RESEARCH METHODS IN

MANAGEMENT
(Study material)

Chapter 1
Research Design
1.1. Formulating Variables
A Variable is a characteristic of a person,
object or phenomenon which can take on
different values. The data is the observed
outcome of these variables. A variable can
be numerical or categorical.
Variables

Categorical Numerical

Ordinal Nominal Discrete Continuous

Fig. 1.1 Type of Variables

1
1.1.1. Numerical Variables
Numerical variables are expressed in
numbers. They can be classified into two
discrete and continuous variables.
Discrete variables can take certain fixed
integer numerical values, e.g. the number
of cars sold by Maruti Udyog Ltd. in 2005,
number of employees in an organization,
number of books in a library, number of
visits to a clinic (0,1,2,3,4, etc), size of the
family.
Continuous variables can take any
numerical value, for e.g. height in
centimeters (2 cm or 2.5 cm or 2.546 cm
or 2.543216 cm), temperature in degrees
Celsius (37.20C or 37.199990C etc.),
weight in grams/kilograms, length in
centimeter/inches.
1.1.2. Categorical Variables
Categorical variables are those which are
not expressed in numerical terms. They
are expressed in categories. For example,
the variable gender has two distinct
categories, male and female. Categorical
variables can either be ordinal or nominal.

2
 Ordinal variables can be ordered or ranked
in increasing or decreasing order, For
example: Income level - High income
(above Rs. 30,000 per month); Middle
income (Rs. 10,000-Rs. 30,000 per
month); and Low income (less than Rs.
10,000 per month), Agreement with a
statement - fully agree, partially agree,
fully disagree, Customer Satisfaction –
very satisfied, somewhat satisfied,
somewhat dissatisfied, very dissatisfied.
Nominal variables do not have an order or
ranking in them. For example: religion
(Hindu, Muslim, Christian, Sikhism, etc),
language (English, Hindi, French, German,
Tamil etc.). The only comparisons that can
be made between these variable values
are equality and inequality. There are no
"less than" or "greater than" relations
among the nominal variables, nor
operations such as addition or subtraction.
E.g. the variables of interest can be
behaviour (Past, present or intended),
demographic characteristics (age, sex,
income, and occupation), level of
knowledge, attitudes and opinion.
1.2. Type of Research Designs

3
 There are two primary types of research
designs – QUANTITATIVE and
QUALITATIVE. Quantitative research
focuses on outcomes while qualitative
research focuses on process. Quantitative
research involves collecting numbers and
qualitative is concerned with collecting
observations.
1.2.1. Quantitative Research Design
QUANTITATIVE research approaches
involve the collection of numerical data in
order to explain, predict, and/or control
phenomena of interest. Data analysis is
mainly statistical. Quantitative research
design can be classified into Exploratory,
Descriptive, Correlation, Causal-
comparative and Experimental Research.
1.2.1.1. Exploratory
It has the goal of formulating problems
more precisely, clarifying concepts,
gathering explanations, gaining insight,
eliminating impractical ideas, and forming
hypotheses. Exploratory research can be
performed using
 A literature search
 Surveying certain people about their
experiences:

4
 When surveying people, exploratory
research studies would not try to acquire a
representative sample, but rather, seek to
interview those who are knowledgeable
and who might be able to provide insight
concerning the relationship among
variables.
 Focus groups and case studies:
Case studies can include contrasting
situations or benchmarking against an
organization known for its excellence.
Exploratory research is characterized by
its flexibility.
1.2.1.2. Descriptive
Descriptive research is more rigid than
exploratory research and seeks to
describe users of a product, determine the
proportion of the population that uses a
product, or predict future demand for a
product. Descriptive research involves
collecting data in order to test hypotheses
or describe a behavior. In contrast with
the qualitative approach the data are
numerical. The data are typically collected
through a questionnaire, an interview, or
through observation. Generally these data
are analyzed using Chi Square,

5
 percentages, standard deviations and so
forth. A question, which could be
answered by this type of research is, “Is
there a difference between how
engineering students perceive science and
how arts students perceive science as a
list of facts to be memorized?” In
descriptive research, the investigator
reports the numerical results for one or
more variables on the subjects of the
study. For example: Attitude of employees
regarding decentralization in decision-
making; Customer satisfaction about a
particular product; Way in which
unproductive employees spend their time.
1.2.1.3. Co-relational
Co-relational research attempts to
determine whether and to what degree, a
relationship exists between two or more
numerical variables. Statistics used for
this analysis are the correlation coefficient
to report the results. Some examples of
co-relational research are: The
relationship between intelligence and self-
esteem; The relationship between anxiety
and achievement; The use of aptitude test
to predict success in management course.

6
1.2.1.4. Causal-comparative
Causal-comparative research attempts to
establish cause-effect relationships among
the variables of the study. The attempt is
to establish that values of the
independent variable have a significant
effect on the dependent variable. This
type of research usually involves group
comparisons. The groups in the study
make up the values of the independent
variable, for example gender (male versus
female). However, in causal-comparative
research the independent variable is not
under the researchers control but has to
take the values of the independent
variable as they come. The dependent
variable in a study is the outcome
variable. The statistic generally used to
analyze this type of research is an analysis
of covariance.
Effect-
Cause
outcome
(Independent
variable) (Dependent
variable)
Fig 1.2 Cause-effect
diagram

7
 It accomplishes this goal through
laboratory, field experiments, treat people
differently and see how they respond,
notice if there are differences based on
variables you cannot control.
For example: The effect of smoking
(independent variable) on lung cancer
(dependent variable); Effect of high
achievement in a on positive attitude;
Effect of latitude and altitude on
temperature at a place; Effect of amount
spent on advertisement and sales
generated.
1.2.1.5. Experimental
Experimental research or intervention
approach, the researcher manipulates a
situation and measures the effects of this
manipulation.
In an EXPERIMENTAL STUDY, individuals
are randomly allocated to at least two
groups. One group is subject to an
intervention, or experiment, while the
other group is not. The outcome of the
intervention (effect of the intervention on
the dependent variable/problem) is
obtained by comparing the two groups.

8
 Study
Population
(Sampling)

Sample Population
(randomization)

EXPERIMENTAL, or CONTROL, or
study group comparison group

First data collection First data collection

(before intervention) X (same time as in study group) A

Period of INTERVENTION/ NO INTERVENTION/

MANIPULATION MANIPULATION

Last data collection

Last data collection
(after intervention)
(same time as in study group)Z
Y

COMPARE
Treatment Effect = (Y – X) – (Z – A)

Fig 1.3 Before and After with Control

Design

9
For example, a researcher plans to study
the effect of a new drug. He plans to
include 300 patients in the study who are
currently receiving the standard treatment
for the same condition for which the new
drug has been designed. He explains the
study to the patients asking their consent
to be divided into two groups on a random
basis. One group will receive the
experimental drug while the other group
will continue to receive the standard
treatment. He makes sure that the
medications are disguised and labelled in
such a manner that neither the research
assistant administering them nor the
patient know which drug is used.
Another type of design that is often
chosen because it is quite easy to set up
uses only one group in which an
intervention is carried out. The situation is

10
analysed before and after the intervention
to test if there is any difference in the
observed problem. This is called a In
BEFORE-AFTER experimental study
involves neither randomization nor the use
of a control group.
The Hawthorne effect.
Students of business are aware of the
effects Elton Mayo found in his work at the
Hawthorne Works. When researchers
became interested in people they begin
acting differently and the changes can
continue even after the researchers have
gone – Hawthorne effect.
Study group Study group
before Intervention after

COMPARE

Fig 1.4 Diagram of BEFORE - AFTER

Study

11
In QUASI-EXPERIMENTAL experimental
study, one characteristic of a true
experiment is missing, either
randomization or the use of a separate
control group. A quasi-experimental study,
however, always includesStudy group the
manipulation
Study group
of Intervention
an independent after variable
which is the intervention. One of theCOMPARE most
common quasi-experimental designs uses
Control group
twoControl
(or more)
group groups, one of which
after serves
before
as a controlFiggroup in which no intervention
1.5 Quasi-experimental design with two groups
takes place. Both groups are observed
before as well as after the intervention, to
test if the intervention has made any
difference.

12
A researcher plans to study the effects of
health education on the level of
participation of a village population in an
immunization campaign. She decides to
select one village in which health
education sessions on immunization will
be given and another village which will not
receive health education and serves as a
control. The immunization campaign will
be carried out in the same manner in both
villages. A survey will then be undertaken
to determine if the immunization coverage
in the village where health education was
introduced before the campaign is
significantly different from the coverage in
the ‘control’ village which did not receive
health education. (Note: The study is
quasi-experimental because the subjects
were not assigned to the control or
experimental groups on a random basis).
Factorial Designs
Sometimes we may suspect more than
one independent variable to be

13
 responsible for changes in a dependent
variable. An experiment which seeks to
test a number of causal factors, singly and
in combination, is known as a factorial
design.
1.2.2. Qualitative Research Design
Qualitative research involve the
identification and exploration of a number
of often mutually related variables that
give Insight in human behaviour
(motivations, opinions, attitudes), in the
nature and causes of certain problems and
in the consequences of the problems for
those affected. Qualitative research
designs can be classified into Historical,
Ethnographic and Case study approaches.
Historical research is involved in the study
of past events to describe trends, factors,
causes, effects, that may explain present
of future events. In this approach the
researcher is concerned with analyzing
previously produced documents and, on
occasion, interviewing witnesses. The
researcher is called upon to formulate and
report conclusions and generalizations.
Some examples of historical research
studies can be: Factors leading to the

14
 development and growth of team learning;
Effects of decisions of the Supreme Court
on Private Technical Education, Trends in
reading instruction, 1970-1975.
1.2.2.1. Ethnographic research
Ethnography (from the Greek ethnos =
nation and graphein = writing) refers to
the qualitative description of human social
phenomena, based on fieldwork.
Ethnography is a holistic research method
founded in the idea that a system's
properties cannot necessarily be
accurately understood independently of
each other. Ethnographic research also
called as naturalistic method, endeavors
to answer “why” questions. It is involved
in the study of current events rather than
past events. It involves the collection of
extensive narrative data (non-numerical
data) on many variables over an extended
period of time in a naturalistic setting. The
researcher becomes a part of the
environment to be studied and interpreted
from the viewpoint of the subjects (what is
seen). Methods of data collection include
observations, interviews, questionnaires
etc. over a sustained period. An example

15
of a question studied through this
approach can be “Why do students
believe that science is a list of facts to be
memorized?”
1.2.2.1.1. Techniques of conducting
Ethnographic research
1. Direct, firsthand observation of
daily behavior. This can include
participant observation.
2. Conversation with different levels
of formality. This can involve small talk
to long interviews.
3. The genealogical method. This is a
set of procedures by which
ethnographers discover and record
connections of kinship, descent and
marriage using diagrams and symbols.
4. Detailed work with key consultants
about particular areas of community life.
5. In-depth interviewing.
6. Discovery of local beliefs and
perceptions.
7. Problem-oriented research.
8. Longitudinal research. This is
continuous long-term study of an area or
site.
9. Team research.

