2. Meaning and Definition
Before defining ‘marketing research’, let us
determine research.
• Research always starts with a question or a problem.
• Its purpose is to find answers to questions through the
application of the scientific method.
• It is a systematic and intensive study directed towards a
more complete knowledge of the subject studied.
Research can be classified into two broad categories:
5/16/2021 2
Dr. Ravindra, IGU, Meerpur, Rewari
3. Types
• Basic Research
• Applied Research
Basic Research: Basic research is something called
‘fundamental research’, ‘theoretical research’ or ‘pure
research’. Its aim is to expending the frontiers of
knowledge and does not directly involve pragmatic
problems. The essence of basic research is that it
addresses itself to more fundamental questions and not to
the problems with immediate commercial potential.
5/16/2021 3
Dr. Ravindra, IGU, Meerpur, Rewari
4. Types
Applied Research: Which is also called ‘decisional
research’ , on the other hand, proceeds with a certain
problem and it specifies alternative solutions and the
possible outcomes of each alternative. Unlike basic
research, it is prompted by commercial considerations.
Though one may usually be able to distinguish between
basic research and applied research.
Applied research can be of two types:
(i) Problem – solving research
5/16/2021 4
Dr. Ravindra, IGU, Meerpur, Rewari
5. Types
(ii) Problem – oriented research
(i) Problem – solving research: As the name implies, is
concerned with a particular issue or a problem and is
usually proprietary in character.
(ii) Problem – oriented research: Problem – oriented
research on the other hand, is concerned with a class of
issues or problems in which several firms may be
interested. Research of this type is usually concerned with
conceptual aspects but is oriented towards applied
problems.
5/16/2021 5
Dr. Ravindra, IGU, Meerpur, Rewari
6. Definition
Marketing research is a systematic and objective study of
problems pertaining to the marketing of goods and services. It
may be emphasized that it is not restricted to any particular
area of marketing, but is applicable to all its phases and
aspect.
Definition
The American Marketing Association has Defined Marketing
Research as:
“Marketing research is the function which links the
consumer, and public to the marketer through information –
information used to identify and define marketing
opportunities and problems; generate, refine and evaluate
5/16/2021 6
Dr. Ravindra, IGU, Meerpur, Rewari
7. Definition
marketing actions; monitoring marketing performance;
and improve understanding of market as a process.
Marketing research specifies the information required to
address these; designs the method for collecting
information; manages and implements the data
collection process; analyses the results and
communicates the findings and their implications”.
The AMA definition is provide the information to
marketing decision makers, its also indicates the scope
and process of marketing research.
5/16/2021 7
Dr. Ravindra, IGU, Meerpur, Rewari
8. The Growth of Marketing Research
• The first evidence of formal research techniques
employed could be traced as far back as 1824.
• Almost one century thereafter, marketing research
developed rather slowly and sporadically, it was only
1910 that it made rapid strides.
Several factors have contributed to the growth of
marketing research.
• Large scale production.
• Producers do not have direct contact with customers.
5/16/2021 8
Dr. Ravindra, IGU, Meerpur, Rewari
9. The Growth of Marketing Research
• The problems of marketing of goods.
• Marketers do not know the potential areas where his goods
could be marketed.
• Shifting the market from seller’s to buyer’s.
• Emergence of specialist such as; statisticians, psychologists
and behavioral scientists in a fairly large number.
• Increase use of computer.
• Changes in the composition of population
• Changing marketing environment.
5/16/2021 9
Dr. Ravindra, IGU, Meerpur, Rewari
10. Scope of Marketing Research
Many people think that marketing research is just a consumer
survey, asking consumers about certain products or service.
Though consumer research is an integral part of marketing
research, the latter is quite a pervasive activity, covering the
various types of marketing problems that confront the
marketing manager. Here, a broad idea is given about the
scope of marketing research.
(1) Product Research: Product research is concerned with the
design, development and testing of new products as well as
improvement of existing one. Apart form this its includes;
5/16/2021 10
Dr. Ravindra, IGU, Meerpur, Rewari
11. Scope of Marketing Research
product mix and packaging.
(2) Sales Research: As the name implies, sales research is
concerned with all the selling activities of the firm. It
would like to examine various sales territories, product
wise performance along with overall performance,
position of main competitors, effectiveness of the sales
force and identifying main factors that may improve its
performance and alternative methods of distribution to
become distribution effective.
5/16/2021 11
Dr. Ravindra, IGU, Meerpur, Rewari
12. Scope of Marketing Research
(3) Customer Research: Customer research is concerned
with the studies of buyer behavior. It probes into different
factors such as sociological, economical, psychological,
demographical and environmental that influence the
customers in their decision to buy the product. Further the
research may be conducted to know the attitude,
perception, preference and level of satisfaction regarding
firms products and services. The firms may also ascertain
its overall market share as well as different segment level.
(4) Pricing Research: Regardless of the nature of the
5/16/2021 12
Dr. Ravindra, IGU, Meerpur, Rewari
13. Scope of Marketing Research
business, every firm has to decides what price is to be
fixed for its goods and services. Pricing is one of the
major factors that affects the success of the firm. Cos is
the major factors in determining the price, other factors
are; nature of demand, competition level, technological
development, financial status of the customers etc.
(5) Advertising Research: Another area in marketing
research is related to advertising. The main objective of a
firm using advertising is to increase awareness of its
products or service to the target group, leading eventually
5/16/2021 13
Dr. Ravindra, IGU, Meerpur, Rewari
14. Scope of Marketing Research
to increase sales. There are two types of advertising research;
(i) Media research
(ii) Copy research
Media research is concerned with alternative media –
television, radio, newspapers and magazines.
As regards copy testing, there are different methods that are to
be studied to determine the suitability of a particular copy
before it is finally released by the firm as an advertisement
like; them, heading, subjects, appeal type etc.
5/16/2021 14
Dr. Ravindra, IGU, Meerpur, Rewari
15. Limitations of Marketing Research
Previous discussion assume that the marketing research
can solve all the problems of marketing. While it can be
extremely rewarding to a firm, it is wise to know that it is
subject to certain limitations. One must be aware of these
limitations in advance so that one is clear about what
marketing research can and cannot do.
• Marketing research have fragmentary approach.
• Marketing research is become too superficial and faulty in
industry.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 15
16. Limitations of Marketing Research
• Absence of meaningful dialogue between the marketing
management and the marketing research team.
• Marketing research is not exact science.
• Analytical tools are still deficient.
• Marketing research is sometime miss-used.
• Some time its support someone position or viewpoint.
