variables
are on an equal footing, and the analysis is concerned with
the whole set of relationships among the variables that
characterize the objects.

Unlike regression or discriminant analysis, factor


analysis would focus on the whole set of interrelationships
displayed by the variables. At the conceptual level, a
factor is a qualitative dimension of the data that attempts
to depict the "way in which entities differ, much as the
length of an object or the flavour of a product defines a
qualitative dimension on which objects may or may not
differ."
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 255
Factor analysis is a procedure that takes a large number of Notes
variables or objects and searches out factors in common which __________________
account for their inter-correlation. For example, we can __________________
attribute the high association between grades in computer __________________
courses to the factor of intelligence or the association __________________
between certain attributes of coffee and factor of acidity.
__________________

Application of Factor Analysis in Marketing Research __________________


__________________
There are a number of applications in marketing research.
__________________
These includes data reduction, structure identification and
__________________
scaling.
__________________
Data Reduction: Factor analysis is used for reducing a mass
of data to a manageable level. For example, the marketing
researcher who has collected data on 40 attributes of a
brand. This analysis and understanding of this data may
be aided by reducing 30 attributes and having only 10
attributes.

Factor analysis has been employed to “purify” original sets


of scale items by isolating those items that do not reflect a
common core and also to name the dimensions captured by
measure. It is also used in the study of lifestyles and
psychographic research problems in which it is used to
develop consumer profiles that reflect people’s attitudes,
activities, interests, opinions, perceptions and preferences,
so as to better predict their consumption and purchase
behaviour. One of the major impetuses of lifestyles and
psychographic research has been the generally inadequate
nature of market segments defined using traditional
demographic measures.

Structure Identification: Factor analysis may be used to


understand the basic structure underlying a set of measures.
For example, the above 30 attributes may be reduced to ten
factors identified by the researcher.

Scaling: A marketing researcher may develop a scale on a


research topic. One problem that may occur in developing
any scale is in weighing the variables being combined to form
the scale.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
256 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Limitations of Factor Analysis


__________________
(1) The factor analysis is not strongly supported by
__________________
statistical methods.
__________________
__________________ (2) The different results can be obtained by using loading
__________________
factors in different manners.
__________________ (3) A large number of attributes are required to study factor
__________________ analysis.
__________________
(4) In some cases, complex factors are used which
__________________
respondents find difficult to understand.
__________________
14 coffee flavours – Unpleasant flavour
Stagnant, muggy taste – Sparkling, refreshing taste
Mellow taste – Bitter taste
Cheap taste – Expensive taste
Comforting, harmonious, smooth, friendly taste – Irritating
Discordant, rough, hostile taste
Dead, lifeless, dull taste – Alive, lively, peppy taste
Tastes warmed over – Tastes just brewed
Hearty, full-bodied, full flavour – Watery, thin, empty flavour
Pure, clear taste – Muddy, swampy taste
Raw taste – Roasted taste
Fresh taste – Stale taste
Overall preference: Excellent quality – Very poor quality

Note: Ten blank boxes separated each set of opposing statements. Subjects checked
the position that came closest to describing how they felt towards the product.

Figure 8.4: Coffee Attributes and Intercorrelations Among


Attribute Ratings
Table 8.1: Intercorrelations Among Attribute Ratings

1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Pleasant flavor 1.00 .76 .81 .79 .83 .81 .74 .66 .65 .71 .76 .65 .71 .75

2. Sparkling taste 1.00 .78 .85 .77 .87 .83 .65 .70 .78 .85 .69 .74 .83

3. Mellow taste 1.00 .77 .85 .81 .77 .60 .65 .64 .75 .69 .69 .74

4. Expensive 1.00 .78 .87 .83 .76 .69 .81 .81 .64 .71 .87
taste

5. Comforting 1.00 .82 .77 .66 .60 .69 .82 .69 .69 .74
taste

6. Alive taste 1.00 .88 .70 .74 .80 .81 .65 .77 .87

7. Tastes like real 1.00 .67 .76 .75 .79 .62 .76 .87
coffee

8. Deep distinct 1.00 .51 .84 .70 .54 .59 .70


flavour

9. Tastes just 1.00 .67 .65 .67 .80 .75


brewed

10. Hearty flavour 1.00 .83 .65 .72 .76

11. Pure clear 1.00 .83 .65 .72 .76


taste

12. Roasted taste 1.00 .78 .61

13. Fresh taste 1.00 .73

14. Overall 1.00


preference

Mean rating* 4.5 4.3 4.4 4.6 4.4 4.2 4.2 4.3 4.3 4.2 4.3 4.4 4.6 6.9
Standard deviation 1.6 1.3 1.4 1.4 1.3 1.4 1.6 1.5 1.5 1.5 1.4 1.2 1.4 2.7

* The 10 scale categories were assigned successive integers, beginning with 1 at the favourable side of the scale. Thus,
ratings could vary from 1 (very good) to 10 (very bad) on an attribute.
Source: Updated to 1995 from Bishwa Nath Mukherjee “A Factor Analysis of Some Qualitative Attributes of Coffee”,
Journal of Advertising Research, Vol. 5, p. 36, March ’95. Used with permission.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 257
Four factors underlying the 14 attributes. The elements in Notes
this matrix listed under the factors are called unrotated __________________
factor loadings. The loadings measure which variables are __________________
involved in which factor pattern, to what degree and in what __________________
direction. They can be interpreted as correlation coefficients. __________________
The square of the loading equals the proportion of the
__________________
variation that a variable has in common with an unrotated
__________________
factor.
__________________
Another way to conceptualize this relationship is to __________________
remember that a loading is a correlation coefficient between __________________
a variable and a factor. In essence, when we square a loading
__________________
we are calculating a coefficient of determination, h2, between
a variable and a factor. Thus, the squared loading represents
the amount of shared variation between a variable and a
factor.

The h2 measures are called communalities. Communality is


the proportion of a variable’s total variation that is involved
in the factors. Mathematically h 2 equals the sum of the
squared loading of a variable on all factors. For example, for
attribute 1:

h2 = (.86)2 + (-.01)2 + (-.20)2+ (.04)2 = .78

Communality may be interpreted as a measure of uniqueness.


By subtracting h2 from 1.0 the degree to which a variable is
unrelated to the others may be calculated. Here we note that
78 per cent of the variation in scores on attribute 1 can be
predicted from the other variables, leaving 22 per cent
uniquely related to this attribute.

To get the percentage of total variance in the data explained


by the four factors, we simply calculate H where

sum of all h 2s
H= × 100
number of variables
11.61
= × 100
14
=83%

This value is called the common variance explained by the


factors.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
258 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes To calculate the amount of variation in the data accounted


__________________ for by a factor, we square each loading for a factor, add and
__________________ then divide the result by the number of variables. For
__________________ example, for factor 1 the value is
__________________ (.86)2 + (.91)2 + (.86)2 + ......... + (.90)2 = 10.42
__________________
This value is called an eigen value in the vocabulary of factor
__________________
analysis. Thus, the percentage of total variance explained
__________________
by factor 1 is 10.42/14 = 74.4 per cent. The percentage of
__________________
common variance explained by this factor is then 10.42/11.61
__________________ = 90 per cent. We note that the percentages of both common
__________________ variance and total variance are presented in Table 8.2.

In order to obtain an interpretation of the result, we examine


the rotated factors.

Table 8.3 presents one such interpretation. There the factors


have been placed with high-loading variables and each given
a creative name by the author of the study. Note that there
is no unique definition of the meaning of any factor; it is up
to the creativity of the researcher.

Table 8.2: Factor Loadings

Principal factor matrix Rotated (varimax) matrix Oblique factor matrix

I II III IV h2 A B C D A B C D

1 .86 -.01 -.20 .04 .78 .63 .38 .36 .34 .34 .01 .07 -.03
2 .91 -.01 -.01 -.09 .83 .48 .43 .53 .38 .14 .04 .23 .04
3 .86 .11 .28 .002 .83 .70 .26 .38 .36 .36 .13 -.003 -.01

4 .91 .15 -.001 -.10 .87 .46 .53 .54 .29 .16 -.05 .34 -.07
5 .87 .002 -.31 .10 .87 .74 .38 .30 .32 .47 .01 -.004 -.08
6 .93 .03 -.02 -.16 .90 .49 .43 .59 .35 .12 .07 .30 -.01
7 .90 -.02 .04 -.21 .86 .42 .38 .64 .37 .03 .11 .33 .04
8 .77 .36 .11 .16 .77 .31 .74 .27 .22 .24 -.40 .32 -.10
9 .79 -.28 .24 -.09 .76 .23 .24 .52 .62 .15 .11 .14 .37
10 .87 .25 .22 .17 .89 .28 .75 .33 .39 .14 .38 .31 .07
11 .89 .11 .05 .10 .82 .51 .55 .36 .36 .28 -.15 .17 -.01
12 .76 -.29 .04 .27 .74 .43 .28 .16 .67 .18 -.08 .18 .38
13 .84 -.27 .19 .12 .83 .33 .32 .36 .70 .01 -.03 -.001 .41
14 .90 .04 .08 -.23 .86 .38 .43 .65 .34 .002 .08 .39 .01
Percent common variance 90.0 4.1 3.3 2.6
Percent total variance 74.4 3.4 2.7 2.6

Source: Adapted and updated to 1995 from Bishwa Nath Mukherjee. “A Factor Analysis of Some Qualitative Attributes
of Coffee”, Journal of Advertising Research, Vol. 5, p. 37, March 1965. Used with permission.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 259

Table 8.3: Table Interpretation of Factors Notes


__________________

Variable Attribute Varimax Oblique __________________


__________________
Factor A (comforting quality)
1 Pleasant flavour .625 .340 __________________
3 Mellow taste .698 .359
__________________
5 Comforting taste .736 .465
11 Pure clear taste .512 .283 __________________
Factor B (heartiness)
__________________
8 Deep distinct flavour .742 .396
10 Hearty flavour .745 .380 __________________
Factor C (genuineness)
__________________
2 Sparkling taste .524 .232
4 Expensive taste .541 .334 __________________
6 Alive taste .594 .301
7 Tastes like real coffee .636 .328
8 Deep distinct flavour .268 .323
10 Hearty flavour .332 .310
14 Overall preference .653 .387
Factor D (freshness)
9 Tastes just brewed .621 .359
12 Roasted taste .670 .465
13 Fresh taste .698 .238
Source: Updated to 1995 from Bishwa Nath Mukherjee, “A Factor Analysis of Some
Qualitative Attributes of Coffee”, Journal of Advertising Research, Vol. 5, p. 37, March
1965. Used with permission.