16
 Not all of these techniques are used by
ethnographers, with interviews and
participant observation being the most
widely used.
1. Focus group
A focus group is a form of qualitative
research in which a group of people are
asked about their attitude towards a
product, concept, advertisement, idea, or
packaging. Questions are asked in an
interactive group setting where
participants are free to talk with other
group members. In the world of
marketing, focus groups are an important
tool for acquiring feedback regarding new
products, as well as various topics.
In particular, focus groups allow
companies wishing to develop, package,
name, or test market a new product, to
discuss, view, and/or test the new product
before it is made available to the public.
This can provide invaluable information
about the potential market acceptance of
the product.
Also, in the social sciences, focus groups
allow interviewers to study people in a
more natural setting than a one-to-one

17
 interview. In combination with participant
observation, they can be used for gaining
access to various cultural and social
groups, selecting sites to study, sampling
of such sites, and raising unexpected
issues for exploration.
Focus groups have a high apparent
validity - since the idea is easy to
understand, the results are believable.
Also, they are low in cost, one can get
results relatively quickly, and they can
increase the sample size of a report by
talking with several folks at once.
However, focus groups also have
disadvantages: The researcher has less
control over a group than a one-on-one
interview, and thus time can be lost on
issues irrelevant to the topic; the data are
tough to analyze because the talking is in
reaction to the comments of other group
members; observers/ moderators need to
be highly trained, and groups are quite
variable and can be tough to get together.
2. Participant observation
Participant observation is a major research
strategy which aims to gain a close and
intimate familiarity with a given group of

18
individuals (such as a religious,
occupational, or deviant group) and their
practices through an intensive
involvement with people in their natural
environment.
Such research usually involves a range of
methods: informal interviews, direct
observation, participation in the life of the
group, collective discussions, analyses of
the personal documents produced within
the group, self-analysis, and life-histories.
Thus, although the method is generally
characterized as qualitative research, it
can (and often does) include quantitative
dimensions. Participant observation is
usually undertaken over an extended
period of time, ranging from several
months to many years. An extended
research time period means that the
researcher will be able to obtain more
detailed and accurate information about
the people he/she is studying. Observable
details (like daily time allotment) and
more hidden details (like taboo behaviour)
are more easily observed and
understandable over a longer period of
time.

19
It emerged as the principal approach to
ethnographic research by anthropologists
and relied on the cultivation of personal
relationships with local informants as a
way of learning about a culture, involving
both observing and participating in the
social life of a group. By living with the
cultures they studied, these researchers
were able to formulate first hand accounts
of their lives and gain novel insights.
This same method of study has also been
applied to groups within Western society,
and is especially successful in the study of
sub-cultures or groups sharing a strong
sense of identity, where only by taking
part might the observer truly get access to
the lives of those being studied.
1.2.2.2. Case study approach
Case study approach in qualitative
research includes an in-depth study of an
individual, group, institution, organization
or program. Data include interviews, field
notes of observations, archival data and
biographical data. Some of the examples
can be: A multi-case study of employees
who excel despite non facilitating
environments. A case study research by a

20
 teacher for improving student’s writing
skills.
1.2.2.3. Projective techniques
Qualitative research involves projective
techniques, when the respondents are
unwilling to give the information or unable
to provide it. The projective techniques
are used to reveal hidden feelings and
attitudes without being aware of doing so.
For e.g. immediately respondents are
asked to complete a sentence (sentence
completion tests), story completion tests,
respondents are shown cartoons in which
two persons are shown interacting with
one bubble filled and other blank and the
respondents are asked to fill the blank.
A Diagram classifying the types of
Is the data
Research
numerical orDesigns
Narrative ?
Numerical Narrative

Cause-Effect Past Events or

Relationship ? Current Events ?

Yes No Past Current

Independent
Relationship Historical Qualitative
Variable
Predicted ? Research ? Research/
Manipulated ?
Case-Study ?

No
Yes No
Yes

Experimental Causal- Co-relational Descriptive

Research Comparative Research Research
Research

21
1.3 Validity and Reliability of
Research
Validity means that the observations
actually measure what they intend to
measure (the conclusions are true).
Validity refers to the soundness of the
observations and to the accurateness of
the data collected by the research
method. Reliability (repeatability) means
that someone else using the same method
in the same circumstances should be able

22
to obtain the same findings (the findings
are repeatable).

Neither valid Reliable but

nor reliable not valid
The research The research
methods do not methods do not
hit the heart of the hit the heart of
research aim (not the research
‘valid’) and aim, but
repeated attempts repeated
are unfocussed. attempts get
almost the same
(but wrong
results.

Fairly valid but Valid and

not very reliable reliable
The research The research

23
 methods hit the methods hit the
aim of the study heart of the
fairly closely, but research aim,
repeated attempts and repeated
have very attempts all hit
scattered results in the heart
(not reliable) (similar results.
Fig 2.7 Graphical Presentation of Possible
Combinations of Validity and Reliability

QUESTIONS
Q 1. Distinguish between Exploratory,
Descriptive and Experimental Research
with the help of suitable illustrations.
MBA Part I, 2002, MBA Part II -
Three year Executive course, 2003.
Q 2. Discuss different research designs with
suitable examples. MBA Part I, 2004.
Q 3. Differentiate between descriptive and
experimental research designs.
MBA Part –I,
2007.
Q 4. A marketing manager of a local chocolate
manufacturing company likes to
determine the price for each chocolate in

24
comparision with national brands. He
thinks that the price lower than national
brands will be a right decision to increase
sales. He also believes that the chocolate
purchase also depends on customers'
income level. He wishes to test his
assumptionthrough an appropriate
research which will help him to arrive at
appropriate price decision. You are
required to develop an appropriate
research design. MBA
Part-I, 2007.

25
 Chapter 2
Sampling Decisions
2.1. Sampling
Sampling is the process of selecting a
number of study units from a defined
study population. Some studies involve
only small numbers of people and thus all
of them can be included. We might be
interested in the proportion of the Indian
population that will vote for various
political parties, or perhaps how the grain
size of sediment varies down the course of
a river. In both the above cases we wish to
know something about the population (all
the individuals), but the population is too
large to permit us to physically go and find
out the characteristics of each individual.
We therefore sample the population - we
select a manageable sub-set of the
population and measure the
characteristics of this sub-set (or sample).

26
We then use the sample characteristics as
a proxy for the population characteristics.
Often, research focuses on such a large
population that, for practical reasons, it is
only possible to include some of its
members in the study. Then we draw a
SAMPLE from the total population.
Sample provides an attractive way of
learning about a population or process in
terms of reduced cost and time.
Sampling methods

Random sampling Non-Random sampling

methods methods

Simple
Cluster Systemati Quota
Random Convenienc
Samplin c samplin
Samplin e sampling
g sampling g
g
Stratifie Judgementa
Multi-
d l or
stage
Samplin purposeful
sampling
g Sampling

Fig. 4.1 Sampling Methods

27
2.1.1. Random (Probability) Sampling
Simple Random Sampling: A simple
random sampling every member of the
population has an equal and independent
chance of being selected. It is ideal for
statistical analysis. In random sampling
every item is chosen in such a way that
each has equal chance of being selected.
For example, if N is the finite population
size, then NCn possible samples have the
same probability, 1/ NCn of being selected.
For e.g. six elements (N=6), suppose the
sample size is 3, then there are 6C3 distinct
samples, and 1/ 6C3 = 1/20, probability for
each sample of being chosen. Relatively
easy method of obtaining random samples
is by the use of Random Numbers. For
example if we want to take a sample of 10
units from a population of 5000 units
bearing numbers from 3001 to 8000, four-
digit random numbers can be used from
the tables from left to right. One
disadvantage with this method is that all
the members of the population have to be

28
 available for selection, which many times
is not possible. A simple random sample
can also be selected using a ‘lottery’
method, For example, a simple random
sample of 50 students is to be selected
from a school of 250 students. Using a list
of all 250 students, each student is given
a number (1 to 250), and these numbers
are written on small pieces of paper. All
the 250 papers are put in a box, after
which the box is shaken vigorously, to
ensure randomisation. Then, 50 papers
are taken out of the box, and the numbers
are recorded. The students belonging to
these numbers will constitute the simple
random sample.
2.1.1.1. Stratified Sampling
Stratified Sampling is useful when the
population consists of number of
heterogeneous subpopulations (groups)
and the members within a given
subpopulation are relatively homogenous
compared to the population as a whole. A
simple random sample, called as sub-
sample, is drawn from each group, in
proportion to its size. The basis for
forming the group such as location, age,

29
industry type, gross sales or number of
employees, is at the discretion of the
researcher. Individual group samples are
combined into one to obtain an overall
sample for analysis. For this method of
sampling to be more effective in terms of
reliability, efficiency, and precision, any
stratification should be done which
ensures maximum uniformity among
members of each group and largest
degree of variability among various
groups. For example, a survey is
conducted on household water supply in a
district comprising 20,000 households, of
which 20% are urban and 80% rural. It is
suspected that in urban areas the access
to safe water sources is much more
satisfactory. A decision is made to include
100 urban households (out of 4000, which
gives a 1 in 40 sample) and 200 rural
households (out of 16000, which gives a 1
in 80 sample). Because we know the
sampling fraction for both strata, the
access to safe water for all the district
households can be calculated after the
study (by multiplying the findings for the
urban households by 40 and those for the

30
rural households by 80, and then
calculating statistics for the total sample).
Cluster Sampling
Cluster Sampling is useful when the
population consists of very large number
of similar groups which are geographically
distant. The method sometimes known as
area sampling method, has been devised
to meet the problem of costs or
inadequate sampling frames (a complete
listing of all members in a population so
that each member can be identified by a
distinct number). By the use of map
references, the entire area to be analysed
is divided into smaller areas and a sample
of desired number of areas are selected
by simple random sampling method. Such
groups are termed as clusters. The
members of a cluster are called
elementary units. From each of such
cluster we may select a random sample of
the desired size. The value of cluster
sampling depends on how representative
each cluster is of the entire population. If
all clusters are similar in this regard, then
sampling a small number of clusters will
provide good estimates of the population

31
parameters. Cluster sampling is useful
when each cluster is a representative
small-scale version of the entire
population. In practice, it is advisable to
take a sample consisting of a large
number of small clusters rather than a
sample containing only few large clusters.
For example, in a study of the
knowledge, attitudes and practices (KAP)
related to buying preference in rural
communities of a region, a list is made of
all the villages. Using this list, a random
sample of villages is chosen and all study
units in the selected villages are
interviewed. The main advantage of
cluster sampling is that in this type of
sampling a sampling frame of individual
units is not required for the whole
population. Existing sampling frames of
clusters are sufficient. Only within the
clusters that are finally selected is there a
need to list and sample the individual
units. The cluster sample is easier to
select than a simple random sample of
similar size, because the individual units in
the sample are physically together in
groups, instead of scattered all over the

32
study population. The main
disadvantage of this type of sampling is
that compared to simple random
sampling, there is a larger probability that
the final sample will not be representative
of the total study population. The
likelihood of the sample not being
representative depends mainly on the
number of clusters that is selected in the
first stage. The larger the number of
clusters, the greater is the likelihood that
the sample will be representative. Further,
the sampling units at cluster level should
be selected randomly (avoid convenience
sampling).