• Deliberately delaying the decision making.
• It is used to grab the power and authority in an
organization.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 16
17. Marketing Information System
Definition
‘A structured, interacting complex of persons, machines, and
procedures designed to generate an orderly flow of pertinent
information, collected from both intra – and extra – firm
sources for use as the basis for decision – making in specified
areas of marketing management.’
A good marketing information system should determine the
information needs of the organization and generate and
process such information on a continuing bases. It should also
provide for its storage so that it can be used when required.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 17
18. Marketing Information System
It seems worthwhile to specify some of the contrasting
characteristics of marketing research and a marketing
information system.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 18
Marketing Research Marketing Information System
1. Emphasis is on handling external
information.
1. Handles both internal and external data.
2. Concerned with solving problems. 2. Concerned with preventing as well as
solving problems.
3. Operates in a fragmented intermittent
fashion – on a project to project bases.
3. Operates continuously – is a system.
4. Tend to focus on past information. 4. Tends to be future oriented.
5. One source of information input for a
marketing information system.
5. Includes other sub-systems, besides
marketing research.
20. Marketing Information System
As can be seen from the figure MIS has four sub-systems
of which marketing research is one. Further, each sub-
system is linked with the other three and the flow of
information is both ways.
The box in left side of the figure show marketing
environment which marketing managers must understand
and analyze from time to time. This is concerned with
large markets, marketing channels, competitors, publics
and macro-environment forces. The information
emanating from these factors is picked up by the company
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 20
21. Marketing Information System
by one of four sub-systems making up MIS. From these
sub-systems, the information eventually flows to
concerned marketing managers as shown on the right side
of the figure. The information thus received is analyzed by
marketing managers so that the marketing activity of the
firm can be properly planned and implemented. The four
system of MIS are;
(1.) Internal Report System: An internal report system is
maintained by every business firm. Such a system
provides data on sales, costs, inventories, cash flows, and
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 21
22. Marketing Information System
account receivable and payable, etc. to management. Many
firms maintain a computer-aided internal reports system.
(2.) Marketing Intelligence System: A marketing intelligence
system provides marketing managers their day-to-day
information pertaining to the external marketing environment.
In fact, marketing managers themselves generate marketing
intelligence, by reading newspapers and business magazines.
The marketing intelligence received from managers, can be
further improved by the firm by taking certain steps.
(3.) Marketing Research System: A firm at times needs to
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 22
23. Marketing Information System
undertake a specific study on a certain problem/situation so
that it can take the right decision. In such cases, managers
cannot wait for information that comes in parts at different
times. In such situations, marketing research is frequently
resorted tp. Marketing research as defined by the American
Marketing Association has been given earlier in the previous
slide. To reiterate briefly, collection, analysis and
interpretation of data and finding on a specific marketing
situation or problem faced by a firm.
(4.) Analytical Marketing System: The analytical marketing
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 23
24. Marketing Information System
system consists of two sets of tools. The first tool is the
statistical bank. It has a number of advanced statistical
procedures for ascertaining the relationships within a set
of data. The second tool is the model bank. It has those
models which are helpful to marketing managers in
having a better understanding of the given problem. This,
in turn, facilitates them in decision making.
Issues in Marketing Research
There are certain issues which should be sorted out to
ensure further growth of marketing research. These are as;
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 24
25. Marketing Information System
• First the quality of information collected through
marketing research surveys will be considerably affected
on account of lack of transparency in business. This
problem can be overcome by having more and more
professionalization and institutionalization of marketing
research.
• Second, marketing research is frequently seen as a
separate activity undertaken by professionals and is not
linked with short or long term marketing strategies. It is
advisable to involve operational staff in a company with
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 25
26. Marketing Information System
marketing research activity. This will be advantageous to
both professionals and operational staff and will make
marketing research more useful to the business.
• Third, a major issue at the institutional level is the
allocation of budget to marketing research activity. Small
firms may use marketing research at their direction and
that too on modest scale. But large firms may have to use
it more frequently and, as such, they may have to allocate
adequate funds for it.
• Fourth, marketing research has so far remained confined
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 26
27. Marketing Information System
to the urban market. The marketing problems in the rural India
are vast and complicated and have hardly been addressed.
• Fifth, marketing research agencies should not merely have
contextual familiarity with the projects as at present. They
should be very well-informed about each sector such as
telecom research, pharmaceutical research, etc. The emphasis
should be on industry/sector specific research where more
sophisticated techniques can be used.
• Sixth, there is not yet adequate realization of the importance
of marketing information system and marketing decision
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 27
28. Marketing Information System
support system. In view of increasing competition in
business, companies or marketing research agencies must
have databases so that the requisite information can be
retrieved without any delay.
• Finally, as mentioned earlier, so far adequate attention has
not been given to the study of consumer behaviour, which
leads to qualitative research.
Characteristics of Good Research
The characteristics of good research are briefly described
below.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 28
29. Characteristics of Good Research
(1.) Clarity of Purpose: The research should be clear about
the purpose of the proposed research. The research
problem should be formulated carefully, indicating its
scope and limitations. Both the manager and the
researcher should have the same understanding of the
research problem and how they should proceed in
different stages of the research.
(2.) Research Process: The research process should be
properly planned. The researcher should ensure that
significant procedural steps are taken care of. In their
absence, it would not be possible to estimate the validity
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 29
30. Characteristics of Good Research
and reliability of data, which, in turn, will weaken the
confidence of the reader in the research itself.
(3.) Selection of Proper Research Design: There are several
research design, out of which one is chosen. In some cases, a
combination of two research designs may be used. However,
the choice of a research design has to be based on the nature
of the research problem. If a research design involves primary
data collection, a proper method of sampling should be
chosen. The main questions to be answered are; Whether
probability sampling or non-probability sampling?
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 30
31. Characteristics of Good Research
Within the probability sampling, which specific design is
to be used?
(4.) Data Collection: After having selected a research
design, the researcher should take the necessary steps to
collect data. He should exhaust all the sources of
information already available on the research problem. In
the case of primary data, a suitable questionnaire should
be developed. Further, enumerators who are assigned the
job of data collection, should be trained so that reliability
of data can be ensured.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 31
32. Characteristics of Good Research
(5.) Proper Analysis and Interpretation of Data: The
method (s) of analysis used should be appropriate keeping
in mind the nature of the problem and the purpose of the
research study. Hypothesis should be carefully formulated
and appropriate test should used for testing them. A
detailed analysis should be undertaken so that a deeper
insight into the research problem can be obtained.