Unfortunately, researchers can all too easily fool themselves


with wonderful sounding interpretations. Great care must
be taken in this regard.

The presentation of factor analysis has concentrated on the


nature of the input and output. There are many more issues
on all aspects of factor analysis not covered here. You should
consult a more advanced reference if you are interested.

Difficulties, Problems and Cautions


Cost: Factor analysis involves the study of various factors.
To study a large number of factors, the researcher has
to spend a lot of time. Therefore, it is considered an expensive
analysis. But now, we can use computers to get faster
results.

Reliability: A factor analysis starts with a set of imperfect


data. Data change because of changes in the sample, changes
in data gathering procedures or any of the numerous kinds
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
260 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes of measurement errors, the results of any single analysis


__________________ are, therefore, always less than perfectly dependable.
__________________
__________________ Cluster Analysis
__________________ Cluster analysis is used by the researcher to place variables
__________________ or objects into subgroups or clusters. These clusters are not
__________________ defined by the researchers but are formed by the cluster
__________________ analysis procedure itself. The following factors are
__________________ considered for cluster analysis:
__________________ 1. They form subgroupings and assign variables or objects
__________________ to these groups.

2. They take as input a matrix of association between


variables or objects.

3. They assume that natural clusters exist within the data.

Example of Cluster Analysis


A two-dimensional perceptual Figure 8.5 has been drawn
on the basis of data relating to 15 individuals A to O, on the
basis of (i) Number of vacation days (ii) Expenditure on
vacations during a given year.

I
C M
Number of vacation days

H
II
F L
N
O
G K D
III
I B A
J E

Figure 8.5: Cluster Analysis

It will be seen that there are three distinct clusters. The


first cluster comprising five individuals C, F, H, L and M
shows that although these individuals take too many
vacations, they do not spend much on their vacations. The
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 261
second cluster comprising six individuals D, G, I, K, N and Notes
O shows that they take vacations moderately and also spend __________________
moderately. Finally, the third cluster comprising three __________________
individuals A.B. and J shows that they have relatively few __________________
vacation days but spend substantially more on their __________________
vacations.
__________________

Discriminant Analysis __________________


__________________
Discriminant analysis is a technique that is appropriate with
__________________
a nominal dependent variable and interval independent
__________________
variables. Nominal dependent variables are very common in
__________________
marketing. For example, a satisfied customer, an unsatisfied
customer, service seeker customer, non-service seeker
customer. This is the reason why discriminant analysis has
received extensive application in marketing research.

The basic idea is to find out the linear combination of the


independent variables that make the mean scores across
categories of the dependent variable on this linear
combination maximally different. This linear combination is
called the discriminant function. It can be represented in
the following way:

DF = V1 X1 + V2 X2 + ... Vm Xm

The criterion used to decide when group means are maximally


different is the familiar ANOVA (F-Test) for the differences
among means thus, the V’s are derived such that

SS between
F= is maximised
SS within

Confusion matrix for customer satisfaction.


Table 8.4: Predicted Category

Actual category Satisfied Unsatisfied

Satisfied 400 20

Unsatisfied 30 50
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
262 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The output of discriminant analysis includes the values of


__________________ the V’s plus what is called a confusion matrix. This matrix
__________________ compares the category of the dependent variable that the
__________________ discriminant functions predict a subject will be in with the
__________________ category involved in it. The example of pain relievers
explains the complete analysis of discriminants.
__________________
__________________ Discriminant Analysis in Marketing: This analysis has
__________________ been used quite extensively in marketing. The following
__________________ aspects can be studied by this type of analysis:
__________________ 1. Identification of new buyer group.
__________________
2. Consumer behaviour toward new products or brands.

3. Brand loyalty study.

4. Relationship between variables.

5. Checklist of properties of new products.

The targeting strategy of various pain relievers needs to be


discriminant targeting different characteristics as perceived
by consumers. The survey of 1000 consumers have been
marked on a 5 point scale. The pain relievers characteristics
need to be grouped and marketed accordingly.
Table 8.5

I II III IV V VI VII
Pain Relievers
Characteristics
Moov Iodex Zandu Amrutanjan Medicream Relaxyl Spray

1. Joint Pains 4.2 3.4 3.3 3.3 4.3 5.0 2.3

2. Muscular Pain 3.3 4.3 3.0 2.0 4.3 4.3 4.5

3. Backache 4.2 4.2 3.0 1.5 3.2 5.3 3.3

4. Sciatica (Nerve) 3.1 3.1 2.1 2.0 3.2 4.3 4.5

5. Headache 1.0 0.0 4.2 4.0 0.0 0.0 0.0

6. Cold (Coryza) 0.0 0.0 3.3 3.0 0.0 0.0 0.0

7. Rheumatism 2.1 1.0 1.2 1.1 3.2 4.0 1.5

8. Arthritis 2.1 1.0 1.2 1.2 3.2 4.0 1.5

9. Gout 3.2 2.0 1.5 1.2 2.5 4.0 1.0

10. Smell/Aroma 4.5 1.0 2.0 2.0 4.0 4.0 3.5

11. Non Greasiness 4.0 1.0 1.5 1.5 4.0 5.0 5.0

12. Ease of Application 4.5 2.0 2.0 2.5 4.0 4.5 5.0

13. Cooling/Warming 4.0(W) 2.0(W) 3.0(C) 3.5(C) 4.0(W) 3.5(W) 5.0(C)


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 263
Discriminant analysis technique can be used by the Notes
marketing manager who wants to have an idea about the __________________
market in terms of certain profile of pain relievers from a __________________
data about particular types of pain relievers. The method __________________
derives the relative weights to the different types of pain __________________
relievers needs. Thus, the discriminant functions show the
__________________
common patterns of pain relievers for various types of pain
__________________
relievers. In this example, the market is initially segmented
__________________
on the basis of product types and then we attempt to
discriminate each of the products on the basis of the specific __________________

combination of pain relievers needs that they may be serving. __________________


5.0
__________________

4.0

3.0

2.0

1.0

0.0
J MP BA S H C R A G S NG EA C/W

Figure 8.6: Pain Reliever Needs for Moov


5.0

4.0

3.0

2.0

1.0

0.0
J MP BA S H C R A G S NG EA C/W

J=Joint Pains MP=Muscular Pain BA=Backache S=Sciaticia (Nerve)


H=Headache C=Cold (Coryza) R=Rheumatism A=Arthiritis
G=Gout. S=Smell/Aroma NG=Non Greasiness EA=Ease of Application
C/W=Cooling/Warming

Figure 8.7: Pain Reliever Needs for Iodex


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
264 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 5.0

__________________
__________________ 4.0
__________________
__________________ 3.0

__________________
__________________ 2.0

__________________
__________________ 1.0

__________________
__________________ 0.0
J MP BA S H C R A G S NG EA C/W

Figure 8.8: Pain Reliever Needs for Zandu

5.0

4.0

3.0

2.0

1.0

0.0

J MP BA S H C R A G S NG EA C/W
J=Joint Pains MP=Muscular Pain BA=Backache S=Sciatica (Nerve)

H=Headache C=Cold (Coryza) R=Rheumatism A=Arthritis

G=Gout S=Smell/Aroma NG=Non-Greasiness EA=Ease of Application

C/W=Cooling/Warming

Figure 8.9: Pain Reliever Needs for Amrutanjan


OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 265
Notes
5.0
__________________
__________________
4.0
__________________
__________________
3.0
__________________
__________________
2.0 __________________
__________________
1.0 __________________
__________________

0.0
J MP BA S H C R A G S NG EA C/W

Figure 8.10: Pain Reliever Need for Medicream

6.0

5.0

4.0

3.0

2.0

1.0

0.0
J MP BA S H C R A G S NG EA C/W
J=Joint Pains MP=Muscular Pain BA=Backache S=Sciatica (Nerve)

H=Headache C=Cold (Coryza) R=Rheumatism A=Arthritis

G=Gout S=Smell/Aroma NG=Non Greasiness EA=Ease of Application

C/W=Cooling/Warming

Figure 8.11: Pain Reliever Needs for Relaxyl


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
266 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
__________________
__________________ 6.0

__________________ 5.0 ❤
__________________
4.0
__________________
__________________ 3.0

__________________ 2.0
__________________
1.0
__________________
__________________ 0.0

J MP BA S H C R A G S NG EA C/W

J=Joint Pains MP=Muscular Pain BA=Backache S=Sciatica (Nerve)

H=Headache C=Cold (Coryza) R=Rheumatism A=Arthritis

G=Gout S=Smell/Aroma NG=Non Greasiness EA=Ease of Application

C/W=Cooling/Warming

Figure 8.12: Pain Reliever Needs for Spray

5 MOOV

IODEX
4
ZANDU

AMRUTANJAN
3
MEDICREAM

2 RELAXYL

SPRAY

0
J MP BA S H C R A G S NG EA C/W

J=Joint Pains MP=Muscular Pain BA=Backache S=Sciatica (Nerve)

H=Headache C=Cold (Coryza) R=Rheumatism A=Arthritis

G=Gout S=Smell/Aroma NG=Non Greasiness EA=Ease of Application

C/W=Cooling/Warming

Figure 8.13
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 267
Conjoint Analysis Notes
__________________
Conjoint analysis is concerned with the measurement
__________________
a combination of two or more attributes that are important
__________________
in marketing decisions from customers point of view. In
__________________
this analysis, marketing research suggests a new combination
on the basis of existing attributes. For example: A bank __________________

would like to know which is the most desirable __________________


combination of attributes to frequent borrowers – rate __________________
of interest, instalment amount, date of payment of __________________
instalment. __________________

Conjoint analysis is useful for products which can be offered __________________

in different combinations for which the consumer has to


make a trade-off.