Multi-stage sampling
Multistage sampling is useful when the
population is very widely spread and
random sampling is not possible. In such
case the whole population is divided into
number of primary units called stages,
each of which is composed of second
stage units. A series of samples are then

33
taken at successive stages. The sample
size at each stage is determined by the
relative population size of each stage. For
example, for the purpose of a national
pre-election opinion poll, the first stage
would be to draw the sample with
reference to main geographical areas. The
size of the sample that is the number of
interviews, from each region would be
determined by the relative populations in
each region. In the second stage, a limited
number of towns/cities in each of the
regions would be selected, and then in
the third stage, within the selected
towns/cities, a sample of respondents of
the region could be drawn from the
electoral roll of the areas selected in the
second stage.
Systematic Sampling
In Systematic sampling individuals are
chosen at regular intervals (for example
every fifth) from the sampling frame.
Ideally the first individual is randomly
selected to tell where to start selecting
individuals from the list.
For example, a systematic sample is to
be selected from 1200 students of a

34
school. The sample size selected is 100.
The sampling fraction is:
100 ( = sample size) 1
=
1200 ( = study population ) 12

The sampling interval is therefore 12.

The number of the first student to be
included in the sample is chosen
randomly, for example by blindly picking
one out of twelve pieces of paper,
numbered 1 to 12. If number 6 is picked,
then every twelfth student will be included
in the sample, starting with student
number 6, until 100 students are selected:
the numbers selected would be 6, 18, 30,
42, etc. Systematic sampling is usually
less time consuming and easier to perform
than simple random sampling. However,
there is a risk of bias, as the sampling
interval may coincide with a systematic
variation in the sampling frame. For
instance, if we want to select a random
sample of days on which to count college
attendance, systematic sampling with a
sampling interval of 7 days would be
inappropriate, as all study days would fall
on the same day of the week (e.g.,
Tuesdays only).

35
An example of systematic sampling,
taking every 10th door number of a street
(see figure 4.2)
Door numbers
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30
Fig. 4.2 An example of systematic
sampling

Non-Random Sampling
Quota sampling
Quota Sampling In this method the
selection of the respondents lies with the
investigator, although in making such
selection the researcher must ensure that
each respondent satisfies certain criteria
which are essential for the study. For
example, the researcher may choose to
interview ten men and ten women in such
a way that two of them have annual
incomes of more than two lakh rupees,
five of them have annual incomes
between one and two lakh rupees and
thirteen whose annul income is below one

36
lakh rupees. Furthermore some of them
should be between 25 and 35 years of
age, others between 36 and 45 years of
age, and the balance over 45 years. This
means that the researcher choice of
respondent is partially dedicated by these
‘controls’. Quota sampling has been
criticized because it does not satisfy the
fundamental requirement of a sample,
that is, it should be random.
Consequently, it is not possible to achieve
precision of results on a valid basis.
Snowball Sampling
In snowball sampling, an initial group of
respondents is selected, usually at
random. After being interviewed, these
respondents are asked to identify others
who belong to the target population of
interest. Subsequent respondents are
selected based on the referrals. This
process may be carried out in waves by
obtaining referrals from referrals, thus
leading to snowballing effect. Even though
probability sampling is used to select the
initial respondents, the final sample is a
non-probability sample. The referrals will
have demographic and psychographic

37
characteristics that are more similar to the
persons referring them that would occur
by chance.
A major objective of snowball sampling is
to estimate characteristics that are rare in
the population. Examples include users of
particular government or social services,
whose names cannot be revealed, special
census group such as widowed females
under 35; members of a scattered
minority population. The major advantage
of snowball sampling is that it
substantially increases the likelihood of
locating the desired characteristic in the
population. It also results in relatively low
sampling variance and costs.
Judgemental or Purposeful sampling
Judgemental/Purposive Sampling In
this procedure, the researcher selects
units of the sample that he/she feels are
most representative of the population with
respect to population characteristics under
study. For example, a teacher may
select two or three students from his
class, judging that these students would
reflect the general opinion of all students
in the class on certain issues. However,

38
 the validity of the sample results depends
on the proper judgment of the researcher
in choosing the sample. Great precaution
is needed in drawing conclusions based on
judgment samples to make inferences
about a population.
Convenience Sampling
In Convenience Sampling units are
included in the sample at the convenience
of the researcher rather than any
prespecified or known probabilities of
being selected. For example, a student for
his project on ‘food habits among adults’
may use his own friends in the college to
constitute a sample simply because they
are readily available and will participate
for little or no cost. Convenience samples
are easy for collecting data on a particular
issue. However, it is not possible to
evaluate its representativeness of the
population and hence precautions should
be taken in interpreting the results of
convenience samples that are used to
make inferences about a population.
2.2. Sample Size
Key Factors in determining Sample
size

39
 The sample size depends on a number of
things:
 Type of sample
The same level of precision can be
achieved with a smaller sample if
random sampling is used.
 Level of precision and confidence limits
required
The higher the precision levels and
confidence limits, the larger is the
sample.
 Variability of data
The more is the variation in data, the
larger is the sample.
 Time and costs constraints
Data collection is time-consuming and
costly.
 Size of working universe
The larger the working universe, the
larger the sample needs to be.
2.3. Sample Size determination
The sample size for estimating mean in a
population with standard deviation σ ,
with an error not greater than ‘ e ’ and
with confidence corresponding to z is
given by:
z 2σ 2
n= 2
e

40
 where, n = sample size
σ = standard deviation
e = acceptable error ( the
precision )
For finite Population
z 2 .N .σ p
2

n=
( N − 1).e 2 + z 2 .σ p
2

where,
N = Finite population size.
σ = standard deviation of
p

population.
z = standard variate at given
confidence level.
Sample size when estimating a
percentage or proportion
z 2 . p.q
n=
e2
where p = proportion of
success in the sample.
q = (1 – p)
For finite Population
z 2 . p.q.N
n=
( N − 1).e 2 + z 2 . p.q
If the standard deviation of the population
is not available, on the basis of some
estimation about range (difference
between highest and lowest values), a
crude estimate of the standard deviation

41
of the population for getting the working
idea of the required sample size is made
as follows:
Range
σˆ =
6
Example 1
The hotel management is interested in
determining the percentage of hotel’s
guest who stays for more than three days.
The reservation manager wants to be 95
% confident that the percentage has been
estimated to be within ± 2% of the true
value. What is the most conservative
sample size needed for this situation.
Solution
e = 0.02 ( since the estimate should
be within 2 % of the true value)
z = 1.96 (from normal probability
distribution table)
p = 0.5 ( for conservative sample)
z 2 . p.q
n=
e2
n=
(1.96 ) ×0.5 ×0.5
2

( 0.02 ) 2
n = 2401
Thus the sample size should be 2401.

42
Example 2
The axial length of pistons produced
follows normal distribution with the
variance of 64 mm. Determine the sample
size such that the mean length of pistons
is within plus or minus 2 mm with a
confidence level of 95 %.
Solution
The sample size for estimating mean in a
population with standard deviation σ ,
with an error not greater than ‘ e ’ and
with confidence to z is given by:

z 2σ 2
n=
e2
where n = sample size
σ = Variace = 64 = 8

Confidence level (1 −α) = 0.95 or α = 0.05

z = 1.96
e = the precision or the desired
accuracy = ± 2
1.96 2 × 8 2
n=
22
n = 61.5 ≈ 62
Thus the sample size should be 62.
Example 3

43
 The manager of a bank feels that 35 % of
branches will have increased yearly
collection of deposits after introducing a
hike in interest rate. Determine the
sample size such that the mean proportion
is within plus or minus 0.06 at the
confidence level of 90 %.
Solution
e = 0.06
z = 1.64 (from normal probability
distribution table)
p = 0.35, q = 0.65
z 2 . p.q
n=
e2
n=
(1.64 ) × 0.35 × 0.65
2

( 0.06 ) 2
n = 170
Thus the sample size should be 170.
2.4. Sample Statistics
In most statistical research studies,
population parameters are usually
unknown and have to be estimated from a
sample. Sampling means drawing a
sample from the population for estimating
population parameters from sample
statistics. The estimate of the population
parameter may be one single value (point
estimate) or it could be range of values
(interval estimate).

44
Sample statistics
Mean = x and standard deviation s

Population statistics
Mean = µ and standard deviation σ

The sample may be unbiased but it cannot

eliminate sampling variability or sampling
error.
Due to variations in samples, standard
error of the mean (σ ) (standard deviation
x

of sampling distribution) is calculated as

follows:
σ
σx =
n

where,
σ = Standard error of the
x

mean.
n = sample size.
Standard error of the mean for the Finite
population
s N −n
σx = ×
n N −1

where,
N = Population size.
s = Sample standard
deviation.
In estimation of population parameters,
there are two types of estimations.
Point Estimate (for single value)

45
 Population mean, µ = Sample
mean, x
Population Standard deviation, σ =
Sample standard deviation, s
Interval estimate (for range of values)
For Mean µ of the population
σ
x ±zσ or x ± z n
x

Where,
z corresponds to confidence level, (1 −α) ,
corresponding to the area under the
normal curve. The most widely used
values of z are:
Confidence Interval z
90 % 1.645
95 % 1.96
99 % 2.58
Confidence level is the likelihood that the
answer will fall under the precision range.
For example for 95 % of the area

z = ±1.96

where, σ is not known,

46
 s
σ x = sx =
n

where,
s is the sample standard
deviation
∑( x − x )
2

s=
n −1

t - Distribution
t – distribution is used when σ is not
known and the sample size < 30.
Confidence Interval for estimating
population mean.
x ±t df s x

where,
t
df is the t statistic at n – 1 degrees of
freedom (df) at at required level of
significance (α) .
Confidence Interval of Population
proportion
 p.q 
p ± z



 n 

where,
 p.q 


 n



is the standard error of
proportion
p = Sample proportion of
success.
q = (1 − p )
n = size of the sample.
Example 4

47
If the size of the sample is 36, sample
mean is 6.20 and the standard deviation is
3.8, find the point estimation of population
mean and standard deviation. Also find
the interval estimate of mean of the
population for 95% confidence level?
Solution
Given
n = 36, x = 6.20, s = 3.8
Point Estimation
Population mean, µ = Sample
mean, x
i.e. µ = x = 6.20
Population Standard deviation, σ =
Sample standard deviation, s
i.e. σ = s = 3.8
Interval estimate
For Mean µ of the population
σ
x ±zσ OR
x
x ±z
n

For 95 % Confidence Interval,

z =1.96 (from normal curve)
 3.8 
6.20 ±1.96  
 36 

or
  3.8    3.8 
6.20 −1.96   < µ < 6.20 + 1.96  
  36    36 

or 4.96 < µ < 7.44

48
 µ = 4.96 µ

= 7.44
Example 5
In a random selection of 64 of the 2400
intersections in a small city, the mean
number of scooter accidents per year was
3.2 and the sample standard deviation
was 0.8. Make an estimate of the standard
deviation of the population from the
sample standard deviation. Work out the
standard error of the mean for this finite
population. If the desired confidence level
is 90 %, what will be upper and the lower
limits of the confidence interval for the
mean number of accidents per
intersection per year?
Solution
Given
N = 2400
n =64
x =3.2
z =1.645 for 90 % Confidence
s =0.8