(6.) Reporting the Results: The research report should be
written in a simple and unambiguous language. It should
be objective in the presentation of results. In case there are
any limitations, then these should be specified. After
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 32
33. Characteristics of Good Research
reading the report, the reader should be convinced about
the sincerity, objectivity and competence of the
researcher.
(7.) Compliance with Ethical Standards: It is necessary to
ensure that ethical standards or norms are followed
through all the stages of the research process. At times,
the researcher comes across a situation which creates an
ethical dilemma. The dividing line between ethics and
non-ethics becomes so blurred that it may be lead to a
wrong decision, that is, non-compliance with an ethical
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 33
34. Characteristics of Good Research
norm. The researcher has to be extremely vigilant in such
cases. The compliance with ethical norms will improve
the image of the research project as well as the entire
research team.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 34
35. Value of Information
Meaning
Before undertaking marketing research, the management has
to be convinced of its utility. If it feels that marketing research
will not be helpful in decision-making, it will not be
undertaken. Even when a decision to undertake research has
been taken by the management, there are several alternative
research design from which one is to be chosen.
Need for Information
There may be certain situations where management is
sufficiently clear that no additional information is likely to
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 35
36. Value of Information
change its decision. In such cases, it should be obvious that the
value of information is negligible. In contrast, there may be
certain situations where the decisions cry out for information
which is not available at any price. Between these two
extremes, there are a large number of situations where the
availability of additional information will improve the odds of
making a good decision.
Unless the information collected does not lead the manager to
change or modify his decision, the information has no value.
Generally, information is most useful in cases (i) where one is
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 36
37. Value of Information
insure of what is to be done, and (ii) where extreme values,
say, huge profits or losses, are involved.
However, it is not that any type of information available can
serve the purpose. In order that information is really useful to
the researcher, it must be;
• Relevant: This is the most important characteristics. It means
that the information has relevance to the subject of enquiry.
• Available: When a decision is being made, information must
be available.
• Accurate: The information, which is going to be used for
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 37
38. Value of Information
decision-making, must be accurate. Obviously, the use of
wrong information will lead to a wrong decision.
• Current: Information may be current. This implies that it
is not an obsolete one and is applicable to the time period
of the study.
• Adequate: In addition to the foregoing four
characteristics, information must be adequate. Scanty or
piecemeal information will not serve the purpose.
It may be mentioned that the available of information may
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 38
39. Value of Information
not have all the characteristics. As such, tradeoffs are usually
made among these characteristics, depending on their relative
importance to the research problem. However, as the most
important characteristics is relevance, it must not be
compromised in such a trade off.
Since the collection of information involves a cost, it is
necessary to insure that the benefit from information is more
than the cost involved in its collection.
• The Concept of Probability: The concept of probability is
the basis of decision-making under conditions of uncertainty.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 39
40. Value of Information
When we say that probability of certain event is 0.8, we
mean that the likelihood of its occurrence is 8 time out of
10 or 80 percent. The more we think that a particular
event is expected to occur, the greater will be its
probability and vice versa. There are three sources of
assigning probabilities.
• Logic/Deduction: It is quite simple to assign probabilities
in some situations. For example, when a coin is tossed, the
probability of getting a head or a tail is one-half or 0.5.
Likewise, when a dice is rolled, the probability of getting
any one of the six number (1 to 6) is 1/6. However, most
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 40
41. Value of Information
of the marketing research situations are not as simple.
• Past Experience/Empirical Evidence: Another source of
estimating probabilities is the past experience or empirical
evidence. A firm might have faced similar problems in the
past. The experienced gained in the process of resolving these
problems may be helpful in the estimation of probability
• Subjective Estimate: This is another method of estimating
probability and is perhaps the most frequently used basis of
probability estimates. When probability is to be estimated by a
manager, the main consideration is that who should be asked
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 41
42. Value of Information
to do so. Obviously, a well informed and knowledgeable
executive should be entrusted with this responsibility.
There is considerable criticism of the theory of decision
analysis, on account of the subjective method of
estimating probability. As such, one should use as much
data as possible even for assigning probability on a
subjective basis.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 42
43. Research Process
Meaning
The marketing research process involves a number of
inter-related activities which overlap and do not rigidly
follow a particular sequence.
In marketing research, even though our focus is on one
particular step, other inter-related steps or operations are
also being looked into simultaneously. As we complete
one activity or operation, our focus naturally shift from it
to the subsequent one, i.e. the focus is not concentrated
exclusively on one single activity or operation at any
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 43
44. Research Process
particular point of time.
Let us now describe the major steps involved in a
marketing research process.
1. Identify and Formulate the Problem
• Statement of the Problem in General Way.
• Understanding the Nature of the Problem.
• Survey the Available Literature.
• Developing the Idea Through Discussion.
• Rephrasing the Research Problem.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 44
45. Research Process
2. Extensive Literature Survey:
• Conceptual
• Empirical
3. Research Objectives: After detail and intensive literature
survey research is able to decide the research question or
research objective through which researcher wish to get
solution of the problem.
4. Developing the Research Hypothesis:
• Discuss with colleague.
• Examine the data and record available related to the problem.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 45
46. Research Process
• Review of similar studies in the area.
• Exploratory personal investigation or interview on a limited
scale.
5. Research Design: A research design specifies the methods
and procedures for conducting a particular study. Broadly
speaking, research designs can be grouped into three
categories – exploratory studies, descriptive research and
experimental/ causal research.
The choice of research design will depend essentially on the
nature of the problem on which the research is to be
undertaken. Another factor that is also relevant in this respect
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 46
47. Research Process
is the scope of the proposed study. If the scope is wide enough
covering a number of aspects, then the research design will be
very different from a study having an extremely limited scope.
As exploratory research focuses on the discovery of ideas and
is generally based on secondary data. A descriptive study is
undertaken when the researcher wants to know the
characteristics of certain groups such as age, gender,
educational level, income occupation, etc. A causal research is
undertaken when the researcher wants to know the cause and
effect relationship between two or more variables. Such
studies are based on reasoning along well tested lines.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 47
48. Research Process
Type of Research Design
Exploratory Studies Descriptive Studies Causal Studies
Literature Search Longitudinal Cross-sectional Natural Controlled
Experience Survey Experiment Experiment
Focus Groups Sample Survey (Several Types)
Analysis of
Selected Cases
True Panel Omnibus Panel
Time Series Cross-Sectional Combination
and trend design design of these two design
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 48
49. Research Process
(6) Determine the Sources of Data: The next step is to
determine the sources of data to be used. The marketing
researcher has to decide whether he has to collect primary
data or depend exclusively on secondary data. Sometimes,
the research study is based on both secondary and primary
data.