As such it is more useful for products which require a longer


buying process for consumers.

The attributes for which the different levels are to be tested


might not be the ones that the consumer considers important
while making a choice. It is not always right to add the utility
scores across respondents unless the researcher is sure that
they belong to a homogenous group.

Numerical Example
Consider a situation in which a manufacturer different
colours of paints is interested in measuring consumers’ trade
offs among the following attributes:

*Brand *Price
Alpha 50 per litre
Gamma 60 per lire
Beta 70 per litre
*Life *Colours
3 Years Pink
4 Years Green
5 Years

Potential replacement paint purchasers are recruited


and presented with the 18 alternatives. Each respondent
is asked to rate each of the alternatives on a 0-10
scale, where 0 indicates absolutely no interest in
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
268 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes purchasing and 10 indicates extremely high interest in


__________________ purchasing.
__________________
Illustratively, Figure 8.13 shows the 18 evaluations made by
__________________
one respondent. Although a total of 3x3x3x2 = 54
__________________ combinations of attribute levels could be made up, the
__________________ respondent needs to evaluate only 18 of these. However, this
__________________ specific set of 18 should be selected in a particular way, as
__________________ will be noted later.
__________________
Figure 8.14 shows graphs of the implied values for each of
__________________ the attribute levels; these can be obtained from an ordinary
__________________ multiple regression program using dummy variable coding.
All one needs to do to estimate the respondent’s original
evaluations is to add each separate value (The regression’s
intercept term may be added in later if there is interest in
estimating the absolute level of purchase interest.) For
example, to obtain the respondent’s estimated evaluation of
card 1, one reads off the part worths:

Value of alpha = 2.4

Value of 3 years = 0

“Value” for Rs 50 per litre (a Rs 20 cost saving) = 1.6

Value for green colour = 1.2

= 55.2

In this instance we obtain an almost perfect prediction of


the person’s overall response to card 1. Similarly, we can
find the estimated total evaluations for the other 17 options
and compare them with the respondent's original
evaluations.

The regression technique guarantees that the (squared)


prediction error between estimated and actual response will
be minimized.

The information also permits the researcher to find


estimated evaluations for all combinations including the 54–
18=36 options never shown to the respondent. Moreover,
all respondent’s separate part-worth functions as illustrated
for person can be compared in order to see if various types
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 269
of respondents (e.g. high versus low income respondents) Notes
differ in their separate attribute evaluations. __________________
__________________
In short, while the respondent evaluates complete bundles
__________________
of attributes, the technique solves for a set of part-worths–
one for each attribute level that are computed from the __________________

overall trade offs. These part-worths can then be combined __________________


in various ways to estimate the evaluation that a respondent __________________
would give to any combination of interest. It is this high __________________
leverage between the options that are actually evaluated and __________________
those that can be evaluated (after the analysis) that makes __________________
conjoint analysis a useful tool.
__________________

Card 1 Card 2 Card 3


Brand Alpha Brand Alpha Brand Alpha

Life 3 Years Life 4 Years Life 5 Years

Price Rs 50 Price Rs 60 Price Rs 70

Colour Green Colour Green Colour Green

Respondent's Rating....5.2 Respondent's Rating....7.3 Respondent's Rating....5.7

Card 4 Card 5 Card 6


Brand Gamma Brand Gamma Brand Gamma

Life 3 Years Life 4 Years Life 5 Years

Price Rs 60 Price Rs 70 Price Rs 50

Colour Green Colour Green Colour Green

Respondent's Rating....4.8 Respondent's Rating....7.2 Respondent's Rating....9.3

Card 7 Card 8 Card 9


Brand Beta Brand Beta Brand Beta

Life 3 Years Life 4 Years Life 5 Years

Price Rs 70 Price Rs 50 Price Rs 60

Colour Green Colour Pink Colour Green

Respondent's Rating....0.8 Respondent's Rating....3.2 Respondent's Rating....6.4

Contd...
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
270 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes
Card 10 Card 11 Card 12
__________________
Brand Alpha Brand Alpha Brand Alpha
__________________
Life 3 Years Life 4 Years Life 5 Years
__________________
__________________ Price Rs 70 Price Rs 50 Price Rs 60

__________________ Colour Pink Colour Green Colour Green

__________________ Respondent's Rating....2.2 Respondent's Rating....8.1 Respondent's Rating....8.3


__________________
__________________
__________________
Card 13 Card 14 Card 15
__________________ Brand Gamma Brand Gamma Brand Gamma

Life 3 Years Life 4 Years Life 5 Years

Price Rs 50 Price Rs 60 Price Rs 70

Colour Green Colour Pink Colour Green

Respondent's Rating....6.3 Respondent's Rating....7.4 Respondent's Rating....7.3

Card 16 Card 17 Card 18


Brand Beta Brand Beta Brand Beta

Life 3 Years Life 4 Years Life 5 Years

Price Rs 60 Price Rs 70 Price Rs 50

Colour Green Colour Pink Colour Pink

Respondent's Rating....2.2 Respondent's Rating....4.3 Respondent's Rating....5.7

Figure 8.14: Product Descriptions for Conjoint Analysis

Reliability and Validity Checks


In carrying out a conjoint analysis it is useful to include the
following ancillary analysis: (1) Test-retest reliability,
(2) A comparison of actual utilities with those of random
respondents and (3) An internal validity check on model-
based utilities.

The test-retest reliability can be conducted by including a


few replicate judgements at a later stage in the interview.
The purpose here is to see if the judgements are highly
correlated, on a test-retest basis, to justify the analysis of
the respondent’s data.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 271
Notes
__________________
4 4 (3.5)
(3.3)
__________________
3 3
__________________
(2.4) (2.7) __________________
2 2
__________________
1 1
__________________
__________________
0 0
Alpha Gamma Beta 3 4 5
__________________
Brand Life __________________
4 4 __________________

3 3

2 2
(1.5) (1.2)
(1.7)
1 1

0 0
50 60 70 Pink Green

Price Colours

Figure 8.15: Part-worth Functions Obtained from Conjoint


Analysis (Hypothetical Data)

Random-respondent utilities are obtained by running a


sample of pseudo subjects. The pseudo subjects input data
consist of random ratings or rankings as the case may be. As
the sample size increases we would expect, on an average,
equal utilities within each set of factor levels. However, any
specific pseudo subject would depart from this expected value
through sampling variability alone. By analysing a sample
of 100 or so pseudosubjects the researchers can develop a
crude type of confidence interval around each set of averaged
part-worths with which to compare real respondents results.

Guidelines for use of Multivariate Analysis in


Marketing Research
There are a large number of applications of multivariate
analysis which can performed by multivariate analysis.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
272 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Jagdish N. Sheth, in "seven commandments for uses of


__________________ multivariate methods", multivariate methods for market and
__________________ survey research offers a number of guidelines to the
__________________ marketing researcher who intends to use multivariate
__________________ methods.
__________________ 1. The marketing researcher should not try to be
__________________ technique-oriented. If he develops a fancy for one or two
__________________ particular techniques and uses them indiscriminately
__________________ without checking their suitability in a given case, it will
__________________
harm the interests of marketing research. It may also
bring about the downfall of multivariate methods.
__________________
2. He should not be carried away in building models,
disregarding the fact that multivariate models are
information inputs to facilitate management in the
process of decision-making.

3. He should remember that multivariate methods are not


substitutes for his skills and imagination in the design
of research. These are necessary in conceptualizing the
problem.

4. He should not overlook the need for lucid


communication with the management. A first rate study
using sophisticated techniques may go to waste if the
management is unable to understand it.

5. He should not make statistical inferences about the


parameters of multivariate models. Such inferences are
not possible on account of the sizeable existence of non-
sampling or measurement errors in the data in the social
sciences. Such techniques should be seen as descriptive
statistical techniques for reducing large data to a
summarized and meaningful form.

6. Sometimes, the researcher may be carried away by a


random relationship among the variables and put
substantive meanings into the data, which may not be
true. This is especially true of some multivariate
methods such as cluster analysis, multidimensional
scaling and conjoint measurement as there is no
sampling theory behind them. Therefore, these are
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 8 Data Analysis-Methods 273
heuristics. To guard oneself against this error, it may Notes
be advisable to use at least two different techniques. __________________
__________________
7. Finally, he should exploit the complementary
__________________
relationship inherent in the structural and functional
multivariate methods. In other words, he should __________________

substantiate a number of judgements which he has to __________________


make with structural multivariate analysis of data. For __________________
example, it is advisable to use cluster analysis first to __________________
specify mutually exclusive groups before using a __________________
multiple discriminant analysis. __________________

Summary __________________

Multivariate analysis procedures involve simultaneous


analysis of data measured on two or more variables.
Multivariate techniques are largely empirical and deal
with the reality; they possess the ability to analyze
complex data. The basic objective underlying
multivariate techniques is to represent a collection of
massive data in a simplified way.