Interval.
σ = s = 0.8

49
For Mean µ of the population, upper and
the lower limits of the confidence interval.
x ±zσx
s N −n
where, σx =
n
×
N −1 = 0.097
3.2 ±1.645 (0.097 )

or 3.0404 < µ < 3.3596

Example 6
A Market research survey in which 64
consumers were contacted states that 64
% of all consumers of a certain product
were motivated by the product’s
advertising. Find the confidence limits for
the proportion of consumers motivated by
advertising in the population, given the
confidence level to 95 %.
Solution
Given
p =0.64
q =(1 − p ) = 0.36
n =64
z =1.96

Limits of Population proportion

0.64 ×0.36
0.64 ±1.96
64
0.64 ±0.1176
52 .24 %to 75 .76 %

Example 7
From past experience it has been found
that the standard deviation for the days of
absence per worker per year in a

50
particular factory is 2 days. Construct 95
% and 99 % confidence intervals for the
population mean number of absences
given that a random sample of 87 workers
has a mean of 6 days absence.
Solution
Standard deviation σ = 2 days
Sample size n = 87 workers
Population Mean µ = 6 days absence
 σ 
Confidence Interval x ± z n 
For 95 % Confidence Interval z = 1.96
 2 
6 ±1.96  
 87 
6 ±1.96 ×0.2144
6 ± 0.42
5.58 to 6.42 days of
absence
For 99 % Confidence Interval z = 2.58
 2 
6 ± 2.58 
 

 87 
6 ± 2.58 × 0.2144
6 ± 0.55
5.45 to 6.55 days of
absence
Example 8
Twelve bank tellers were randomly
sampled and it was determined they made
an average of 3.6 errors per day with a
sample standard deviation of 0.42 error.

51
Construct a 90% confidence interval for
the population mean of errors per day.
Solution
Here as the sample size is < 30, and the
population standard deviation is not
known, so t – distribution will be used.
Given:
n = 12
x =3.6
s =0.42

Degrees of freedom = n – 1
= 12 – 1 = 11
t( ) =1.796 at 90 %
11 , 0.05

Confidence Interval
x ±t df s x

90 % Limits
 0.42 
3.6 ±1.796 × 
 12 

3.6 ± 0.218
3.382 to 3.818

Example 9
The foreman of a mining company has
estimated the average quantity of iron-ore
extracted to be 36.8 tons per shift and the

52
sample standard deviation to be 2.8 tons
per shift, based on the random selection
of 4 shifts. Construct a 90 % confidence
interval around this estimate.
Solution
Here as the sample size is < 30, and the
population standard deviation is not
known, so t – distribution will be used.
Given:
x =36 .8
s =2.8

Degrees of freedom = n – 1
=4–1=3
t
df = t( ) = 2.353 ( as 5 % or 0.05 area will go
3, 0.05

in one tails )
Confidence Interval
x ±t df s x
 2.8 
36 .8 ± 2.353 × 
 4

33.5058 to 40.0942
Example 10
A sample of 12 measurements of the
breaking strength of cotton threads gave
the mean of 73.8 grams and a standard
deviation of 12.4 grams. Find the 95 %
and 99 % confidence limits for the actual
breaking strength.
Solution

53
Here as the sample size is < 30, and the
population standard deviation is not
known, so t – distribution will be used.
Given:
x =73 .8
s =12 .4

Degrees of freedom = n – 1
= 12 – 1 = 11
t( ) = 2.201 at 95 %
11 , 0.025

t( ) = 3.106
11 , 0.005 at 99 %
Confidence Interval
x ±t df s x

95 % Limits
12 .4 
73 .8 ± 2.201 × 
 12 

73.8 ± 7.9
65.9 to 81.7
99 % Limits
12 .4 
73 .8 ± 3.106 ×  
 12 

73.8 ± 11.1
62.7 to 84.9
QUESTIONS

Q 1. What are different techniques of

sampling? Explain with examples.
MBA Part -1 2006
Q 2. Differentiate between snowballing,
stratified and quota sampling

54
methods.
MBA Part -1 2007

55
 Chapter 3

Software used in Research

Analysis

3.1. Introduction
Computers have become one of the most
useful research tools in modern times.
They are suited for data analysis in large
research projects, concerned with huge
storage of data, their faster retrieval when
required, and faster processing of data,
with the help of various application
softwares. Their use has reduced human
effort and added to the quality of research
activity.
The two most commonly used statistical
softwares used in analysis are described in
below sections:
 Microsoft Excel (of Microfoft Inc.,
USA)
 SPSS (from SPSS Inc. USA)

56
 Other software that are also used
worldwide are:
 Minitab
 SAS
3.2. Microsoft Excel - Statistical
analysis tools
Microsoft Excel provides a set of data
analysis tools — called the Analysis
ToolPak — that can be used to save steps
and assist in developing complex
statistical or engineering analyses. Some
of the statistical functions are built-in and
others become available when you install
the Analysis ToolPak. Accessing the data
analysis tools The Analysis ToolPak
includes the tools described below. To
access these tools, click Data Analysis on
the Tools menu. If the Data Analysis
command is not available, one have to
load the Analysis ToolPak add-in program.
Some of data analysis tools in Excel:
 Anova
 Correlation
 Covariance
 Descriptive Statistics
 Exponential Smoothing
 F-Test Two-Sample for Variances

57
  Histogram
 Moving Average
 Random Number Generation
 Rank and Percentile
 Regression
 Sampling
 t-Test
 z-Test
These are described below:
3.3. Microsoft Excel - Statistical analysis
tools
3.3.1. Anova
The Anova analysis tools provide different
types of variance analysis. The tool to use
depends on the number of factors and the
number of samples you have from the
populations you want to test.
Anova: Single Factor -This tool performs a
simple analysis of variance, testing the
hypothesis that means from two or more
samples are equal (drawn from
populations with the same mean). This
technique expands on the tests for two
means, such as the t-test.
Anova: Two-Factor with Replication This
analysis tool performs an extension of the

58
single-factor anova that includes more
than one sample for each group of data.
Anova: Two-Factor without Replication
This analysis tool performs a two-factor
anova that does not include more than
one sampling per group, testing the
hypothesis that means from two or more
samples are equal (drawn from
populations with the same mean). This
technique expands on tests for two
means, such as the t-test.
3.3.2. Correlation
The Correlation analysis tool measures the
relationship between two data sets that
are scaled to be independent of the unit of
measurement. The population correlation
calculation returns the covariance of two
data sets divided by the product of their
standard deviations based on the
following formulas.
You can use the correlation analysis tool
to determine whether two ranges of data
move together — that is, whether large
values of one set are associated with large
values of the other (positive correlation),
whether small values of one set are
associated with large values of the other

59
(negative correlation), or whether values
in both sets are unrelated (correlation
near zero).
3.3.3. Covariance
Covariance is a measure of the
relationship between two ranges of
data. The Covariance analysis tool returns
the average of the product of deviations of
data points from their respective means,
based on the following formula.
You can use the covariance tool to
determine whether two ranges of data
move together — that is, whether large
values of one set are associated with large
values of the other (positive covariance),
whether small values of one set are
associated with large values of the other
(negative covariance), or whether values
in both sets are unrelated (covariance
near zero).
3.3.4. Descriptive Statistics
The Descriptive Statistics analysis tool
generates a report of univariate statistics
for data in the input range, providing
information about the central tendency
and variability of your data.
3.3.5. Exponential Smoothing

60
The Exponential Smoothing analysis tool
predicts a value based on the forecast for
the prior period, adjusted for the error in
that prior forecast. The tool uses the
smoothing constant a, the magnitude of
which determines how strongly forecasts
respond to errors in the prior forecast.
Values of 0.2 to 0.3 are reasonable
smoothing constants. These values
indicate that the current forecast should
be adjusted 20 to 30 percent for error in
the prior forecast. Larger constants yield a
faster response but can produce erratic
projections. Smaller constants can result
in long lags for forecast values.
3.3.6. F-Test Two-Sample for Variances
The F-Test Two-Sample for Variances
analysis tool performs a two-sample F-test
to compare two population variances.
For example, you can use an F-test to
determine whether the time scores in a
swimming meet have a difference in
variance for samples from two teams.
3.3.7. Histogram
The Histogram analysis tool calculates
individual and cumulative frequencies for
a cell range of data and data bins. This

61
tool generates data for the number of
occurrences of a value in a data set.
For example, in a class of 20 students, you
could determine the distribution of scores
in letter-grade categories. A histogram
table presents the letter-grade boundaries
and the number of scores between the
lowest bound and the current bound. The
single most-frequent score is the mode of
the data.
3.3.8. Moving Average
The Moving Average analysis tool projects
values in the forecast period, based on the
average value of the variable over a
specific number of preceding periods. A
moving average provides trend
information that a simple average of all
historical data would mask. Use this tool
to forecast sales, inventory, or other
trends. Each forecast value is based on
the following formula.
where:
 N is the number of prior periods to
include in the moving average
 Aj is the actual value at time j
 Fj is the forecasted value at time j
3.3.9. Random Number Generation

62
The Random Number Generation analysis
tool fills a range with independent random
numbers drawn from one of several
distributions. You can characterize
subjects in a population with a probability
distribution.
For example, you might use a normal
distribution to characterize the population
of individuals' heights, or you might use a
Bernoulli distribution of two possible
outcomes to characterize the population
of coin-flip results.
3.3.10. Rank and Percentile
The Rank and Percentile analysis tool
produces a table that contains the ordinal
and percentage rank of each value in a
data set. You can analyze the relative
standing of values in a data set.

Regression
3.3.11.
The Regression analysis tool performs
linear regression analysis by using the
"least squares" method to fit a line
through a set of observations. You can
analyze how a single dependent variable

63
is affected by the values of one or more
independent variables.
For example, you can analyze how an
athlete's performance is affected by such
factors as age, height, and weight. You
can apportion shares in the performance
measure to each of these three factors,
based on a set of performance data, and
then use the results to predict the
performance of a new, untested athlete.
3.3.12. Sampling
The Sampling analysis tool creates a
sample from a population by treating the
input range as a population. When the
population is too large to process or chart,
you can use a representative sample. You
can also create a sample that contains
only values from a particular part of a
cycle if you believe that the input data is
periodic.
For example, if the input range contains
quarterly sales figures, sampling with a
periodic rate of four places values from
the same quarter in the output range.
3.3.13. t-Test
The t-Test analysis tools test the means of
different types of populations.