When a study is to be based on secondary data, whether
partially or fully, it is necessary to satisfy oneself that the
data are quite suitable for the objectives spelt out by the
study. It is also advisable to evaluate secondary data in
detail to avoid possible sources of error. To begin with,
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 49
50. Research Process
one should be familiar with the authentic sources of relevant
data, their periodicity, the agency publishing or having such
data, the concepts used in compilation and their limitation, if
any. A sincere effort must be made to look into the existing
data with a view to examining their suitability for the
research. It is only when such secondary data are unavailable,
inadequate, or unreliable, that a researcher should decide on
collecting fresh data.
(7.) Designing Data Collection Forms: Once the decision in
favour of collection of primary data is taken, one has to decide
the mode of collection. The two methods available are
observational methods, and survey method.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 50
51. Research Process
• Observation: This method suggests that data are collected
through one’s observation. If the researcher is a keen
observer, with integrity he would be in a position to observe
and record data faithfully and accurately. This method may be
suitable for some cases, but, attitude, opinion, motivations
and other intangible states of mind cannot be observed.
• Survey: In marketing research, field surveys are commonly
used to collect primary data from the respondents. Survey can
be (i) personal (ii) telephonic (iii) by mail and (iv) by diary.
Of these personal and mail surveys are more frequently used
in India. A choice has to be maid regarding the type of survey
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 51
52. Research Process
for collecting data. There are certain advantages and
limitations of each type of survey. Broadly speaking,
telephone survey is suitable when very limited
information is sought in a short period of time. In
contrast, survey based on personal interviews are suitable
when detailed information is to be collected. Sometimes
combination of two or more methods could also be used.
(8.) Sampling Design and Size: Another aspect which
forms a part of research process is the sampling plan.
When the marketing researcher has decided to carry out a
field survey, he has to decide whether it is to be census or
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 52
53. Research Process
sample survey. In almost all cases, a sample survey is
undertaken on account of its overwhelming advantages
over a census survey.
When a decision is in favour of a sample survey has been
taken, it is necessary to have a clear definition of the
population from which the sample is to be drawn, before
deciding on the type of sample design to be used. First, a
broad choice is to be made between probability sampling
and non-probability sampling. The researcher then selects
a specific type of sampling design from a number of
sample designs. The type of sample design chosen will
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 53
54. Research Process
depend on its suitability and also the availability of the
requisite sample frame.
As regards sample size there are two basic approaches –
the as statistical approach. Although the former is widely
used in marketing research, it should be the endeavor of
the researcher to follow the statistical approach which is
based on well – defined principles. The size of the sample
will depend on the degree of precision required as also on
the cost consideration.
(9.) Organizing and Conducting the Field Survey: Having
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 54
55. Research Process
prepared the questionnaire and selected the sample design
and size of the sample, the next step is to organize and
conduct the field survey. Two important aspects should be
looked into – interviewing and the supervision of the
field work. The task of interviewing seems to be simple
but, in reality, it is one of the most difficult tasks in
marketing research. This because the respondents are
generally hesitant in giving information unless approached
with fact, initiative and intelligence. Supervision of field
work is equally important to ensure timely and proper
completion of the field survey. Neglecting these aspects
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 55
56. Research Process
would result in interviewing errors, which, in turn, would
undermine the utility of the survey.
(10.) Processing and the Analyzing the Collected Data:
Once the field survey is over and questionnaires have
been received, the next task is to aggregate the data in a
meaningful manner. A number of tables are prepared to
bring out the main characteristics of the data. The
researcher should have a well though out framework for
processing and analyzing data, and this should be done
prior to the collection. It is advisable to prepare dummy
tables, as such an exercise would indicate the nature and
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 56
57. Research Process
extent of tabulation as also the comparisons of data that can
be undertaken.
In order to drive meaningful results from the statistical tables,
the researcher may use one or more of the following four
steps. The first step is to calculate relevant measures of central
tendency as also of dispersion, highlighting the major aspects
of data. The second is to cross – tabulate the data to ascertain
some useful relationships. The third is to calculate the
correlation coefficient and undertake a regression analysis
between variables. The fourth is to undertake a multivariate
analysis.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 57
58. Research Process
(11.) Preparing the Research Report: Finally, the
researcher has to prepare the report of what has been done
by him. Writing of report must be done with great care
keeping in view the following.
(a) The layout of the report should be as follows: the
preliminary pages, the main text, and end matter.
In its preliminary pages the report should carry title and
date followed by acknowledgements and foreword. Then
there should be a table of contents followed by a list of
tables and list of illustrations, if any, given in the report.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 58
59. Research Process
The main text of the report should have the following
parts;
• Introduction: It should contain a clear statement of the
objective of the research and an explanation of the
methodology adopted in accomplishing the research. The
scope of the study along with various limitations should as
well be stated in this part.
• Summary of the findings: After introduction there would
appear a statement of findings and recommendations in
non – technical language. If the findings are extensive,
they should be summarized.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 59
60. Research Process
• Main Report: The main body of the report should be
presented in logical sequence and broken –down into
readily identifiable sections.
• Conclusion: Towards the end of the main text, researcher
should again put down the results of his research clearly
and precisely. In fact, it is the final summing up.
At the end of the report, appendices should be enlisted in
respect of all technical data. Bibliography, i.e. list of
books, journals, reports, etc., consulted, should also be
given at the end. Index should also be given specially in a
published research report.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 60
61. Research Process
(b) Report should be written in a concise and objective style
in simple language avoiding vague expressions such as ‘it
seems’, ‘there may be’, and the like.
(c) Chart and illustrations in the main report should be used
only if they present the information more clearly and
forcibly.
(d) Calculated ‘confidence limits’ must be mentioned and the
various constraints experienced in conducting research
operations may as well be stated.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 61
62. Error in Research Process
The forgoing discussion provides some idea of the steps
involved in the research process. It also gives list of question
that need to be addressed at various stages of the research
process. This apart, an additional aspect need to be discussed
here. The researcher should insure that the research does not
have a high degree of error. The errors are of two types –
sampling errors and non-sampling errors. These two together
constitute the total error.
Sampling Errors: Quite frequently, marketing research
studies are based on samples of people or products or stores.