It transfers a mass observation into a smaller number


of composite scores in such a way that they reflect as
much information as possible contained in the raw data
obtained concerning a research study. Multivariate
analysis is specially important in behavioural sciences
and applied researches for most of such studies involve
problems in which several response variables are
observed simultaneously.

Review Questions
1. What do you mean by multivariate techniques? Explain
their significance in the context of research studies.

2. What is the significance of using multiple discriminant


analysis? Explain in brief the technical details involved
in such a technique.

3. Write short notes on:

(a) Cluster analysis


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
274 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (b) Conjoint analysis


__________________
(c) Factor analysis
__________________
__________________ (d) M.D.S
__________________
4. Name the important multivariate techniques and
__________________ explain the important characteristics of each one of such
__________________ techniques.
__________________
5. Differentiate between bivariate and cluster analysis.
__________________
__________________ 6. Explain in brief the ‘correspondence analysis’ technique.
__________________
7. Differentiate between univariate and multivariate
analysis.
275

Unit 9 Notes
__________________
__________________
Report Writing __________________
__________________
__________________
__________________
__________________
__________________

Objectives __________________
__________________
After reading this unit you will be able to:

y Discuss a market research proposal.

y Understand cost elements in a market research proposal.

y Discuss-how to prepare a research proposal.

Introduction
The formal marketing research proposal can be viewed as a
series of steps called the research process. To effectively
conduct a research process it is essential to anticipate all
the steps and recognize their interdependence. In research
proposal, the researcher should get the details of sponsor.
The research proposal could be framed up for one’s own
company as an aid to marketing decisions, for a client for
commercial purpose or by a student of marketing as an
academic exercise. In all the three cases, a researcher has to
obtain certain details, which would facilitate the researcher
to design the market research proposal. The researcher can
collect the following information from the sponsor:

1. The main objective of research and the detailed


objectives of the study and how the relevant information
will be used by the sponsor.

2. The inhouse information available to assist in


preparing the research plan. It should include the
details on marketing mix, marketing strategies and
sales data.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
276 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 3. The market competition and industry conditions in the


__________________ market.
__________________
4. The time framework for the completion of the project.
__________________
__________________ 5. Cost considerations, if any, which may have implications
__________________ on the mode of operations and the area to be covered.
__________________ 6. Any other special considerations.
__________________
After collecting the above information and also personal
__________________
discussion with sponsor, the research agency/researcher can
__________________
prepare a proposal for carrying out the research work. The
__________________ following Example will help in understanding details
required from the Hero-Honda company regarding market
survey on “psychographic analysis of Hero-Honda splendor
owners”.

Example 9.1
Details from M/s Hero-Honda Ltd. for the market survey of
Hero-Honda splendor owners:

1. Product details: __________________


(technical aspects). __________________

2. Give details on the __________________


exact product mix that is __________________
to be covered for the __________________
market survey. __________________

3. List all present dealers __________________


of the product. __________________

4. List all facilities __________________


available to the customers. __________________

5. Which are the different __________________


market segments available for __________________
the product? __________________

6. Which are the known competitive __________________


products available in the market __________________
which fulfil the same needs _________________
of the users? _________________
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 277
7. Who are the major known _________________ Notes
competitors at present in _________________ __________________
the market? _________________ __________________
__________________
8. What is main objective of _________________
this market survey? _________________ __________________
__________________
9. What coverage is desired _________________
__________________
for this market survey? _________________
__________________
10. What assistance will be provided _________________ __________________
by your company? _________________ __________________

— Past sales records _________________ __________________

— Industry demand _________________

— Sales literature _________________

— Sales display _________________

Research Proposal
The research organization will require a team depending on
the scope of the research project. An organization set-up as
given below could normally be used.

Research Organization

Research Project Incharge

Research Officers

Research Assistants

Field Staff

Figure 9.1
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
278 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes The following Table shows job description at each level could
__________________ be as under:
__________________
Table 9.1
__________________
Research Research Officer Research Assistants Field Staff
__________________ Project Incharge

__________________ He is Research officer The main job They will carry


in-charge will select the would be the out the actual
__________________ of the samples, will supervision over the field work for
research design the field staff. They data collection and
__________________ project. questionnaire will also help in will conduct
He is and will provide compilation of all personal
__________________ directly the necessary the data in a discussions with
involved with the training to systematic manner, the samples.
__________________ sponsor and assistants and properly segmented
the research team. field staff. and in tabular
__________________ He will do He is also forms for easy analysis.
the costing, involved in the
preparation analysis of the data
of time-schedule collected, arriving
and report writing at conclusions
and preparing
the draft of the report.

Costing
The following headings are to be covered in market research
proposal:

1. Man-hour costs ___________________


(Man-hours × cost + per hour) ___________________

2. Travelling costs ___________________

3. Daily allowances ___________________

4. Cost of stationery, postage, ___________________


telephone, fax, internet, etc. ___________________

5. Typing, binding ___________________

6. Admn–charges
(Normally taken as 7.5% of 1-5) ___________________

7. Profit Margin ___________________


(Normally taken as 25% of 1-6) ___________________
Total cost proposed to client ___________________

The Research Proposal


After several weeks of preparation, a research proposal is
developed and sent to the sponsor for review. At a subsequent
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 279
meeting with the research firm and after a few changes, the Notes
proposal is approved. The following is the summary of the __________________
finalized proposal submitted to M/s Electrolux by Admar on __________________
“market mapping study of washing machine”. __________________
__________________
Rationale for the Study
__________________
The rationale is to have an in-depth understanding of the __________________
overall durable market as well as that of the washing machine __________________
market. This will aid in identifying opportunity segments
__________________
and the need gaps present in the market.
__________________
Based on this, they will be able to develop appropriate __________________
positioning strategy for Electrolux. Moreover, it will help in
providing specific inputs for development of an effective
communication strategy and plan.

Research Objectives
(a) To understand consumer psychographics with respect
to ownership and purchase of durables in general and
washing machines in specific.

(b) To understand existing habits and beliefs that influence


the needs and expectations with respect to ownership
and purchase of washing machines, its type and brands
in specific.

(c) To map positions of different brands of washing


machines in the consumer’s mind by perceived strengths
and weaknesses of each of its contributors.

(d) To gauge the existing equity of Electrolux and its


variance across different levels of familiarity.

Research Methodology
u The research design will be an exploratory-cum-
diagnostic qualitative research study.

u It will involve personal, face-to-face and in-depth


interviewing technique with the help of an interview
guide.

u The research is to be conducted by the field staff.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
280 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Target Respondents


__________________
To give the premium pricing strategy of Electrolux range of
__________________
washing machines the following could be our target groups:
__________________
__________________ u Monthly household income of Rs 10,000 plus or owning
__________________ a new generation car and air-conditioner at home.
__________________ u Potential buyers on active search.
__________________
u Recent buyers (within 3 months) of washing machines.
__________________
__________________ u Households having a high level of education as well as
__________________ high income.

u Either replacement or first time buyers of washing


machines.

Research Centres
The centres, we have chosen by contrasting weather and
buying habits are two of the largest cosmopolitan cities and
the most potential markets in India.

Mumbai — Humid and high monsoon market. A city of


middle-class earners.

Delhi — Dry and low monsoon market. A city


constituting mostly of business-class earners.

Sample Size and Sample Composition


Delhi Mumbai Total

Recent Buyers 15 15 30
Potential Buyers 15 15 30
Total 30 30 60

u A quota of about 10 to 15 buyers of front loading washing


machines.

u 60:40 split between replacement and first time buyer.

u 50:50 split of executives/professionals and businessmen.

Sampling Procedures
u The procedure adopted for this project is purposive
sample method.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 281
u The listing of respondents from both households and Notes
dealers outlets is spread across 6-8 areas within each __________________
city. __________________

u Appointments will be fixed during listing and __________________

respondents interviewed at home. __________________


__________________
Research Project Execution __________________
u The entire recruitment is carried out and managed by __________________
the general field force of Admar. __________________
__________________
u Research executives carry out all interviews.
__________________
u Usual norms of quality are followed both at the
recruitment as well as before the interviewing stage to
ensure data integrity.

u Research executives analyse all response contents.

Research Units
u Most of the houses surveyed were nuclear families living
in flats with 4-6 members.

u Out of them 2-3 were families with school going children.

u The living room wears a functional look with a sofa or


sofa-cum-bed, a centre table, wall paintings/hangings,
bronze plates, wall paper decorations.

u The flats usually have attached dining and drawing


rooms.

u The television and music systems are normally kept in


the living room (in Mumbai) and in the bedroom (in
Delhi). Normally there is a trend of multiple television
ownership in Delhi.

u Washing machines are normally kept in the toilet or


dining area.

Samples
u Life revolves around the house and family.

u In Delhi, mainly full-time housewives, whereas in


Mumbai, it’s the reverse.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
282 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes u Cooking–she does it herself.


__________________
u Bartan dhona, jharru-pochha and dusting-the maid/
__________________
servant does it.
__________________
__________________ u Kapra dhona–she does it herself.
__________________ u Bartan dhona, jharru-pochha and dusting–the maid/
__________________ servant does it.
__________________
u Kapra dhona–she does it herself in Delhi but not so in
__________________
Mumbai.
__________________
__________________ Research at a Glance
u Typically the middle class and a few upper middle class
families have a functional outlook.

u The housewife almost always holds the rein of the family


with the husband as the financier.

u Most of the families apparently seem quite content with


what they own.

u Household entertainment and kitchen gadgets form the


basic possessions everywhere.

u There is wide variation evident across cities surveyed


for items like car, cooler, air-conditioner, oven, vaccum
cleaner.

u Washing machine—emerging as a necessity.