64
t-Test: Two-Sample Assuming Equal
Variances This analysis tool performs a
two-sample student's t-test. This t-test
form assumes that the means of both data
sets are equal; it is referred to as a
homoscedastic t-test. You can use t-tests
to determine whether two sample means
are equal.
t-Test: Two-Sample Assuming Unequal
Variances This analysis tool performs a
two-sample student's t-test. This t-test
form assumes that the variances of both
ranges of data are unequal; it is referred
to as a heteroscedastic t-test. You can use
a t-test to determine whether two sample
means are equal. Use this test when the
groups under study are distinct. Use a
paired test when there is one group before
and after a treatment.
t-Test: Paired Two Sample For Means
This analysis tool and its formula perform
a paired two-sample student's t-test to
determine whether a sample's means are
distinct. This t-test form does not assume
that the variances of both populations are
equal. You can use a paired test when
there is a natural pairing of observations

65
 in the samples, such as when a sample
group is tested twice — before and after
an experiment.
3.3.14. z-Test
The z-Test: Two Sample for Means analysis
tool performs a two-sample z-test for
means with known variances. This tool is
used to test hypotheses about the
difference between two population means.
For example, you can use this test to
determine differences between the
performances of two car models.
3.4. SPSS
SPSS is a computer program used for
statistical analysis and is also the name of
the company (SPSS Inc) that sells it.
SPSS (originally, Statistical Package for the
Social Sciences) was released in its first
version in 1968, and is among the most
widely used programs for statistical
analysis in social science. It is used by
market researchers, health researchers,
survey companies, government, education
researchers, and others. In addition to
statistical analysis, data management and
data documentation are features of the
base software.

66
The letters "SPSS" mean "Statistical
Package for the Social Sciences" is a
worldwide software for data analysis and
research, developed by the company SPSS
Inc., headquarters at Chicago, Illinois USA.
SPSS has made difficult analytical tasks
easier through advances in its usability
and data access, enabling more people to
benefit from the use of quantitative
techniques in making decisions.
SPSS is a software system based on the
idea of using statistics to turn raw data
into information essential to decision-
making. The SPSS analysis is the driving
force behind sound decision-making—the
DNA of intelligence. It was originally
developed in 1968 by Stanford university
doctoral candidates when there was a
need of quickly analyzing volumes of
social science data gathered through
various methods of research.
SPSS customers are in more than 100
countries around the world, using SPSS
analytics, the top 10 largest market
research firms use SPSS software, more
than 90% of the top U.S. universities use
SPSS software, more than 85% of the top

67
consumer packaged goods companies use
SPSS analytics, the top 10 pharmaceutical
companies in the world use SPSS
analytics, the largest newspaper
companies in the U.S. use SPSS software.
The Statistics included in the base
software:
 Descriptive statistics: Cross tabulation,
Frequencies, Descriptives, Explore,
Descriptive Ratio Statistics
 Bivariate statistics: Means, t-test,
ANOVA, Correlation (bivariate, partial,
distances), Nonparametric tests
 Prediction for numerical outcomes:
Linear regression
 Prediction for identifying groups:
Factor analysis, cluster analysis,
Discriminant Analysis.
Add-on modules provide additional
capabilities.
The available modules are:
 SPSS Programmability Extension.
Allows Python programming control of
SPSS.
 SPSS Data Validation. Allows
programming of logical checks and
reporting of suspicious values.

68
  SPSS Regression Models - Logistic
regression, ordinal regression,
multinomial logistic regression, and
mixed models.
 SPSS Advanced Models - Multivariate
GLM and repeated measures ANOVA.
 SPSS Classification Trees. Creates
classification and decision trees for
identifying groups and predicting
behaviour.
 SPSS Tables. Allows user-defined
control of output for reports.
 SPSS Exact Tests. Allows statistical
testing on small samples.
 SPSS Categories
 SPSS Trends™

 SPSS Conjoint
 SPSS Missing Value Analysis. Simple
regression-based imputation.
 SPSS Map
 SPSS Complex Samples. Adjusts for
stratification and clustering and other
sample selection biases.

3.4.1. Some features of SPSS 15.0

69
3.4.1.1. Complex Samples Plan
To help through the planning stage in the
analytical process, SPSS Complex Samples
provides with specialized tools and
procedures for working with sample
survey data. SPSS Complex Samples Plan
(CSPLAN) procedure can be used to
specify the sampling frame to create a
complex sample design or analysis.
CSPLAN can be used to specify how to
draw or analyze stratified, clustered or
multistage complex sample designs, with
or without replacement.
3.4.1.2. Sample Power
SPSS SamplePower saves time, eneggy
and resources by finding the right sample
size for the research in minutes and
testing the possible results before
beginning the study. If the sample size is
too small important research findings are
missed and if it's too large, valuable time
and resources are wasted.
SamplePower helps in finding the best
sample size for your research and
different study parameters and their
effects can be compared with the flexible
analytical tools. SamplePower is designed

70
 to cover Means and differences in means,
Proportions and differences in proportions,
Correlation, Oneway and factorial Analysis
of Variance (ANOVA), Analysis of
Covariance (ANCOVA), Regression,
Survival analysis, Equivalence tests.
3.4.1.3. SPSS Conjoint
SPSS Conjoint helps in discovering in a
realistic way what peaple value and how
individual product attributes affect
consumer and citizen preferences. With
SPSS Conjoint, the tradeoff effect of each
product attribute can be measured in the
context of a set of product attributes as
consumers value when making purchasing
decisions.
3.4.1.4. SPSS Data Entry
SPSS Data Entry products work together to
enable a fluid survey research process.
From creating effective surveys and
forms, to collecting accurate data for
analysis, the SPSS Data Entry system
covers all of the survey research needs.
All SPSS Data Entry products offer
complete integration with SPSS® for
Windows®, to move from data collection
to analysis in a single step.

71
3.4.1.5. SPSS Dimensions
Through SPSS Dimensions survey research
technology platform, a clear, detailed
understanding of what the customers
think, do, and what they will do in the
future can be gained. And you can
continually update this information, easily
and cost effectively. Using Dimensions
survey research software products
continuously customer satisfaction and
feedback can be tracked and be proactive
in retaining your best customers and
increasing their lifetime value, develop
and deliver market-driven products by
conducting ongoing surveys of target
customers, Optimize marketing efforts,
targeting the right customer segments
with appropriate offers delivered through
the right channel, Strengthen the brand
across geographies and markets, Support
all customer relationship management
(CRM) initiatives.
Dimensions can Collect data in any
language, for any format, Field paper
surveys that can be scanned to capture
responses, Conduct computer-assisted
personal interviewing (CAPI) using laptops

72
 or handheld devices, computer-assisted
telephone interviewing (CATI), including
text messaging on mobile phones, or
online surveys , Plug into your existing
customer relationship management
systems, to enrich your customer profile
database, Provide real-time results to
colleagues and clients through a secure
online portal.
3.4.1.6. SPSS Missing Value Analysis
Missing data can seriously affect the
results. If missing data is ignored or
assumed there is a risk of reaching invalid
and insignificant results. SPSS Missing
Value Analysis part of your data
management and preparation step
ensures that one enters the data analysis
stage using data that takes missing values
into account. SPSS Missing Value Analysis,
an SPSS add-on module, is a critical tool
for data validty, survey researchers, social
scientists, data miners, and market
researchers.
SPSS Missing Value Analysis can easily
examine data from several different
angles using one of six diagnostic reports
to uncover missing data patterns. It can

73
 then estimate summary statistics and
impute missing values through regression
or expectation maximization algorithms
(EM algorithms).
SPSS Missing Value Analysis can be used
to improve survey questions that have
been identified as possibly troublesome
or confusing, based on observed missing
data patterns.
3.4.1.7. SPSS Text Analysis for Surveys
Open-ended questions are included in the
surveys because there’s great value in
letting people express their views in their
own words. But it can be tedious, time-
consuming, and costly to categorize or
"code" text responses.
SPSS Text Analysis for Surveys to help in
gaining full value from text responses
without the drudgery and expense
associated with manual coding. SPSS Text
Analysis for Surveys is based on natural
language processing (NLP) software
technologies. SPSS Text Analysis for
Surveys offers both automated and
manual techniques, giving greater control
over the process and your results.
3.4.1.8. SPSS Regression Models

74
 SPSS Regression Models can be used in
many disciplines, including Market
research: Study consumer buying habits,
Medical research: Study response to
dosages through probit analysis,
Institutional research: Measure academic
achievement tests, Loan assessment:
Analyze good and bad credit risks.
SPSS Regression Models includes these
procedures:
 Multinomial logistic regression
(MLR): Predict categorical outcomes with
more than two categories
 Binary logistic regression: Easily
classify your data into two groups
 Nonlinear regression (NLR) and
constrained nonlinear regression (CNLR):
Estimate parameters of nonlinear
models
 Probit analysis: Evaluate the value
of stimuli using a logit or probit
transformation of the proportion
responding
3.4.1.9. SPSS Trends
Reliable forecasts can have a major
impact on your organization’s ability to
develop and implement successful

75
 strategies. SPSS Trends, an add-on
module to SPSS Base, can be used to
predict trends and develop forecasts
quickly and easily.
Unlike spreadsheet programs, SPSS
Trends has the advanced statistical
techniques needed while working with
time-series data. Regardless of level of
experience through SPSS Trends historical
data can be analysed and trends can be
pridicted faster, and deliver information in
ways that your organization’s decision
makers can understand and use.
3.4.1.10. SPSS Tables
SPSS Tables, an add-on module for SPSS
for Windows, enables to better understand
the data, and easily report the results to
those who need them.
More than a simple reporting program,
SPSS Tables provides comprehensive
analysis capabilities that help you learn
from your data. For example:
 Compare means or proportions for
demographic groups, customer
segments, time periods, or other
categorical variables, when you include
inferential statistics

76
  Select summary statistics—from
simple counts for categorical variables
to measures of dispersion—and sort
categories by any summary statistic
used
 Choose from three significance
tests: Chi-square test of independence,
comparison of column means (t test), or
comparison of column proportions (z
test)
3.4.1.11. SPSS Maps
Where are your best prospects located?
What are customers buying and where are
they buying it? How do you find the best
site for your next store or office? Which
neighborhoods, cities, and states or
provinces show high employment rates?
SPSS Maps uncovers answers to questions
like these for improved decision making.
The demographic data can be turned into
easy-to-read visual output with SPSS
Maps. SPSS Maps empowers to create
easy-to-read output using your geographic
data.
QUESTIONS
Q 1. “What is SPSS? What are its main
features? MBA Part -1 2006

77
Q 2. Enumerate different computer
softwares used for the analysis of data
with suitable illustrations.
MBA Part -1 2004

78
Most frequently
Non parametric tests used non-
parametric tests
One sample tests Two sample tests K-samples test

Chi-square Sign
Run test Median test Kruskal-
test test
Wallis test

79
Median Two Rank sum
test sample test
sign test
 Fig. 14.1 Classification of Non-parametric
tests

Some of the most Commonly used Tests

are explaned below
3.5. The sign test for paired data
It assumes population as continous
symmetrical and non-normal, such that
the probabilty of having a sample value
less than mean value, as well as
probability of having a sample value more
than the mean value is ½. Hence:

80
modified sample value = +, if sample
value X > µ
= – , if X < µ
This means that if a sample value is more
than µ (mean/median), then it is replaced
by ‘+’ sign, and if a sample value is less
than µ (mean/median), then it is replaced
by ‘–’ sign.
The Null hypothesis is:
1
H0 : p =
2
Alternative hypothesis can be: (one tail
test)
1
2 (right tail test)
H :p>
1

1
Or H : p < 2 (left tail test)
1

Example 1
A medical representative visited at
random 9 clinics and found that the
waiting times in minutes before meeting
doctors in these clinics are 16, 30, 35, 22,
17,11, 19, 10, 8. Use the sign test at 5%
level of significance to test the medical
representative claim that on an average
he has to wait more than 15 minutes for
the doctor.
Solution
The Null hypothesis is:

81
 1
H0 : p =
2
Alternative hypothesis: (one tail test)
1
2 (right tail test)
H :p>
1

Here the sample size is n = 9 (small

sample n<30). On classifying each
observation into ‘+’ sign if it is more than
15, and ‘–’ sign, if it is less than 15, the
signs are as follows: +,+,+,+,+,–,+,–,–.
The number of plus (+) signs = 6. The
number of minus (–) signs = 3.
The binomial probability that the number
of plus signs more than or equal to 6 is
computed below:
P ( X ≥ 6, n = 9, p = 0.5) =
= P ( X = 6) + P ( X = 7) + P( X = 8) + P ( X = 9)
6 3 7 2 8 1 9 0
1  1  1  1  1  1  1  1 
=9C6     +9C7     +9C8     +9C9    
2 2 2 2 2 2 2 2
9
1   9! 9! 9! 9! 
=   + + + 
 2  6!×( 9 −6 )! 7!×( 9 −7 )! 8!×( 9 −8)! 9!×(9 −9 )!
9
1   9! 9! 9! 9! 
=   + + + 
 2  6!×3! 7!×2! 8!×1! 9!×0!
9
 1  9 ×8 × 7 9 ×8 9 1
=   + + + 
 2   3 × 2 ×1 2 ×1 1 1
9
1 
=   [84 + 36 + 9 +1]
2
= 0.001953125 ×130
= 0.2539
Since P( X ≥ 6) is more than the given
significance level 0.05, the binomial
statistics fall in the acceptance region.
Hence, accept Ho.

82
 Thus the waiting time of the medical
representative is not more than 15
minutes.
In case of two tail test, the calculated
probability is compared with ½ of the
significance level.
When the sample size is greater than 30,
sample happen to be large, the binomial
distribution can be approximated to
normal distribution, with mean, µ = np ;
standard deviation, σ = npq
X −np
∴Z =
npq where q = (1 – p);
X = number of + signs.
3.6. The one sample Runs test
In Run test for randomness,
The null and the alternative hypothesis in
this test are as follows:
H : The occurrence of the runs in the
0

given stream of symbols is random.

H : The occurrence of the runs in the
1

given stream of symbols is not

random.
In this situation, one can approximate the
sampling distribution of r to normal
distribution with the following mean and
variance:

83
  2.n .n 
Mean, µr =  1 2 +1
 n1 + n2 
2.n1.n2 .( 2.n1.n2 − n1 − n2 )
Standard deviation, σr =
(n1 + n2 ) 2 .( n1 + n2 −1)

where,
1n = frequency of occurrence of a

particular symbol in the whole

stream of symbols.
2n = frequency of occurrence of

another symbol in the whole

stream of symbols.
r = the number of runs.
The formula for standard normal z statistic
to test the significance of r is given by:
r − µr
∴Z =
σr
The run test can be best explained by an
example shown below:
Example 2
An economic researcher wants to find out
if there is any pattern in arrivals at the
entrance of the shopping mall in terms of
male and females arriving or whether such
arrivals are simply random. One day, he
stationed himself at the main entrance
and recorded the gender of first 30
shoppers who came in. The results are as
follows: -

84
 MMFMFFFMMMFFMFMMFFFFMMMMM
FFMMM
Use the runs test for randomness at 0.05
level of significance. What are the
limitations of use of non-parametric test?
MBA Part –I, 2005
Solution
Number of runs using underlines to
combine the letters which constitute the
runs
MM F M FFF MMM FF M F MM FFFF MMMMM FF MMM
1 2 3 4 5 6 7 8 9 10 11 12 13

No. of runs ( r) = 1
n = number of occurrences of M =
1

17
n = number of occurrences of F =
2

13
The sampling distribution of ‘ r – statistic ’,
the number of runs, is to be used and this
distribution has its mean ( µ ) given by: r

 2.n .n   2 ×17 ×13 

µr =  1 2 +1
 n1 + n2 
=> µr = 
 30
+1

=> µr = 15 .733

and the standard deviation (σ ) given by: r

2.n1.n2 .( 2.n1.n2 − n1 − n2 )
σr =
(n1 + n2 ) 2 .( n1 + n2 −1)

In this case:
2 ×17 ×13 × ( 2 ×17 ×13 −17 −13 )
σr =
(17 +13 ) 2 .( 30 −1) => σ r = 2.6414

85
For testing the null hypothesis concerning
the randomness of the pattern of arrivals
at the entrance of the shopping mall in
terms of males and females at 5 % level of
significance, we assume that the sampling
distribution of r can be closely
approximated by the normal distribution.

Acceptance
2.5 % area Region
(95 % area) 2.5 % area

- 1.96 1.96

10.556 15.773 20.91

Limits of r : 10.556 – 20.91

By using the table of area under the

normal curve, the value of Z for 0.475 of
the area under the curve is 1.96. The
limits of acceptance region:
Upper limit = µ + ( 1.96 ) × ( σ )
r r

= 15.773 + 1.96 × 2.6414 = 20.91

Lower limit = µ ( 1.96 ) × ( σ )
r r

= 15.773 1.96 × 2.6414 =

10.556
We now find that the observed number of
runs ( i.e. r = 13 ) lies within the
acceptance region and therefore we
accept the null hypothesis of randomness

86
 of the pattern of arrivals at the entrance of
the shopping mall in terms of males and
females at 5 % level of significance (α = 0.05 ) .

Chapter 4
Attitude Measurement

4.1. Introduction
Attitude refers to likes and dislikes
towards products, services, advertisement
and media. There are many factors which
affect our attitude towards products like
moisturising effect for Dove, longetivity
and durability for Bajaj, economy for T-
Series cassettes.
Change in attitude can be due unforseen
circumstances like for e.g. a teenager
thinking of pruchasing a ‘TVS Scooty’ and
immediately father looses the job so he
may postpone it, time gap as attitudes are
dynamic, conflict of attitude like for e.g.
compromise for a particular ice-cream
flavor for spouse.
4.2. Selected Attitude Measurement
Scales

87
4.2.1. Guttman scale or Scalogram Analysis
This is a procedure to determine whether
a set of items can be rank-ordered on an
unidimensional scale. It utilizes the
intensity structure among several
indicators of a given variable. Statements
are listed in order of importance. The
rating is scaled by summing all responses
until the first negative response in the list.
The Guttman scale is related to Rasch
measurement; specifically, Rasch models
bring the Guttman approach within a
probabilistic framework.
A Guttman scale is a psychological
instrument developed using the scaling
technique developed by Louis Guttman in
1944 called Guttman scaling or scalogram
analysis. A primary purpose of the
Guttman scaling is to ensure that the
instrument measures only a single trait (a
property called unidimensionality, a single
dimension underlies responses to the
scale). Guttman's insight was that for
unidimensional scales, those who agree
with a more extreme test item will also
agree with all less extreme items that
preceded it.

88
A perfect Guttman scale
A hypothetical, perfect Guttman scale
consists of a unidimensional set of items
that are ranked in order of difficulty from
least extreme to most extreme position.
For example, a person scoring a "7" on a
ten item Guttman scale, will agree with
items 1-7 and disagree with items 8,9,10.
An important property of Guttman's model
is that a person's entire set of responses
to all items can be predicted from their
cumulative score because the model is
deterministic.
Applications
The Guttman scale is used mostly when
researchers want to design short
questionnaires with good discriminating
ability. The Guttman model works best for
constructs that are hierarchical and highly
structured such as social distance,
organizational hierarchies, and
evolutionary stages.
Example
Here is an example of a Guttman scale -
the Bogardus Social Distance Scale:
(Least favourable)

89
 1. Are you
willing to permit immigrants to live in your
country?
2. Are you
willing to permit immigrants to live in your
community?
3. Are you
willing to permit immigrants to live in your
neighbourhood?
4. Are you
willing to permit immigrants to live next
door to you?
5. Would you
permit your child to marry an immigrant?
(Most favourable)
E.g., agreement with item 3 implies
agreement with items 1 and 2.
Guttman’s scalogram analysis or
Cumulative scale is a scale for measuring
attitude, in which statements for
agreement and disagreement are
arranged in cumulative series. Attitude
items can be arranged in such a order that
a respondent who positively answers to a
particular item also responds positively to
all other items having a lower rank. The
assumption is that ordering of certain

90
stimuli is possible i.e. if an individual
dominates a particular stimulus, he will
also dominate all the stimulus ordered
below that stimulus.If he fails to dominate
a particular stimulus, he will fail to
dominate any of the stimuli above the
stimulus in the order.
For example five arithmetic problems
arranged in increasing level of difficulty:
(i) 6 + 4
(ii) 80 + 33
(iii) 59 – 21 + 87
(iv) 15 ( 28 – 13 ) + 5
(v) ( 25 x 12 ) / ( 10 – 4 + 29 )
A person who does the fifth problem right
is expected to give all the earlier right
answers.
The scalogram (Table containing individual
responses) for the above arithmetic
problem is shown below:
(i) (ii) (iii) (iv) (v) Score
+ + + + + 5
+ + + + – 4
+ + + – – 3
+ + – – – 2
+ – – – – 1
– – – – – 0

91
 In case of attitude statements, an
individual with more favourable attitude
score than another individual must also be
just as favourable or more favourable in
his response to every statement in the set
than the other individual. The objective of
scalogram analysis is to ascertain whether
a given set of attiude statements falls
along a unidimentional continuum from
the least to the most favourable.