The results emerging from such studies are then generalized,
i.e. applied to the entire population. For example, if a study is
done amongst Maruti Car Owners in a city to know their
average monthly expenditure on the maintenance of their car,
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 62
63. Error in Research Process
it can be done either by covering all Maruti Car Owners
residing in that city or by choosing a sample, say 10%, of
the total Maruti Car Owners. In the latter case, the study
may give a different average than the actual average if the
entire population is covered. This difference between the
sample value and the corresponding population value is
known as the sampling error.
Non-sampling Errors: Non-sampling errors as the name
implies, are all those errors which occur in different stages
of research except in the selection of sampling. Obviously,
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 63
64. Error in Research Process
these errors are many and varied. A non-sampling error can
arise right at the beginning when the problem is defined
wrongly. It can also occur in any of the subsequent stages
such as in designing a questionnaire, non-response of the
questionnaire, in the analysis and interpretation of data, etc.
An important point to note is that sampling error is measurable
while it is not easy to measure a non-sampling error. Another
point to note is that sampling error decreases as the sample
size increases. This is not necessarily so in the case of non-
sampling error. As the chance of non-
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 64
65. Types of Non-Sampling Errors
sampling errors are several during the course of study, the
marketing researcher has to be very particular to ensure an
effective control over them. Different types of non-
sampling errors are briefly explained below.
Types of Non-sampling Errors
(i) Defective Problem Definition: Problem on which
research is to be undertaken should be precisely defined.
For example, a study on unemployment must be clear as
to the concept of unemployment, the reference period, the
geographic area to be covered, and so on. If any of these
basic concepts has wrong connotation, the results of the
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 65
66. Types of Non-Sampling Errors
study would turn out to be wrong.
(ii) Defective Population Definition: If the population is not
well-defined and does not fit to the objects of research study
then an error occurs.
(iii) Frame Error: The sampling frame is the list of all units
comprising the population from which a sample is to be
drawn. If the sampling frame is incomplete or inaccurate, its
use will give rise to this type of error.
(iv) Surrogate Information Error: This type of error occurs
when the information sought by the researcher is different
from the information needed to solve the problem. A simple
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 66
67. Types of Non-Sampling Errors
example of this error is when price of brand is taken to
represent its quality. In such a case, it is presumed that higher
the price of the brand, the better is its quality. This may or
may not be true.
(v) Non-response Error: It is almost impossible to obtain data
from each and every respondent covered in the sample. There
are always some respondents who refuse to give any
information. Thus, non-response error occurs when
respondents refuse to cooperate with the interviewer by not
answering his questions.
(vi) Measurement Error: This is caused when the information
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 67
68. Types of Non-Sampling Errors
gathered is different from the information sought. For
example, respondents are asked to indicate whether they own
a colour television set. Some of them may respond in the
affirmative just to boost their image before an interviewer,
even though they may not be owning a colour television set.
Such response will result in measurement error. This type of
error occurs frequently in research studies.
(vii) Experimental Error: An experiment aims at measuring the
impact of one or more independent variables on a dependent
variable. Take the case of the impact of training on the
performance of salesmen. During the period when the training
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 68
69. Types of Non-Sampling Errors
is given, there may be a decline in competition and as a
result sales performance may improve. The results of such
an experimental study will be misleading.
(viii) Poor Questionnaire Design: a questionnaire is an
instrument to collect the data from respondents in a
survey. If the questionnaire is defective, the data collected
on that basis will be misleading.
(ix) Interviewer Bias: This error occurs on account of
interviewer’s influence in conducting an interview or
wrong recording by him. By putting emphasis on a certain
word or phrase in a questionnaire, interviewers can
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 69
70. Types of Non-Sampling Errors
influence respondents to answer in particular way.
(x) Data Processing Error: After the data have been
collected, they are to be processed. This involve coding
the responses, recording the codes, etc., so that data
collection can be transformed into suitable tables. Mistake
can occur during the processing stage of data.
(xi) Data Analysis Error: As in the case of data processing,
error can occur on account of wrong analysis of data.
Apart from simple mistakes in summation, division, etc.,
more complex errors can occur.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 70
71. Types of Non-Sampling Errors
(xii) Interpretation Error: Sometime wrong interpretation
of data can cause this type of error. In order to support a
particular line of action, the researcher may deliberately
misinterpret data. Wrong interpretation of data may occur
at time without one being aware of it.
◌ ------------------- ◌
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 71
72. Research Design
As research should be based on scientific method, we
should first know what the scientific method is. The
scientific method encourages doubts and criticism so that
what emerges is the real evidence which has stood the test
of reasoning. It make science progressive as it is never too
sure about its results. A unique characteristics possessed
by this method is self-correction. A scientist does not
believe any proposition without testing it. He has a
number of built-in checks all along the way to enable him
to adhere to the right path and arrive at the ‘truth’. Such
check are free from personal beliefs, attitudes and values.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 72
73. Research Design
Karl Pearson, observes that there are three distinct
characteristics of the scientific method;
(a) Careful and accurate classification of facts and observation
of their correlation and sequence.
(b) The discovery of scientific laws with the aid of the creative
imagination and
(c) Self-criticism and the final touchstone of equal validity for
all normally constituted minds.
Research Design
Having discuss about the scientific method, we now turn to
research design.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 73
74. Research Design
According to Kerlinger: ‘Research design is the plan,
and strategy of investigation conceived so as to obtain
answer to research questions and to control variance’.
According to Green and Tull: ‘A research design is the
specification of methods and procedures for acquiring the
information needed. It is the over all operational pattern or
framework of the project that stipulates what information
is to be collected from which sources by what
procedures’.
From the forgoing definitions it is evident that research
design is more or less a blueprint of research. It can be
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 74
75. Research Design
compared with the plan of house, which lay down the
method and procedure for the collection of requisite
information and its measurement and analysis with a view
to arriving at certain meaningful conclusions at the end of
the proposed study.
Types of Research Design
At the outset it may be noted that there are several ways of
studying and tackling a problem. There is no single
perfect design. Research designs have been classified by
various authors in different ways. Different types of
research designs have emerged on account of the different
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 75
76. Research Design
perspectives from which a research study can be viewed.
However, a frequently used classification system is to group
research designs under three broad category – exploratory,
descriptive and causal/experimental.
1. Exploratory Research: As its name implies, the researcher
explore the nature of the problem and is free to proceed in
whatever manner he wants to. Here, the focus is on the
discovery of ideas as hardly anything is known about the
problem to be studied. When the researcher embarks on an
exploratory study, his approach is very flexible. In the
beginning, he may be having vague ideas but as he goes on
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 76
77. Research Design
getting more and more information, he more clear about the
research problem. He may get new ideas to be examined to
ascertain how far they are relevant to the problem.