Example 9.2
Research Project Synopsis
Consumer Behaviour of Titan Quartz Watch

1. Introduction

A business firm comes into being because of the existence of


consumers who have unfulfilled needs and wants. To fulfil
consumers’ needs, organizations are created. The consumer
behaviour can be defined as, “the decision process and
physical activity engaged in when evaluating, acquiring,
using or disposing of goods and services”. The term consumer
is a far wider term as it comprises of not only the actual
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 283
buyer or consumer but also users, i.e., customers. There are Notes

two situations when this distinction between consumers and __________________


customers may occur: __________________
__________________
(a) When the service or product is provided free.
__________________
(b) When the customer is not the actual user of the product __________________
or is only one of the many users. __________________

2. Objectives of the Study __________________


__________________
The objectives of the study are: __________________

(a) To find out the consumers of Titan watches. __________________

(b) To find out why do these consumers buy this brand and
not the competitors’ brands.

(c) To analyse as to how watch consumers perceive Titan


products to fulfil their needs.

(d) To evaluate whether they feel satisfied after having


bought Titan product.

(e) To study the feelings of the consumers as reflected in


their behaviour as consumers.

3. Research Methodology used for Carrying out the


Study

(a) Data collection: The study conducted is exploratory in


nature. The data will be collected for 100 people–both
males and females.

(b) Data analysis: The technique of sampling used is


stratified random sampling. Here the sample is based
on three groups. First group is upto post graduates,
second group is employees and third group is above 45
years of age.

(c) Statistical analysis: Consumer behaviour of users


of Titan’s Quartz watch is analyzed by getting
information from friends, newspapers, magazines,
television and hoardings about the price and quality of
the watch.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
284 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 4. Application of Consumer Behaviour in Marketing


__________________
Marketing is defined as ‘human activity directed at satisfying
__________________
needs and wants through the exchange process’. Consumer
__________________
behaviour and marketing go hand in hand. Consumer
__________________ behaviour has a number of applications in the area of
__________________ marketing as given below:
__________________
(a) Analysis of market opportunity.
__________________
__________________ (b) Selection of the target market.
__________________
(c) Determination of the product mix.
__________________
(i) Product

(ii) Price

(iii) Distribution

(iv) Promotion

(d) Non-profit and social marketing.

(i) Analysis of market opportunity: The study of


consumer behaviour helps in identifying needs and
wants which are not fulfilled. This is done by
examining income status, consumers’ lifestyle and
emerging influences. Its rapidly rising sales graph
is an indication of how well the product has
satisfied the consumers’ needs.

(ii) Selecting the target market: The study of


consumers’ trend would reveal distinct groups of
consumers with very distinct needs and wants.
Study of potential consumers for wrist watches
reveal that there is a class of consumers who
prefer Quartz watches on special occasions and
use normal watches in their daily routine.

(iii) Determining the product mix: Having identified


the unfulfilled needs slot and having modified the
product to suit heterogenous consumers tests, the
markets now have to get down to the brass tacks
of marketing. They have to determine the right
mix of product, price, promotion and advertising.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 285
(a) Product: The marketer has the product that Notes
satisfies hitherto unfulfilled consumer __________________
needs, but he must decide the size, shape __________________
and attributes of the products. Titan __________________
must figure out conspicuously to have __________________
predominant quality over other products to
__________________
choose from. Does the product require any
__________________
special kind of packaging? Does it need any
__________________
guarantee or after sales services? The study
of consumer behaviour also guides the __________________

marketer in making decisions regarding __________________


packaging. __________________

(b) Price: Titan company has already fixed the


price the marketer should charge for the
product. Should it be the same as that of
the competing product or lower or
higher? Should the price be market on the
product or left to the discretion of the
retailer?

(c) Distribution: Having determined the


product size, packaging, shape and price, the
next decision the marketer has to make is
the distribution channel. What type of retail
outlet should sell the products? Does the
consumer look for the nearest convenient
location or is he willing to travel some
distance for buying the product?

(d) Promotion: The marketer here is concerned


with finding the most effective methods of
promotion which make the product stand
out amongst the clutter of so many other
brands and products which help increase
the sales objective and yet be within the
budget.

(iv) Use in non-profit and social marketing: The


knowledge of consumer behaviour is also useful
in the marketing of non-profit or social or
governmental service or institution.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
286 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes 5. Contribution from the Study


__________________
Consumer behaviour comprises the entire spectrum of
__________________
activities and processes in which individuals engage when
__________________
buying, using, acquiring or disposing of goods and services.
__________________ The purchase is only one part of this range of activities. An
__________________ understanding of consumer behaviour is of paramount
__________________ importance to all persons engaged in a marketing activity.
__________________ Its importance can be measured from the fact that:
__________________
(a) Behaviour of consumers can be understood.
__________________
__________________
(b) Behaviour can be influenced.

(c) The marketer can manipulate the influencing variables


to his advantage. Having understood the behaviour of
consumers and knowing that their behaviour can be
influenced, the company can initiate a number of steps
to do so.

Research Report Writing


The last step in the research process is the preparation and
presentation of the research report. The best of research
efforts will be of little value unless the result can be
summarized and communicated to marketing management
in a form that is both understandable and useful.
Preparation and presentation of the research report is
the most important part of the research process. If the
report is confusing or poorly written, all the time and effort
spent on gathering and analysing data would not bring any
result.

The purpose of this chapter is to give guidance in preparing


a research report, which is simply the presentation of
research findings to a specific audience. We will discuss
two vital ingredients of this communication process: (1) The
written research report, and (2) The oral presentation
typically provided for those who occupy executive
positions. Along the way, we will review some of the
more popular devices for graphical and visual support
and how statistics can be used or misused to suit your
needs.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 287
Types of Research Reports Notes
__________________
Depending on its intended audience, the research report may
__________________
by either technical or popular in orientation. While both
__________________
approaches describe the research study, its methodology,
findings, conclusions and recommendations, they can differ __________________

considerably in terms of detail, writing style, use of technical __________________


terms and length. In general, the higher the executive status __________________
of the audience, the shorter the report will tend to be. __________________
__________________
Technical Report
__________________
The technical report is generally intended for other __________________
researcher, or for research managers. The report should
enable another researcher to be a critic of methodology, check
calculations and accuracy and to follow everything which is
done on a step-by-step basis. A brief definition of any
technical term should be given.

The Popular Report


The popular report is intended for a more general audience,
one that is not conversant with the details of research
methods and terminology. Compared to the technical report,
the presentation will be a bit more lively with increased
attention to headlines, flow diagrams, charts, tables and
occasional summaries for the purpose of stressing major
points.

Because different kinds of audience may be interested in


the results of the same marketing research study, it is
sometimes necessary to write both a technical report and a
popular report.

Guidelines for Writing a Report


Researchers who are effective in report writing agree that
there are a series of guidelines which should be followed.

Consider the Audience: Make the report clear; use only


words familiar to the readers and define all technical terms.
To make the comparison of figures easier, use percentages,
rounded off figures, ranks or ratios; put the exact data in a
table within the text or in the appendix. Use graphic aids
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
288 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (charts, graphs, pictures, etc.) wherever they help clarify the
__________________ presentation of data.
__________________
Address the Information Needs: Remember the research
__________________
report is designed to communicate information to decision
__________________ makers. Make sure it clearly relates the research findings
__________________ to the objectives of management.
__________________
Be Concise, Yet Complete: Most managers will not
__________________
want to read about the details of a research report.
__________________
Knowing what to include and what to leave out is a
__________________ confusion.
__________________
Steps in Report Writing Process
Two main attitudes must be cultivated while preparing to
write a project report:

Attitude 1: Adopt the fresh mind approach (FMA).

Attitude 2: Adopt the kiss approach (keep it short and


simple).

Fresh-Mind Approach (FMA)


Normally we proceed with a preoccupied mind as if the
solution is available and the report is to be made suiting
that solution. It is essential that a researcher must clear
his mind of all preoccupations about the solution to the
problem. If the data collection, investigation and analysis
are done properly, the interpretation and recommendations
may be quite different from what the researcher expected.
Developing the fresh mind approach will help in avoiding
the mindset about a situation, which is too common
when dealing with a familiar work situation. We are
always able to find fault with someone else’s work, while
we fail to see what is wrong with our own. In light of this,
the person who is going to assess your work will adopt the
above strategy of evaluating your work if sufficient
originality, genuineness and a fresh approach has not be
adopted by you. The project assessor certainly possesses the
skills of assessing whether the work is done with FMA or
otherwise.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 289
Kiss Approach (Keep it Short and Simple) Notes
__________________
Your project work is going to be read and evaluated by
__________________
someone who:
__________________
u Maybe outside your organization. __________________

u May not be quite familiar with the field of study. __________________


__________________
u May not be conversant with technical terms.
__________________
Therefore, you must write the project study such that: __________________
__________________
u It is understandable.
__________________
u It is simple.

u It provides sufficient background information.

u It creates a clear picture in the mind of your work


environment.

u It helps him understand the problem under study.

For the above you must:

1. Use short sentences.

2. Use simple language.

3. Use logic.

Short sentences will help in avoiding grammatical mistakes,


thus creating a good impression of the researcher. Avoid use
of technical jargon. This may cause harassment to the
assessor. If the use of technical terms in unavoidable, it is
desired to include glossary of terms in the project report.
Sometimes the researcher uses abbreviations. It is
recommended that a page explaining the full form of
abbreviations may also be included. In a nutshell, short
sentences and simple language should be used throughout
the report.