4.2.2. Bogardus Social Distance Scale

It measures the degree to which a person
is willing to associate with a class or type
of people. It asks how willing the
respondent is to make various
associations. The results are reduced to a
single score on a scale. There are also
non-comparative versions of this scale.
The Bogardus Social Distance Scale is a
psychometric scale to empirically measure
people's willingness to participate in social
contacts of varying degrees of closeness
with members of diverse social groups,

92
 such as other racial and ethnic groups,
etc.
The scale asks people the extent to which
they would be accepting of each group (a
score of 1.00 for a group is taken to
indicate no social distance):
As close relatives by marriage (score
1.00)
As my close personal friends (2.00)
As neighbors on the same street (3.00)
As co-workers in the same occupation
(4.00)
As citizens in my country (5.00)
As only visitors in my country (6.00)
Would exclude from my country (7.00)
The Bogardus Social Distance Scale is a
cumulative scale (a Guttman scale),
because agreement with any item implies
agreement with all preceding items.
4.2.3. Thurstone scale
In psychology, the Thurstone scale was
the first formal technique for measuring
an attitude. It was developed by Louis
Leon Thurstone in 1928, as a means of
measuring attitudes towards religion. It is
made up of statements about a particular
issue, and each statement has a

93
numerical value indicating how favorable
or unfavorable it is judged to be. People
check each of the statements to which
they agree, and a mean score is
computed, indicating their attitude.
In constructing a Thurstone scale large
number of evenly graduated favourable to
unfavourable attutude statements
pertaining to the subject of enquiry are
formed by either studying existing
literarture, personal experience, or with
discussion with knowledgeable persons.
The statements should be brief, non-
ambiguous, and their acceptance or
rejection should indicate exactly the
respondent’s attitude. Then a scale on
equal appearing continuum is made.
QUESTIONS

Q 1. What are Scalograms? How are they

different from Thurston's scale?
MBA Part -1 2006

94
 Chapter 5

Motivational Research

5.1. Introduction
Motivational research is a type of
qualitative research that attempts to
explain why people behave as they do.
Motivational research seeks to discover
and comprehend what people do not fully
understand about themselves. Implicitly,
motivational research assumes the
existence of underlying or unconscious

95
 motives that influence human behavior.
Motivational research attempts to identify
forces and influences that consumers may
not be aware of (e.g., cultural factors,
sociological forces). Typically, these
unconscious motives (or beyond-
awareness reasons) are intertwined with
and complicated by conscious motives,
cultural biases, economic variables, and
fashion trends (broadly defined).
Motivational research in marketing,
attempts to examine all of these
influences and factors to unravel the
mystery of consumer behavior as it relates
to a specific product or service, so that the
marketer better understands the target
audience and how to influence that
audience.
5.2. Major Techniques
The three major motivational research
techniques are observation, focus groups,
and depth interviews.
Observation

Qualitative Focus group

Research
Depth Interview

96
 Fig. 5.1 Major Techniques of Motivational
(Qualitative) Research

5.2.1. Observation
Observation can be a fruitful method of
deriving hypotheses about human
motives. Anthropologists have pioneered
the development of this technique. All of
us are familiar with anthropologists living
with the “natives” to understand their
behavior. This same systematic
observation can produce equally insightful
results about consumer behavior.
Observation can be accomplished in-
person or sometimes through the
convenience of video. Usually, personal
observation is simply too expensive, and
most consumers don’t want an
anthropologist living in their household for
a month or two.
It is easier to observe consumers in buying
situations than in their homes, and here

97
the observation can be in-person or by
video cameras. Generally, video cameras
are less intrusive than an in-person
observer. Finding a representative set of
cooperative stores, however, is not an
easy task, and the installation and
maintenance of video cameras is not
without its difficulties. In-store observers
can be used as well, so long as they have
some “cover” that makes their presence
less obvious. But, observation by video or
human eye cannot answer every question.
Generally, observation must be
supplemented by focus groups or depth
interviews to fully understand why
consumers are doing what they do.

5.2.2. Focus Group

A Focus group discussion (FGD) is a group
discussion of approximately 6 - 12 persons
guided by a facilitator, during which group
members talk freely and spontaneously
about a certain topic. A FGD is a
qualitative method. Its purpose is to
obtain in-depth information on concepts,
perceptions and ideas of a group. A FGD

98
aims to be more than a question-answer
interaction. The idea is that group
members discuss the topic among
themselves, with guidance from the
facilitator.
FGD techniques can be used to Focus
research and develop relevant research
hypotheses by exploring in greater depth
the problem to be investigated and its
possible causes, Formulate appropriate
questions for more structured, larger scale
surveys, Help understand and solve
unexpected problems in interventions,
Explore controversial topics.
The focus group in the hands of a skilled
moderator can be a valuable motivational
research technique. To reach its full
motivational potential, the group interview
must be largely nondirective in style, and
the group must achieve spontaneous
interaction. It is the mutual reinforcement
within the group (the group excitement
and spontaneity) that produces the
revelations and behaviors that reveal
underlying motives. A focus group
discussion dominated by the moderator
will rarely produce any motivational

99
insights. A focus group actively led by the
moderator with much direct questioning of
respondents will seldom yield motivational
understanding. But the focus group is a
legitimate motivational technique.
5.2.2.1. How to conduct a FGD
A FGD can be regarded as a mini-study. It
requires clear objectives to guide the
research team in the formulation of
discussion questions. Any FGD requires
good knowledge of prevalent conditions
for situational analysis. Participants should
be different categories, like different
socio-economic group or age and gender
composition to facilitate free discussion.
These differences will be reflected in their
perceptions of the problems they suffer
from and possible solutions. A researcher
must be aware of these differences,
otherwise (s)he may miss important
groups of participants or obtain a
hotchpotch of information. Similarly, (s)he
must know which key persons or
organisations could be good entry points
for the selection of participants in the
FGDs. If a FGD forms part of a bigger
study, or project, it may be easy to define

100
target groups for the discussions.
Otherwise, the first task of the
researcher(s) will be to explore the area
and identify possible target groups.
Interviews with some key informants and
a basic or elementary situation analysis
are then required. Communication and
interaction during the FGD should be
encouraged in every way possible.
Arrange the chairs in a circle. Make sure
that there will be no disturbances,
sufficient quietness, adequate lighting,
etc. Try to hold the FGD in a neutral
setting which encourages participants to
freely express their views. A health centre,
for example, is not a good place to discuss
traditional medical beliefs or preferences
for other types of treatment.
There should be a written list of topics to
be covered. It can be formulated as a
series of open-ended questions. One of
the members of the research team should
act as ‘facilitator’ or ‘moderator’ for the
focus group discussion. One should serve
as ‘recorder’. The facilitator should
preferably be as close as possible to the
participants in their characteristics (same

101
sex, roughly same age, etc.). The
facilitator should NOT act as an expert on
the topic. His or her role is to stimulate
and support discussion.
For encouraging involvement some useful
techniques include:
— Asking for clarification:
‘Can you tell me more about. . . ?’
— Reorienting the discussion when it goes
‘off the track’:
Saying: ‘Wait, how does this relate
to. . . ?’
Saying: ‘Interesting point, but how
about. . . ?’
— Using one participant’s remark to direct
a question to another, for example,
‘Mrs. X said . . . , but how about you,
Mrs. Y?’
— When dealing with a dominant
participant, avoiding eye contact or
turning
slightly away to discourage the person
from speaking, or thanking the person
and changing the subject.
— When dealing with a hesitant
participant, using the person’s name,
requesting

102
 his/her opinion, making more frequent
eye contact to encourage his/her
participation.
The recorder should keep a record of the
content of the discussion as well as
emotional reactions and important aspects
of group interaction. Assessment of the
emotional tone of the meeting and the
group process will enable to judge the
validity of the information collected during
the FGD. Items to be recorded include:
Date, time, place; Names and
characteristics of participants; General
description of the group dynamics (level of
participation, presence of a dominant
participant, level of interest); Opinions of
participants, recorded as much as possible
in their own words, especially for key
statements; Emotional aspects (e.g.,
reluctance, strong feelings attached to
certain opinions); Vocabulary used -
particularly in FGDs that are intended to
assist in developing questionnaires. It is
highly recommended that a tape-recorder
be used to assist in capturing information.
Even if a tape-recorder is used, notes
should be taken as well, in case the

103
machine malfunctions and so that
information will be available immediately
after the session for discussion. If there is
no reliable tape-recorder available, it is
advisable to have two recorders.
After each focus group session the
facilitator and recorder should meet to
review and complete the notes taken
during the meeting. This is the right
moment to evaluate how the focus group
went and what changes might be made in
the topics when facilitating the next focus
group. Immediately afterwards a full
report of the discussion should be
prepared which reflects the discussion as
completely as possible, using the
participants’ own words. List the key
statements, ideas, and attitudes
expressed for each topic of discussion.
Present the findings, following the list of
topics discussed and guided by the
objective(s) of your FGD.
5.2.3. Depth Interview
The heart and soul of motivational
research is the depth interview, a lengthy
(one to two hours), one-on-one, personal
interview, conducted directly by the

104
motivational researcher. Much of the
power of the depth interview is dependent
upon the insight, sensitivity, and skill of
the motivational researcher. The
interviewing task cannot be delegated to
traditional marketing research
interviewers—who have no training in
motivational techniques.
During the personal interview, the
motivational researcher strives to create
an empathic relationship with each
respondent, a feeling of rapport, mutual
trust, and understanding. The researcher
creates a climate in which the respondent
feels free to express his feelings and his
thoughts, without fear of embarrassment
or rejection. The researcher conveys a
feeling that the respondent and his
opinions are important and worthwhile, no
matter what those opinions are. The
motivational researcher is accepting,
nonthreatening, and supportive. The
emotional empathy between motivational
researcher and respondent is the single
most important determinant of an
effective interview.

105
The motivational researcher relies heavily
upon nondirective interviewing
techniques. Her goal is to get the
respondent to talk, and keep talking. The
researcher tends to introduce general
topics, rather than ask direct questions.
She probes by raising her eyebrows, by a
questioning look upon her face, by
paraphrasing what the respondent has
said, or by reflecting the respondent’s own
words back to the respondent in a
questioning tone. Nondirective techniques
are the least threatening (and the least
biasing) to the respondent.
Projective techniques can play an
important role in motivational research.
Sometimes a respondent can see in others
what he cannot see—or will not admit—
about himself. The motivational
researcher often asks the respondent to
tell a story, play a role, draw a picture,
complete a sentence, or associate words
with a stimulus. Photographs, product
samples, packages, and advertisements
can also be used as stimuli to evoke
additional feelings, imagery, and
comment.