With this brief introduction, we now turn to some methods
that are generally used in exploratory studies. The following
are the sources of exploratory studies;
• Literature survey
• Experience survey
• Focus group
• Case studies
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 77
78. Research Design
2. Descriptive Studies: Having looked into exploratory research,
we now turn to descriptive studies. A good amount of
research in marketing can be considered descriptive research.
Descriptive research is conducted when the purpose is as
follows;
(i) To determine the characteristics of a certain group.
(ii) To estimate the proportion of people in a given population
who have behaved in a particular manner.
(iii) To make specific predictions.
There is a general belief that descriptive studies are factual
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 78
79. Research Design
and are very simple. This is not necessarily true.
Descriptive studies can be complex, demanding a high
degree of scientific skill on the part of the researcher.
Unlike an exploratory study which is flexible, a
descriptive study is considered rigid. The objective of
such a study is to answer the “who, what, when, where,
and how” of the subject under investigation.
Descriptive studies are well-structured. As was mentioned
earlier, an exploratory study needs to be flexible in its
approach, but a descriptive study, in contrast, tends to be
rigid and its approach cannot be changed every now and
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 79
80. Research Design
then. It is therefore, necessary that the researcher gives
sufficient thought to framing research questions and
deciding the type of data to be collected and the procedure
to be used for this purpose. If he is not careful in the initial
stages, he may find that either the data collected are
inadequate or the procedure used is cumbersome and
expensive.
Descriptive studies can be divided into two broad
categories – cross-sectional and longitudinal. These two
types further categories into four categories.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 80
81. Research Design
Cross-sectional Studies: A cross-sectional study is concerned
with a sample of elements from a given population. Thus, it
may deal with households, dealers, retail stores or other
entities. Data on a number of characteristics from the sample
elements are collected and analyzed. Cross-sectional studies
are of two types
• Field studies
• Surveys
The distinction between them is not clear-cut, there are some
practical differences which need different techniques and
skills.
Field studies are ex-post-facto scientific enquiries that aim at
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 81
82. Research Design
finding the relations and inter-relations among variables in
a real setting. Such studies are done in life situation like
communities, schools, factories, organizations, and
institutions.
Another type of cross-sectional study is survey research.
A major strength of survey research is its wide scope.
Detailed information can be obtained from a sample of a
large population. Besides, it is economical as more
information can be collected per unit of cost. Also, it is
obvious that a sample survey need less time than a census
inquiry.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 82
83. Research Design
Longitudinal Studies: As mentioned earlier, longitudinal
studies involve panels. There are two types of panels; true
panels and omnibus panels. True panels involve
repeated measurements of the same variables. It is
important to know that each sample member is measured
each time on the same characteristics over a period of
time, such data will reflect changes in the buying
behaviour of families. In an omnibus panel, a sample of
elements is still selected and maintained, but the
information collected from the sample members varies.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 83
84. Research Design
At one time, it may be the evaluation of an advertising copy
while at another time the attitude of panel members towards a
new soft drink. This shows there is a clear difference between
the true panel and the omnibus panel.
Having discussed two researches – exploratory and descriptive
now turn to casual research design. As the name implies, a
casual research design investigates the cause and effect
relationship between two or more variables.
Concept of Causality
The concept of causality is complex. Let us take, for example
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 84
85. Research Design
the statement “X causes Y”. This implies that any change
in Y is on account of X. This is how we commonly
understand. However, a scientist will interpret this
statement in a different way. He would say X is only one
of the factors causing a change in Y. Further, while
everyday interpretation implies a deterministic
relationship, in contrast, the scientific interpretation
implies a probabilistic relationship. Yet, another point of
difference between scientific interpretation and everyday
interpretation is that the former implies that it can never
be proved that X causes Y. We can only infer on the basis
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 85
86. Research Design
of the data that a relationship between X and Y exists. The
inference is based on some observed data that might have
been obtained in a very controlled experimental setting. Even
so, the scientific notion recognizes the feasibility of such
procedures.
Conditions for Causality
This poses the major question; What kind of evidence is
needed to support causal inferences? In this respect, it is
necessary to satisfy three conditions, namely; concomitant
variation, time order of occurrence of variables, and
elimination of other possible causal factors. It may be noted
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 86
87. Research Design
here that these conditions are necessary but not sufficient
for causality. This means that even when all the three
conditions are satisfied in a given case, it cannot be said
decisively that there exists a causal relationship.
Concomitant Variation: Concomitant variation, or
invariant association, is a common basis for ascribing
cause. Concomitant variation is the extent to which a
cause, X, and an effect , Y, occur together or vary together
in the way predicted by the hypothesis under
consideration. Suppose a company spends each year on
advertising its products in different parts of the country.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 87
88. Research Design
At the end of a year, the company finds that high sales
with limited advertising expenditure.
On the basis of this information, we can say that there is
concomitant variance between advertising campaign and
sales. Note that the implied hypothesis here is that
advertising campaign cause sales to increase. It may be
emphasized that we cannot decisively say that advertising
is the cause of increased sales. At the most, we can say
that the association between advertising expenditures and
sales performance makes the hypothesis more tenable.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 88
89. Research Design
Time Order of Occurrence of Variables: A casual
relationship between two variables on the basis of time
order of their occurrence is conceptually simple. It implies
that a variable cannot be considered the cause of another
if it occurs after the other event. The occurrence of the
casual factor must occur earlier or simultaneously with the
occurrence of an event. The conceptually it is simple, but
its application is not that simple. The researcher must have
a thorough understanding of the time sequence governing
the phenomenon.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 89
90. Research Design
Elimination of other Possible Casual Factors: The
elimination of other possible casual factors indicates that
when a casual relationship between two variables is being
studied, then if there is one or more other possible factors that
may effect that relationship, it should be eliminated. The
relationship should be exclusively between the two factors.
It is worth nothing that even when these three conditions,
namely, evidence of concomitant variation, time order of
occurrence of variables, and elimination of other possible
casual factors are combined, it is not possible to say
conclusively that a casual relationship exists. However, when
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 90
91. Research Design
more evidence pointing to the same conclusion become
available from several investigations, our confidence on
the casual relationship will increase. Again, if we have
deep conceptual knowledge of the problem situation,
confidence is further strengthened.
Some Definition and Concepts
1. Experiment: 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 survey and observational studies, there is no
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 91
92. Research Design
manipulation of independent variables by the researchers. This
is the fundamental difference between experimental and non-
experimental research.