Tips to be followed.

u Use a friendly and informal tone.

u Explain things clearly.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
290 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes u Avoid overloading with information.


__________________
u Use relevant anecdotes.
__________________
__________________ u Keep sentences, and paragraphs, short.
__________________
u Choose simple and familiar words.
__________________
__________________ u Avoid long phrases. Say although’ rather than "inspite
of the fact that”.
__________________
__________________ u Explain jargon and technical terms clearly.
__________________
Size Report
__________________
Size of report will depend upon the subject matter and what
is involved in the study.

All India Management Association, Centre for Management


Education, suggests that the report should be covered
between 60 to 100 typed pages (double space). Each
page must be typed only on one side leaving a wide
margin.

School of Management Studies, Indira Gandhi National Open


University (IGNOU) suggests that the length of the report
may be about 50 to 60 double spaced typed pages
not exceeding 18000 words (excluding appendices and
exhibits). However, 10 per cent variation on either side is
admissible.

Faculty of Management Studies, University of Rajasthan


suggests an average size of the report is between 100 -150
typed pages in 3/2 space.

Bearing in mind the time available to students to complete


the report, and to facilitate the understandable anxiety of
students who are doing this exercise for the first time, the
size of the report is fairly good.

In practice, there is no mandatory length and the students


should be guided by their project guides on this matter,
in fact, the researcher should deal with the report without
regard to its length ensuring that all the information is
recorded and the report produced conforms to the
requirements of the objectives of the study.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 291
Notes
Check list
__________________
1. Check whether all essential information is included.
__________________
2. Check for proper structure. __________________

3. Check for clarity. __________________


__________________
4. Check for completeness.
__________________
5. Check there is no unnecessary pending.
__________________
6. Check that it essentially contributes to the objectives. __________________

7. Check if anything can be curtailed if it is too voluminous. __________________


__________________
In other words, write everything about the subject under
study and structure it with the help or your guide. It is
desirable to get the first draft typed in double space with a
two inch margin on the right hand side of the page. This will
give the opportunity to see:

u What the report looks like?

u What modifications are necessitated?

The margin and the spacing helps in entering corrections


and additional information if desired. This also facilitates
editing of the report.

Steps in Writing Report


The method of carrying out the study, assembling of data
and compiling of a report should be carried on in the following
eight stages:

1. Preparation

2. Gathering the information

3. Sorting, selecting, arranging and recording the material

4. Writing the introduction

5. Recording inferences

6. Writing recommendations

7. Preparing the report presentation

8. Typing the report.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
292 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Preparation
__________________
Ready yourself mentally for the task of project writing as
__________________
explained in Fresh Mental Approach (FMA).
__________________
__________________
Be sure you accurately understand the purpose of your
report and its objectives as defined in the project synopsis
__________________
or terms of reference.
__________________
__________________
Put yourself in the position of the evaluator and ask yourself
these questions:
__________________
__________________ 1. What information, i.e., evidence and justification, would
__________________
I require to arrive at a logical inference and make the
right decision?

2. As an evaluator, how much, if anything, do I already


know about this subject? Decide on the following issues:

u Method of study.

u What information do you need?

u Who is going to be involved?

u With whom you going to carry out your study?

u When are you going to carry out your study?

u Project plan–time schedule to complete the task.

It is extremely useful to relate the plan to a time schedule


so that you know what to do and when, in order to complete
the report by the required deadline. (Prepare a CPM by
breaking down the project activities into sub-tasks and
allocate a duration to each sub-task).

Gathering the Material


At this stage, don’t attempt to write the report. Collect all
the information i.e., facts, figures and ideas, about the subject
by whatever means available. Some of the means used in
gathering the information are briefly given as under:

By Observation

Don’t rely on your memory. While observing an operation or


coincidence or a situation, record the sequence in detail. Look
at it, watch it, time it and record everything you see.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 293
By Reading Notes
__________________
Published work of other researchers on similar issues, factual
__________________
documentation produced within the organization (if the study
__________________
is about an organization), documentation of professional
bodies will often provide enough information relative to your __________________

subject. Give credit in your report to the source of such __________________


information. __________________
__________________
By Interviews
__________________
Interviews can be formal or informal depending upon their __________________
nature and the person being interviewed. Keep the following __________________
points in mind:

u Plan your interview

u Give your interviewee notice of your subject

u Prepare your questions beforehand

u Don’t rush an interview

u Focus on substance

u Be accurate.

By Discussion

Hold discussions with your colleagues, guide, and any


interested person in the subject. Respect their views.

By Surveys

Surveys are carried out to assess customer opinion and


reaction. These are also useful in understanding people’s
attitudes. Questionnaires are useful instruments in carrying
out surveys. Care must be taken in preparing the
questionnaire and, if in doubt, seek the advice of your guide.
Include copies of the questionnaires and of their analysis in
your report.

By Enquiries

Enquiries can be made from manufacturing or service


industry through letters, fax or telephone.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
294 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Explain
__________________
u Who you are?
__________________
__________________ u Purpose of your enquiry.
__________________
Be polite while asking for their help. Generally you will
__________________ receive a favourable response and these sources of
__________________ information can be most valuable. Give credit to such
__________________ informants and always thank them for their help.
__________________ Acknowledge their support in your report. Use your
__________________ knowledge and experience
__________________ By experiment

Try something and see if it works. Record the details and


verify the results. If there is any doubt about the accuracy of
the information, try and check and verify it from another
source.

If it is not possible to verify the information, discard it or


quote it as uncorroborated. Always be sure of the facts.

Make liberal use of visuals, i.e., diagrams, maps, charts,


photographs, statistical tables. Even sometimes it may be
appropriate to attach sample products to illustrate and
explain your text. This will make your report visually
attractive and interesting.

Avoid using memory and prejudice. These can often lead to


distorted facts. As a report writer always carry a notebook
with you and note down pieces of information or ideas or
things to do at the time they come to you, instead of relying
upon the memory. Once you have written it down you can
concentrate on something else.

At this stage you will have sufficient information in the form


of notes, collections from various published data, survey data
and so on. The next step is to sort it all out.

Sorting, Selecting, Arranging and Recording Material


Having assembled all your material, check to ensure that
you have all you need to support your arguments. Arrange
it in logical sequence, start to finish, ensuring a natural flow.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 295
The questionnaires, surveys and any other similar document Notes
should be explained properly. To ensure clarity to the __________________
readers adequate notations should be used for visuals, i.e., __________________
diagrams, maps, charts, photographs, statistical tables. __________________

If illustrations are used to explain a particular part of the __________________

text and are not going to be referred to several times, insert __________________
them in the appropriate position in the text. __________________
__________________
If the evaluator needs to refer to these illustrations several
__________________
times or if they form a supportive material it should be put
in the appendices section at the end of the report. __________________
__________________
At this stage, clear your mind and go through what you have
written. Put following questions to yourself:

1. Is it clear?

2. Does it make any sense?

3. Does it clearly specify or indicate problem areas?

4. Does it lead you to the “conclusion” section logically?

Writing the Introduction


While writing the introduction take following points into
consideration:

1. Create a background of the organization and work


situation for which the project work is being done.

2. Describe the purpose of study.

3. Describe the method of study.

4. Explain difficulties encountered in the preparation of


report.

5. Introduce yourself, your institution/organization briefly.

6. Ensure continuity between all sections.

Recording Inferences
The conclusions or findings should flow from the outcome of
the investigations. These findings should lead to concrete
recommendations. While writing conclusions, the objectives
of the study should be kept in mind. This will facilitate in
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
296 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes understanding the quality of investigations and gauge the


__________________ success of work. Summary of the problem areas followed by
__________________ explanation of the various alternative solutions to the
__________________ problem should be undertaken and the reason for accepting
__________________ or rejecting the possible alternative should be explained
logically.
__________________
__________________ Writing Recommendations
__________________
While writing recommendations you must adopt a positive
__________________
attitude and enlist those points first, which you really want
__________________
to be done. You must adopt a constructive approach. This
__________________
section may be started as follows.

It is recommended that (provide a list of recommendations


or action you feel should be taken):

1.

2.

3.

4.

5.

6.

7.

If the recommendations involve financial figures of


statements, these should be shown against each item. Don’t
repeat what you have written. In conclusion, you may give a
cross reference of a relevant paragraph if you want to include
something in this section.

Give your Report a Positive Ending


At this stage the text of the report is complete and you are
ready to go the next stage i.e. presentation of the report.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 297
Report Presentation Notes
__________________
Stitch together all the different threads of information into
__________________
a coherent whole. Present the data in the form as suggested
__________________
in the chapter plan. Write the first draft. Discuss with your
guide and incorporate the suggested modifications. Revise __________________

the document and type it finally. The graphics play a __________________


significant role in report image and care must be taken about __________________
the visual aspects of the report. __________________
__________________
Visual Aspects of the Report
__________________
Many reports require communication of voluminous and __________________
complex information. It is not possible to do this job with
words only. Graphics portray things that would take volumes
to explain through words. Technical reports, in particular,
cannot be completed without use of graphics. The main role
of graphics in a report is:

u To support text to communicate report content. To


emphasize key points.

u To make the report attractive.

u To add interest.

u To save time.

u To increase credibility.

u To add variety.

u To create impact.

u To help the reader to remember main points.

Before organizing the report, you must think and plan:

u What graphics are to be used?

u How many graphics are to be used?

u Where to use graphics in the text?

The graphics used in reports are normally:

u Charts
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
298 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes u Graphs
__________________
u Photographs
__________________
__________________ u Maps
__________________
How many graphics are to be used
__________________
__________________ Each graphic should be selected such that it serves to
__________________ communicate a specific point of the report. Random use of
__________________
graphics should be avoided. There is no set rule to determine
the number of graphics. However, plan what is required to
__________________
be communicated effectively through graphics and include
__________________
these graphics only.

Where to use graphics in the text


The graphics should be placed in such a position that they
are likely to be there where they are required to be seen.

Consider following points while placing graphics:

1. Place them close to the related text.

2. If the graphic is large, and is of full size, place it on the


page following the text.

3. If the graphic is small, place it appropriately and


surround it with the related text.

4. Graphics which are not discussed in the main report


but support the discussion should be appended.

5. Always make reference of the graphic in the text to invite


to reader to refer it.

6. Always use rules and borders to enhance presentability.

7. Use colour if you have access to colour printers.

8. Number graphics and title appropriately.

Types of Visual Aids/Graphics


The common type of visual aids used in communicating data
are:

1. Bar charts
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 299
2. Line or area charts Notes
__________________
3. Pictogram
__________________
4. Pie charts __________________
__________________
5. Pictorial diagrams
__________________
General Points about Charts
__________________
Your illustrations should include
__________________
u Appropriate title and units __________________
u Scales or labels on axes __________________

u Key __________________

u Source of data (if from outside source)

u Caption to the chart

Bar charts
Bar charts are:

u Easy to draw

u Familiar to all

u Good for making comparison

While using bar charts ensure:

u The order of categories should be logical and match any


tallies of the same data.

u The width of all the bars should be kept same.

u Start the vertical axis at zero (0), if possible.

u If it is unavoidable to start at other than zero (0), indicate


this clearly by a break in each bar or in the vertical axis,
with a zig-zag or wavy stroke across each end of the
break.

Stacked Bar Chart


These charts show the total well. It is difficult to compare
heights of the sections with stack bar charts. The stacked
bar charts should not be used if the total height has no
meaning.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
300 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Line or Area Chart


__________________
Line and area bar charts are used when there is logical order
__________________
of points. The order of North (N), South (S), West (W), and
__________________
East (E) could be changed giving a completely different
__________________ pattern to the graph so it is not a good chart for this data.
__________________
__________________ Pictograms
__________________ A picture or symbol standing for a word or group of words A
__________________ chart on which symbols are used to represent values.
__________________
Pie Charts
__________________
Pie charts are the commonly used visual aids. The
disadvantage of pie chart is that it is not possible to show
any category that has zero (0) frequency.

Use a pie chart when you want to compare a number of


elements to each other and in relation to a whole. If you
compare the same information expressed as regular
rectangular blocks to a pie graph, you see that the bars are
more precise in comparing elements to each other, but that
the pie graph is much more effective comparing elements to
the whole. One of the fundamental rule of using pie chart is
“never include more than 6 elements.” If you have more than
this number only show six main elements and include the
others under a diverse heading. If necessary break down the
diverse category into a second pie graph.

The pie graph can be made more attractive by:

u Adding pattern to each part

u Using colours

u Adding depth

Two Variables Diagrams


These diagrams are very important when illustration of more
than one variable is needed. It is of great significance to
determine, which variable should be put on which axis.

They illustrate the total number of each sex respondents


questioned, but this may not be what you want to illustrate.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 301
They also illustrate the total number people of in each Notes
category clearly. __________________
__________________
It is more logical to have ‘sexes’ on the horizontal axis for
__________________
a percentage stacked bar chart where the response of
sexes are being compared. Illustrates a percentage __________________

stacked bar chart where responses of the sexes are __________________


compared. __________________
__________________
For showing the comparison, multiple bar charts are
__________________
constructed. Always provide a gap between the bars for
clarity. __________________
__________________
Area Charts
Area charts illustrate how it stacks the data like a stacked
bar chart and so indicates total record data.

Typing the Report


For report typing, A4 size of paper is recommended. It is of
a standard size and will present no difficulty with covers,
binding, photocopying and dividing cards between the
sections. The margins should be kept as suggested earlier in
this chapter.

Make sure:

1. Each section of the report starts on a new page.

2. Indicate where capitals, italics and underlining are


required.

3. If using DTP indicate bold, capital, front type and font


size required.

4. All headings and sections and subsections are clearly


indicated.

5. Indicate where any indentations of the text are required


for the ease of reading.

It is recommended that a clear system of identification is


issued for sections, subsections and paragraphs for the speed
of reference.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
302 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes A typical example is as follows:


__________________
Section Headings 1 2 3
__________________
Sub-sections 1.1 2.1 3.1
__________________
1.2 2.2 3.2
__________________
__________________ 1.3 2.3 3.3
__________________ Paragraphs 1.1.1 2.1.1 3.1.1
__________________ 1.1.2 2.1.2 3.1.2
__________________ 1.1.3 2.1.3 3.1.3
__________________
1.2.1 2.2.1 3.2.1
__________________
1.2.2 2.2.2 3.2.2
Sub-paragraphs 1.1.1(a) 2.1.1(b) 3.1.1(a)
(b) (b) (b)
(c) (c) (c)

Important Note
The sample reports given in the book, the main reports with
appended parts. These are only typical reports to support
the report writers in understanding the principles of report
writing. Each report will be unique depending upon the topic
of research.

To submit the report, the prefatory parts should be added


in the main report. The typical sample of a prefatory part is
given in the book at appropriate places.
Check List for project report before final binding (Typical)

Ensure the prefatory parts are enclosed in the following sequence after
the tittle page:

1. Approved project proposal proforma

2. Synopsis

3. The certificate of originality of work

4. Letter of authorization

5. Letter of transmittal

6. Acknowledgements

Contd...
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 303
Ensure the main report contains: Notes
__________________
1. Table of contents
__________________
2. Table of illustrations, glossary of terms, abbreviations, executive
__________________
summary, introduction, main report with chapter plan
recommendations and limitations. __________________
__________________
Ensure appended parts contains:
__________________
1. Bibliography
__________________
2. Appendix
__________________
3. Any other source of information. __________________
__________________
Final Remarks
As soon as your project has been typed, read it carefully and
correct any errors. Consider following questions about your
report.
1. Does it look a presentable document?

2. Does it make sense?

3. Does it meet objectives of study?

4. Is it clear and understandable?

5. Is it factual, logical, conclusive?

6. Will the evaluator/reader be stimulated to read it?

7. Are you satisfied with it?

8. Is your guide satisfied with it?

The researcher should get the report signed by the guide.


Determine the number of copies required before going for
photocopying. It is preferable, but not necessary, to reproduce
photographs and other illustrations in original. The original
copy which goes to the assessor and the copy which goes to
the library must be presentable. The library copies in all
probability may be referred by your juniors.

A Bit of Advice
Your guide is the person who is supposed to be with you
throughout the preparation of the report. Don’t hesitate to
bring your problems to his notice. Seek his advice. Be guided
by him. Respect him. Involve him.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
304 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes Documentation and Bibliography


__________________
While writing your report you are likely to use the
__________________
information which is not your own information. The
__________________
information which is developed through secondary sources
__________________
must be foot-noted. It should also be listed in the
__________________ bibliography. It is always advisable to acknowledge the
__________________ referred material when it is not a general knowledge. For
__________________ acknowledging the sources a standard (e.g., BIS, Harward,
__________________ APA, Chicago, MLA) reference system should be selected. It
__________________ should be followed throughout the report consistently. Some
__________________
typical examples of recording informations are discussed
below:

Guidelines for Getting Information


Determination is a major word to keep in mind when looking
for facts. If you don’t find what you need in one reference,
look for other sources. Never give up and stop with one or
two attempts.
Guidelines for Getting Information

u Scan material to find out if it is relevant to your project study.

u Check the credibility of the material by considering its text:

Source

Publisher;

Author; and

Current data

u Everything in print is not necessarily true. Check for genuineness.

u Look for several viewpoints on a subject. Don’t depend on one.

Guidelines for Recording Information


Although you may find considerable information on
a subject, the real test of your research ability comes
in getting the important facts down in a form that
is easily understandable. Volumes of inconsistent
information do not yield adequate information for preparing
reports.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 305
Follow these guidelines when taking notes: Notes

Guidelines for Recording Information __________________


u Use separate sheets of paper or help cards for each source. __________________
u Make a tentative bibliography by writing down the titles which __________________
appear to offer the information you require.
__________________
u Keep the papers or help cards in alphabetical order.
__________________
u Write in ink.
u Use a consistent bibliographical form. __________________
u Prepare an outline that will give direction to your investigation. __________________
Note: Usually, bibliography material is recorded on 3 × 5 inch help __________________
cards, and notes are made on 4 × 6 inch help cards. Cards are made
from thick sheets of paper. __________________

u Place a heading, keyed to your outline, on each help card. __________________


u Use direct quotations only when they are significant. Enclose quoted
material in quotation marks, taking care to copy each word exactly
as written.
u Use footnotes to give credit for another person’s idea which you are
using.
u Arrange your notes in the order of the outline to facilitate their use
for the basis of the report.

Preparing Footnotes
Footnotes are single spaced and indented five spaces with
footnote typed slightly above the line. The names of authors
are given in normal order. Commas are used to separate main
items. Each footnote ends with a period.

These examples will guide the author to write footnotes:

Books
(One author)
1. Robert Cameron, Statistics for Business (New York:
Oxford Press, 1971), p. 35.

(Two authors)
2. M.K. Rampal and S.L. Gupta, Cases and Simulation in
Marketing (New Delhi: Galgotia Publishing Company,
1999), p. 101.