2. Treatments: Treatments are the alternatives or independent
variables that are manipulated and whose effects are
measured. Examples in marketing include product
composition, advertising executions, price level etc. In a
measurement sense, treatments need only from a normal scale.
3. Test Units: Test units are the entities to whom the treatment
are presented and whose response to the treatments is
measured. It is common in marketing for both people and
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 92
93. Research Design
physical entities, such as stores or geographic areas, to be used
as test units. For example, people may be asked to try a
product and then have their attitude toward it measured. Here
people are the test units.
4. Dependent Variables: These are the measures taken on the
test units. Typical marketing examples include sales,
preference, awareness, etc. In a measurement sense, the
dependent variables must form an internal scale.
5. Extraneous Variables: These are the variables other than the
treatments that affect the response of the test units to the
treatments. These variables can distort the dependent variable
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 93
94. Research Design
measures in such a way as to weaken or invalidate one’s
ability to make casual references.
6. Experimental Design: As experimental design involves
the specification of (1) treatment that are to be
manipulated, (2) test units to be used, (3) dependent
variables to be measured, and (4) procedures for dealing
with extraneous variables.
Symbols
To facilitate our discussion of specific experimental
designs, we will make use of a set of symbols that are now
almost universally used in marketing research.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 94
95. Research Design
• X represents the exposure of a test group to an empirical
treatment, the effect of which are to be determined.
• O refers to the processes of observation or measurement of the
dependent variable 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 treatment.
• 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
event s that occur simultaneously.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 95
96. Casual Design
Casual Studies
Having considered the meaning of causation and the types
of evidence required to infer causal relationships in the
preceding slides, we now turn to some specific designs of
casual studies.
Casual inference studies can be divided into two broad
categories – natural experiments and controlled
experiments. The main point of distinction between the
two is the degree of intervention or manipulation
exercised by the investigator in a given study. Thus, a
natural experiment will involve hardly any intervention of
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 96
97. Casual Design
the investigator, except to the extend required for
measurement. A controlled experiment, in contrast, will
involve his intervention to control and manipulate variables.
Natural Experiments: There are three classes of designs for
natural experiments – (i) time series and trend designs, (ii)
cross sectional designs, and (iii) a combination of the two.
(i) Time series and trend design: In a time series design, data
are obtained from the sample or population at successive
intervals. Generally, current data are obtained from a panel of
individuals or households. Other panels such as wholesalers,
retail stores or manufacturer can also be used. While time
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 97
98. Casual Design
data relate to the same sample, trend data relate to
matched samples drawn from the same population at
successive intervals. Thus, there is no continuity in the
sample in trend designs, as a result of which data can be
analyzed in the aggregated form. Since time-series data
relate to the same sample over time, change in individual
sample units can be analyzed. An analysis of this type is
also called longitudinal analysis.
There can be several variants of time series and trend
designs. At one extreme they need at least one treatment
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 98
99. Casual Design
and a subsequent measurement, at the other extreme, they may
have several treatments and measurements. A few types of
time-series and trends designs are briefly discussed below.
(a) After-only without control group: The design is also
known as the ‘try out’ or one –shot case study design. It is
the simplest and can be shown symbolically as follows:
X O (1)
where ‘X’ indicates the exposure of a subject or group to an
experimental treatment whose effect is to be observed, and
‘O’ indicates the observation or measurement taken on the
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 99
100. Casual Design
subject or group after an experimental treatment. Suppose
that we provide training to a group of salesmen (X) for a
certain period and then measure the sales effected by this
group of salesmen (O). Since we do not have a prior
measurement of sales made by each of the salesmen
trained, it is not possible to measure the effect of training.
On account of this major limitation, the design is used
only in exploratory research. It should be avoided as far as
possible.
(b) Before-after without control group: This design differs
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 100
101. Casual Design
from the preceding one in one respect, i.e. it has a prior
measurement as well. Symbolically, it can be shown as;
O1 X O2 (2)
and in an extended form as;
O1 O2 O3 X O4 O5 O6 (3)
Taking our earlier example, the sales made by salesmen at
period (1) are known to us. We now provide them training for
a certain period, and then measure their sales. A comparison
of sales after training is made with sales made during the
corresponding period before training. Thus the effectiveness
of training can be measured by O2 – O1. The extended form
of the design (3), is an improvement over design (2) as it show
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 101
102. Casual Design
that sales made by a group of salesmen X are measured
for three successive periods prior to training and three
successive period after their training.
(c) Multiple time series: Another time-series design
involves the control group, symbolically;
O1 O2 O3 X O4 O5 O6
O’1 O’2 O’3 O’4 O’5 O'6 (4)
where the O’s represent measurement of the control
group. This design is an improvement over design (3) as it
measures the effect of a specific treatment on the
experimental group and compares it against the control
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 102
103. Casual Design
group. Thus taking our earlier example of training
salesmen, two comparable groups of salesmen are
selected. Before the treatment, i.e. training, sales made by
them for three successive time periods are measured for
both the groups. Now the experimental group is given
training. After the training, sales made by the
experimental group as also those made by the control
group (which was not given training) are measured. The
difference between the average sales for the experimental
group and the control group may then be attributed to
training.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 103
104. Casual Design
(ii) Cross-sectional Designs: In cross-sectional designs, the effect of different
levels of treatments are measured on several groups at the same time.
Symbolically, a cross-sectional design may be shown as follows;
X1 O1
X2 O2
X3 O3
X4 O4
Thus, subscripts 1, 2, 3, 4 show different groups of X given differing
treatments. The corresponding measure after the treatments are indicated by
O1, O2, O3 andO4 . Cross-sectional designs are used when varying levels of
advertising is done for the same product but in different territories or
when varying prices are fixed in different territories. The impact of
varying levels of treatment is studied on the basis of the sales of the
product in different territories.
The design also suffers from some of the limitations applicable to earlier
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 104
105. Casual Design
designs. Thus, there may be extraneous factors that may affect
the sale in a particular territory.
(iii) Combination of Cross-sectional Time-series Design:
These designs, as the name implies. Combine the time-series
and cross-sectional designs. While there can be many variants,
a more frequently used design is the ex-post-facto test-control
group. The design can be shown symbolically as follows;
O1 X O3
O2 O4 (6)
Such a design is well suited to continuous panel data. A
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 105
106. Casual Design
certain advertisement (X) is run and panel members are
then asked if they had seen it either. Those who had many
constitute a test or experimental group and those who had
not form the control group. It may be noted here that the
experimental and control groups would not be known until
after the advertisement was run. The impact of the
advertisement is measured by comparing the difference in
purchases made by the experimental and control groups
before and after the advertisement.