(Three authors)
3. Ruth Goldhammer, Arthur McGille, and Clyda Bailey,
Marketing Ideas for Today (Columbus: University
Printers, 1973), p. 44.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
306 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (More than three authors)


__________________ 4. Carla samples et al. Survey Methods for Business (New
__________________ York: Capital Publishers, 1970), p. 58.
__________________
(An editor)
__________________ 5. M.K. Rampal, Cases in Service Marketing (New Delhi:
__________________ Galgotia Publishing Company, 2000), p. 84.
__________________
(Corporate authorship)
__________________
6. S.R. Lubeck Committee, Administrative Techniques for
__________________
Corporations (San Francisco: American Administrative
__________________ Association, 1970), p. 223.
__________________
Magazines and Newspapers
(Article in a weekly magazine with no author given)

1. “Trends in Office Equipment,” Office Equipment,


January 1977, pp. 88-90.

(Article in a weekly magazine with author given)

2. Rajni Rampal, “Motivating Executives”, “Marketing


News, 20, No.5 (1999),52.

(A news story with author given)

3. Ashok Galgotla, “Current Legislation Affecting


Business,” Delhi Monitor, 16 Feb, 1998, p. 8, col. 4.

(A news story with no author given)

4. “Coping with Income Taxes,” The Daily News, 4 July


1995, p. 3, col. 1.

Encyclopedias
(An article in reference work with author given)

1. “Facts About the Bureau of the Census,” Trinity


Encyclopedia, 1970, VII, 493.

(An article in a reference work with no author given)

2. Ben Miller, “Secretarial Skills for the Seventies",


Modern Encyclopedia, 1972, X, 477.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 307
Bulletins and Pamphlets Notes
__________________
Guidelines for managers (Washington, D.C.: Associated
__________________
Managers of the United States, 1977), p. 17.
__________________
Highlights of the Annual Meeting of the Personnel __________________
Management Association, Personnel Management __________________
Association Publication No. A-22 (Washington, D.C.:
__________________
Washington University Press, 1970), p. 35.
__________________
Footnote Pointers __________________
__________________
Acknowledgments
__________________
u Footnotes are used to acknowledge and document
material taken from sources. Failure to give credit for
borrowed material is plagiarism.

Placement

u Footnotes may appear at the bottom of the page on which


the quoted material is used or at the end of each chapter
or at the end of an entire work.

Numbering

u Footnotes may be numbered consecutively throughout


the entire work or numbered consecutively by chapter.

Printing

u Footnotes appear in print as single spaced with double


spacing separating one footnote from another. However,
in manuscript copy, footnotes are double spaced with
triple spacing separating one footnote from another.

Referencing the Same Source

u Use ibid., (in the same place), to indicate that the source
is exactly the same one as in the immediately preceding
footnote.

Example 9.3
(First reference)

1. Rampal M.K., Industrial Marketing (New Delhi: Galgotia


Publishing Company, 2001), p. 80.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
308 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (Repeated reference)


__________________
2. Ibid., p. 70.
__________________
__________________ Note: If the page number is not different, simply use Ibid.
__________________
Preparing Bibliography
__________________
__________________ Bibliography is a list of publications or works which the
__________________ researcher has used in the preparation of the report. In other
__________________
words, this should be a list of acknowledgements of sources
of information other than personal contacts.
__________________
__________________ The need for referencing is two fold:

1. It enables the researcher and the evaluator to identify


the source of the information or argument, they may
consult that source directly when the need arises.

2. It helps you to avoid plagiarism, the effects of which


may be the payment of a severe penalty.

Referencing Guidelines
Reference is required whenever you:

u Quote anything. If the quotation is taken from a book


or article by a third person, then you need to make
this clear in your reference; thus you would cite
the original source, indicating that it has been
quoted by that third person–identifying also that
source.

u Refer explicitly to the work of a particular writer, even


where no direct quotation is involved.

u Say something which is the most important element in


your argument, for which a reference can be given.

No reference is required to support:

u Factual Information or general statements which are


clearly common knowledge.

u An argument that derives from large number of


different sources.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 309
The Reference Form Notes
__________________
Once again there are no common rules that all researchers
__________________
follow, but there are two key principles which every effective
__________________
form of citation follows:
__________________
1. Whatever system of citation you employ in a particular __________________
piece of work, it is crucial that you are consistent. Thus,
__________________
once you have decided on the set of conventions you are
__________________
going to follow, make sure that you follow it rigorously
__________________
and consistently throughout the piece of writing.
__________________
2. References should always be full enough to enable the __________________
reader to follow them up with a minimum of difficulty.
Thus, it is essential that every reference identifies the
particular page number(s). A reference that leaves the
reader to find his or her own way to the appropriate
part of the book is of very little value, unless of course
it is in fact the whole book to which you want to refer.

Books

(One author)

Singh Harpal, Indian Company Law’, New Delhi: Galgotia


Publishing Company, 1999.

(Two authors)

Bailey, Clyda, and John O’Donnell, The Economic World,


American Economists Association, Chicago: Cronkite
Publishing House, 1979.

(Three authors)

Gupta Nirmal, Jain R.K. and Malik R.K., Elements of


Accounting. New Delhi: Galgotia Publishing Company, 1998.

(More than three authors)

Griffith, Donna Jean, et al, Problems in Editing. Boston:


Hartcourt and Company, 1976.

(An editor)

Cammero, T.J., ed. Corporations and Organizational


Psychology. New York: Wilson Publishers, 1979.
OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
310 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes (Corporate authorship)


__________________
Committee on Business Research. Research Design for
__________________
Business. Pittsbugh: University Press, 1974.
__________________
__________________ Magazines and Newspapers
__________________
(Article in a weekly magazine with no author given)
__________________
__________________ “Today’s Stocks,” Business World, 2 April 1977, pp, 21-23.
__________________ (Article in a weekly magazine with author given)
__________________
Kristen, Dorothy. “Women in Personnel Management.”
__________________
American Management Association, 12 March 1978, pp. 18-
20.

(A news story with author given)

Clee, Peter, “Financial Trends.” New York Daily, 26 April


1973, p. 4, cols. 3-4.

(A news story with no author given)

“The Rampage of Inflation.” The Cleveland Monitor, 16


December 1972, p.5. cols. 1-2

Encyclopedias
(An article in 8 reference work with author given)

Johnson, Samuel T. “ The Design of Organizations."


Encyclopedia Britannica, 1970, X, 776.

(An article in a reference work with no author given)

“Career Guidance.” The Modern Encyclopedia, 1975, XII, 456-


51.

Bulletins and Pamphlets


Organizations Planning, Washington, D.C.: American
Association of Administrators, 1978.

Motivating Employees. Dept. of Health, Education, and


Welfare. Washington, D.C.: U.S. Government Printing Office,
1971.
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 311
Citing Electronic References Notes
__________________
This is a relatively new area and stilt in a state of flux
__________________
however, the basic components of the reference citation are:
__________________
u Author [surname comma first initials]. __________________

u Title of work. __________________


__________________
u Title of complete work.
__________________
u Protocol and address. __________________
__________________
u Path.
__________________
u Date of message or visit.

Abbreviations
cf. Compare. Also abbreviated cp.

Et. Al, And others [et al l aliae].

e.g. for example [example gratial]. Not in italics;


preceded by comma.

etc. And so forth. Should never be used in formal writing

ff. following

ibid in the same place [ibidem]. Used to refer to the


book, chapter, article or page cited just before. Not
in italics.

Idem the same as; as mentioned before. Used to indicate


a reference previously mentioned, is generally used
to refer to the author’s name while ibid. refers to
the book title.

i.e. that is [id est). Not in italics and preceded by a


comma.

inter alia among other things. In italics.

loc. cit. in the place cited [loco citato]. Not in italics.

op. cit. in the work quoted (opere citato). Not in italics.

sup. above [supra]. In italics.


OLEUM &
E TR EN
FP

ER
YO

GY
u

UNIV ER SIT

ST U DI ES
312 Research Methodology fo
n ~;
k ; k ‘ k f D r o’
k~ e

Notes [sic] thus; so. Used to indicate that a surprising or dubious


__________________ word, phrase, or fact is not a mistake and is to read
__________________ as it stands. With brackets.
__________________
Ivde see.
__________________
__________________ vide infra see before. In italics.
__________________ see supra see above. In italics; abbreviated v.s
__________________
vis – a – vis face-to-face. With hyphens.
__________________
__________________ viz. namely. Not in italics.
__________________
Summary
A good marketing research report possesses of adhering
to the study objectives. A marketing research
report must ensure that it has covered all the
issues. The marketing researcher should logically
justify their choice of data, presentation of data and
findings.

The standard format of the research report may also be


used for report purpose. Deviations from this are
allowed to make it interesting and informative to the
reader by incorporating the nature of the product and
highlights of the research findings. The user must also
have an inherent understanding of the research process
used, the assumptions made and limitations of the
research, to rule out being taken for a ride on statistical
figures which are distorted at times to meet the
requirements of the marketer.

1. A research report is the presentation of the


research findings for a specific purpose. The
presentation can be written or given orally or both.

2. The guidelines for written and oral reports are


(a) consider the audience (b) be concise yet complete
and (c) be objective yet effective.

3. The main elements of the written report are (a) the


title page (b) the table of contents (c) the
Contd....
OLEUM &
E TR EN
FP
ER
YO

GY

u
UNIV ER SIT

ST U DI ES

fo
n ~;
k ; k ‘ kf D r o’
k~ e
UNIT 9 Report Writing 313
Notes
management summary (d) the body of the report
__________________
(e) the conclusions and recommendations and (f) the
__________________
appendix.
__________________
4. The research data can be presented in tabular or __________________
graphic form. The three types of graphic forms are
__________________
(a) the pie chart (b) the bar chart and (c) the line
__________________
chart.
__________________
__________________
__________________
__________________

You might also like