It will be seen that after the experimental and control
groups were formed on the basis of whether the panel
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 106
107. Casual Design
members had seen the advertisement or not. This self-
selection feature of the design may be a source of
systematic error. Besides, the testing effect may contribute
to inaccuracy. Despite these limitations, this design
provides data both cheaply and promptly if the panel
already exists.
Controlled Experiments
In controlled experiments, two kinds of intervention on
the part of the researcher are required. The first related to
the manipulation of at least one assumed independent or
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 107
108. Casual Design
causal variable. In order to measure the effect of one or
more treatments on the experimental variable, it is
necessary that the researcher manipulates at least one
variable. The second intervention relates to the assignment
of subjects to experimental and control groups on a
random basis. This is necessary so that the effect of
extraneous factors can be controlled. As the size of the
experimental and the control group increases, the effect of
extraneous factors on these groups can be equalized or
balanced by using a random selection procedure.
A few controlled experimental designs are as follow;
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 108
109. Casual Design
(i) After-only with Control Group: This is the simplest of all
the controlled experimental designs. In this design, only one
treatment is given and then both the experimental and the
control groups are measured. Symbolically, it can be shown
as follows:
R X O1
R O2 (7)
It has been criticized on the ground that it does not concern
itself with the pre-test. However, by avoiding the pre-test,
the design provides control over the testing and instrument
effects. This design is particularly suitable in those cases
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 109
110. Casual Design
where before measurement or pre-testing is not possible or
where testing and instrument effects are likely to be serious.
(ii) Before-after with one Control Group: This design provides
for pre-testing or before measurement. It can be shown
symbolically as follows:
R O1 X O2
R O3 O4 (8)
Unlike design (6), this design provides for the selection of the
experimental and control groups through the random method.
The design is able to control most of the sources of systematic
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 110
111. Casual Design
error. Both maturation and the testing effect may be taken
as controlled in this design because of their presence in
both the experimental and control groups. As the two
before measurements, O1 and O3 and the two after
measurements, O2 and O4 are made at the same points in
time, the design is able to control history.
With the help of this design, one can measure the effect of
treatments in three ways; O2 – O1, O2 –O4 and (O2 –O1) –
(O4 – O3). If these measurement show similar results, the
effects of experimental treatment can be inferred with
greater confidence.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 111
112. Casual Design
(iii) Four-group, Six-study Design: When the investigator
has to obtain data from respondents in an undisguised
manner, the ‘before-after with control group’ design, such
as the preceding one, is not suitable. This is because both
the experimental and control groups are likely to be
influenced by the before measurement. To overcome this
difficulty, a four group, six-study design may be used.
Such a design is extremely suitable in all those cases
where some sort of an interaction between the respondent
and the questioning process take place. Symbolically, the
design can be shown as follows:
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 112
113. Casual Design
R O1 X O2
R O3 O4
R X O5
R X O6 (9)
This is combination of designs (7) and (8). The effect of
the treatment can be measured in several ways such as O2
– O1, O4 – O2, O6 –O5, O4 – O3 and (O2 –O1) – (O4 – O3).
The after measurements can be shown in a 2 × 2 table us
follows:
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 113
114. Casual Design
No X X
Before Measurement taken: O4 O2
No before measurement taken: O6 O5
The difference between the ‘No X’ and ‘X’ column means
shows the effect of the treatment. Similarly, the difference
between the row means indicates the basis for estimating
the testing effect. Further, the interaction of testing and
treatment can be estimated from the differences in the
individual cell means. Finally, the combined effect of
history and maturation may be estimated by O6 –O1, O6 –
O3 and O4 –O1.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 114
115. Experimental Errors
After having described the different types of experiments,
we now turn to sources of potential errors in experiments.
There are several errors which may distort the accuracy of
an experiment. These are briefly described below.
1. History: History refers to the effect of extraneous
variables as a result of an event that is external to an
experiment occurring at the time as the experiment.
2. Maturation: Although maturation is similar to history, it
differs from it as the actual outcome is usually less
evident. Maturation refers to a gradual change in the
experimental units arising due to the passage of time.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 115
116. Experimental Errors
3. Pre-measurement Effect: this error is caused on account
of the changes in the dependent variable as a result of the
effect of the initial measurement. For example, consider
the case of respondents who were given a pre-treatment
questionnaire. After their exposure to the treatment, they
were given another questionnaire, an alternative form of
the questionnaire completed earlier. They may respond
differently merely because they are now familiar with the
questionnaire. In such a case, respondents’ familiarity
with the earlier questionnaire is likely to influence their
responses in the subsequent period.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 116
117. Experimental Errors
4. Interactive Testing Effect: This error arising on account
of changes in the independent variable as a result of
sensitizing effect of the initial measurement. In other
words, the first observation affects the reaction to the
treatment. For example, consider the case that respondents
have been given a pretreatment questionnaire that asks
questions about various brands of hair oil. The
pretreatment questionnaire may sensitizes them to the hair
oil market and distort the awareness level of new
introduction, i.e. the treatment. In such a case, the
measurement effect cannot be generalized to non-
sensitized persons.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 117
118. Experimental Errors
5. Instrumentation: Instrumentation refers to the changes in
the measuring instrument over time.
6. Selection Bias: Selection bias refers to assigning of
experimental unit in such a way that the groups differ on
the dependent variable even before the treatment. Such a
situation arises when test units may choose their own
groups or when the researcher assigns them to groups on
the basis of his judgment. To overcome this bias, it is
necessary that test units are assigned to treatment groups
on a random basis.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 118
119. Experimental Errors
7. Statistical Regression: Statistical regression effect occurs
when test units have been selected for exposure to the
treatment on the basis of an extreme pretreatment
measure. For example, a training programme may be
devised only for salesmen whose performance have been
very poor. Sales increases in the post-treatment period
may then be attributed to the regression effect. This is
because random occurrences such as weather, health or
luck may contribute to the better performance of salesmen
in the subsequent period. Thus the effect of training
programme may get distorted on account of this factor.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 119
120. Experimental Errors
8. Mortality: Mortality refers to the loss of one or more test
units while the experiment is in progress. It may be
emphasized that mortality leads to the differential loss of
respondents from the various groups. This means that
respondents who left, say group A are different from those
who left group B, thus making the groups incomparable.
In case the experimental pertains to only one group,
mortality effect occurs when responsiveness of the
respondents who have remained in the experiment differs
from responsiveness of those who have ceased to be in the
experiment.
5/16/2021 Dr. Ravindra, IGU, Meerpur, Rewari 120