106
 During the interview, the researcher
watches for clues that might indicate that
a “sensitive nerve” has been touched.
Long pauses by the respondent, slips of
the tongue, fidgeting, variations in voice
pitch, strong emotions, facial expressions,
eye movements, avoidance of a question,
fixation on an issue, and body language
are some of the clues the motivational
researcher keys on. These “sensitive”
topics and issues are then the focus of
additional inquiry and exploration later in
the interview.
Each interview is tape-recorded and
transcribed. A typical motivational study,
consisting of 30 to 50 depth interviews,
yields 1,000 to 2,000 pages of typed
verbatim dialogue. During the interview,
the motivational researcher makes notes
about the respondent’s behavior,
mannerisms, physical appearance,
personality characteristics, and nonverbal
communication. These notes become a
road map to help the researcher
understand and interpret the verbatim
transcript of the interview.
5.2.3.1. The Analysis of Depth Interview

107
The motivational researcher reads and
rereads the hundreds of pages of verbatim
respondent dialogue. As she reads, the
researcher looks for systematic patterns of
response. She identifies logical
inconsistencies or apparent contradictions.
She compares direct responses against
projective responses. She notes the
consistent use of unusual words or
phrases. She studies the explicit content
of the interview and contemplates its
meaning in relation to the implicit content.
She searches for what is not said as
diligently as she does for what is said. Like
a detective, she sifts through the clues
and the evidence to deduce the forces and
motives influencing consumer behavior.
No one clue or piece of evidence is treated
as being very important. It is the
convergence of evidence and facts that
leads to significant conclusions. In the
scientific tradition, empiricism and logic
must come together and make sense.
The analysis begins at the cultural level.
Cultural values and influences are the
ocean in which we all swim and, of which,
most of us are completely unaware. What

108
we eat, the way we eat, how we dress,
what we think and feel, and the language
we speak are dimensions of our culture.
These taken-for-granted cultural
dimensions are the basic building blocks
that begin the motivational researcher’s
analysis. The culture is the context that
must be understood before the behavior
of individuals within the context can be
understood. Every product has cultural
values and rules that influence its
perception and its usage.
Once the cultural context is reasonably
well understood, the next analytic step is
the exploration of the unique motivations
that relate to the product category. What
psychological needs does the product
fulfill? Does the product have any social
overtones or anthropological significance?
Does the product relate to one’s status
aspirations, to competitive drives, to
feelings of self-esteem, to security needs?
Are masochistic motives involved? Does
the product have deep symbolic
significance? And so on. Some of these
motives must be inferred since
respondents are often unaware of why

109
they do what they do. But the analysis is
not complete.
The last major dimension that must be
understood is the business environment,
including competitive forces, brand
perceptions and images, relative market
shares, the role of advertising in the
category, and trends in the marketplace.
Only part of this business environment
knowledge can come from the respondent,
of course, but understanding the business
context is crucial to the interpretation of
consumer motives in a way that will lead
to useful results. Understanding the
consumer’s motives is worthless unless
somehow that knowledge can be
translated into actionable marketing and
advertising recommendations.
Sometimes a motivational study is
followed by quantitative surveys to
confirm the motivational hypotheses as
well as to measure the relative extent of
those motives in the general population.
But many times motivational studies
cannot be proved or disproved by survey
research, especially when completely
unconscious motives are involved. In

110
 these cases, the final evaluation of the
hypothesized motives is by the testing of
concepts (or advertising alternatives) that
address the different motives, or by other
types of contrived experiments.

5.3. Areas of Application of Qualitative

Research in Marketing
 When little is known about a new
territory. When considering products or
new markets, qualitative research can
deliver an early landscape profile of
consumer or business buyer attitudes
and behavior.
 Customer perceptions or attitudes
may be hidden from easy view. When
the product category may represent
unspoken meaning to buyers, qualitative
market research may provide needed
tools.
 Generate ideas for products,
advertising, or brand positioning. The
nuances of buyer attitudes and beliefs
can often provide stimulus for fresh new
ideas, and feed a formal idea generation
process.

111
 Screening ideas and concepts.
Qualitative market research can be a
useful first step, prior to quantitative
research, to screen new advertising,
product, or positioning concepts. This
allows time for refining concepts prior to
quantitative market research.
QUESTION

Q 1.Write a detailed note on

Motivational Research. MBA
Part -1 2006

112
 Chapter 6
Scaling Techniques
6.1. Introduction

113
 Measurement means assigning numbers
or other symbols to characterstics of
objects. For example we measure
consumer (object) preferences,
perceptions, attitude or other relevant
characterstics. Scaling is the extention of
measurement. Scaling involves creating a
continuum upon which the measured
objects are located. For a scale from 1 to
100 for locating consumers according to
the characterstic “attitude toward
department stores.” Each repondent is
assigned a number from 1 to 100
indicating the degree of
(un)favourableness, with 1 = extremely
unfavourable, and 100 extremely
favourable. Measurement is the actual
assignment of a number from 1 to 100 to
each respondent. Scaling is the process of
placing the respondents on a continuum
with respect to their attitude toward
departmental stores.
6.2. Primary scales of Measurement
There are for primary sclaes of
measurement: nominal, ordinal, interval
and ratio.
 Nomianl Scale

114
 A nominal scale is a figurative labeling
scheme in which the numbers and
alphebets serve only as labels for
identification. Common examples include
numbers assigned to football players,
identifying respondents, brands,
attributes, stores and other objects. The
permissible descriptive statistics on
nominal variables are percentages and
mode and inferencial variables are Chi-
square test.
 Ordinal scale
An ordinal scale is a ranking scale in which
numbers are assigned to objects to
indicate the relative extent to which the
objects possess some characterstic, but
not the magnitude of the differences
between the objects. An ordinal scale
allows to determine wheteher an object
has more or less of a characterstic than
some other object, but not how much
more or less. The object ranked first has
more of the characterstic as compared to
the object ranked second, but whether the
object ranked second is a close second or
a poor second is not known. Common
examples of ordinal scales is quality

115
 rankings, ranking of teams in a
tournament, rank of most admired
companies. Measurement of this type
include “grater than” or “less than”
judgements from the respondents. In a
ordinal scale in adiion to the counting
opeartion allowable for nominal scale
data, ordinal scales permit the use of
statistics like median, rank-correrelation.
 Interval Scale
In an Interval scale, numerically equal
distances on the scale represent equal
value in the chracterstics being measured.
An interval scale contains all the
information of a ordinal scale, but it also
allows to compare the differences
between objects. The difference between
any two scale values is identical or
constant or equal interval. For example
difference between 1 and 2 is the same as
the difference between 2 and 3 which is
the same between 5 and 6. A common
example in everyday life is a temperature
scale (Fahrenheit, Centigrade). In
management a common example is
attitudenal data, opinions, obtained from
rating scales.

116
In an interval scale, the location of zero
point is not fixed. Both the zero points and
the units of measurement are arbitrary.
Hence any positive linear transfomation of
the form y = a + bx will preserve the
properties of the scale. Here x is the
original scale value, y is the transformed
scale value, b is a positive constant, and a
is a contant. In addition to the statistical
techniques that can be applied to nominal
and ordinal data, range, mean, standard
deviation, correlation, t-tests, ANOVA,
regression, factor analysis etc. can be
applied.

Primary Scales Help Domino’s to become a Primary Competitor

For understanding the customers’ needs and improving service different scales are
utilized:
 Name, phone number, and (e-mail) address [ Nominal Scale ]
 Preference for pizza restaurants in the local area [ Ordinal Scale ]
 Impressions on the service offered by Domino’s Pizza as a
whole[Interval Scale ]
 Assessments on the products and price [ Interval Scale ]
 Customer satisfaction [ Interval Scale ]
 Amount spent on pizza and fast foods [ Ratio Scale ]

117
 Ratio scale
A ratio scale possesses all the properties
of the nominal, ordinal and interval scales
and in addition, an absolute zero point.
Ratio scales allow proportionate
transfromation of the form y = bx , where
b is a positive constant. A example of this
transformation is the conversion of yards
to feet (b=3). The comparision between
the objects are identical whether made in
yards or feet. Common example of this
scale are length, weight, age, income,
costs, sales, market shares etc. All
statistical techniques can be applied to
ratio data.
6.3. Scaling Techniques
The scaling techniques commonly
employed in business research can be
clasified into comparative and
noncomparative scales. In comparative
scaling, the items are directly compared
with each other (example : Do you prefer
Pepsi or Coke?). Comparative scale data
must be interpreted in relative terms and
have only ordinal and rank properties.
Compararive scales include paired
comparisions, rank order, constant sum

118
scales and other procedures. The major
benefit of comparative scaling is that
small differences between stimulus
objects can be deteceted. As they
compare the stimulus objects,
respondents are forced to choose between
them. Comparative scales are easily
understood and can be easily applied. The
major disadvantage of comparative scales
is the ordinal nature of data and inability
to generalize beyond the stimulus objects
scaled. For instance to compare Miranda
with Pepsi, the researcher woluld have to
a new study. These disadvantages are
substantially overcomed by the non
comparative scaling techniques. In
noncomparative scaling each item is
scaled independently of the others
(example : How do you feel about Coke?).
The resulting data is generally assumed to
be interval or ratio scaled. For example,
respondents may be asked to evaluate
Coke on a 1-to-6 preference scale ( 1 =
not at all preferred, 6 = greatly preferred).
Noncomparative scales can be continuous
or graphic rating or itemised rating scales.
In Continuous rating scales the rater

119
points along the line, depending on the
respondent frame of mind. The Itemised
rating scales are numerical scales in which
the respondent selects one as best
reflecting his evaluation. The itemised
rating scales can be further classified as
Likert, semantic differential, or Stapel
scales. Noncomparative scales scaling is
mostly used in business research.
Scaling
Techniques

Comparativ Non-
e Comparative
Scales (Rating) Scales

Rank Constan Continuous Itemized

Paired
Orde t Rating Rating
Comparison
r Sum Scales Scales

Likert Semantic Stapel

Differential

Fig. 6.1 Classification of Scaling

Techniques

Comparative Scales

120
6.3.1.1. Paired Comparison
In this scaling a respondent is presented
with two objects and asked to select one
according to one criteria.
E.g. obtaining shampoo preferences
using paired comparisions.
Instructions
We are going to present you with 10
pairs of shampoo brands. For each pair,
please indicate which one of the two
brands of shampoo in the pair you
would prefer for personal use.
Recording Form Clinic plus
Organics Ayur Pantene Head &
Shoulders
Clinic plus -
0 0 1 0
Organics 1
- 0 1 0
Ayur 1
1 - 1 1
Pantene 0
0 0 - 0
Head & Shoulders 1
1 0 1 -

Number of times

121
 Preferred 3
2 0 4 1
A 1 means that the brand in that
column was preferred over the brand in
the corresponding row. A 0 means that
the row brand was preferred over the
column brand. The number of times a
brand was preferred is obtained by
summing the 1’s in each column.
The above figure shows the paired
comparion data obtained to access a
respondent’s shampoo preferences. The
respondent made 10 comparisions to
evaluate five brands. In general with n
brands, [n ( n – 1 ) / 2 ] paired
comparisions are required to include all
the possible pairs of objects. To arrive at a
rank order, the researcher determines the
number of times each brand is preferred
by summing the column entries. Therefore
the above respondent order of preference,
from most to least preferred, is Pantene,
Clinic Plus, Organics, Head & Shoulders,
Ayur.
Paired comparision scaling is useful when
the number of objects (e.g. brands) is
limited, because it requires direct

122
 comparision and forced choice. Other
disadvantage of paired comparision is that
it bears little resemblance to the
marketplace situation that involves
selction from multiple alternatives. Also
respondents may prefer one object to
certain others, but they may not like in
absolute sence.
For example, the response pattern of 100
persons in comparing New flavor brand
and established brand:
New Flavour
Established Brand
New Flovour -
70
Established Brand 30
-
Total 30
70
The above data of paired comparision
shows that 70 persons preferred
established brand as compaired to new
flovour.
6.3.1.2. Rank order
In this scaling the respondents are
presented with several objects

123
simultenously and asked to order or rank
them according to some criterion.
For example, repondents may be asked to
rank brands of toothpaste according to
overall preference.
Instructions
Rank the various brands of toothpaste in
oder of preference. Begin by picking out
the one brand that you like most and
assign it a number 1. Then find the
second most preferred brand and asign
it a number 2. Continue this procedure
until you have ranked all the brands of
toothpaste in order of preference. The
least preferred brand should be assigned
a rank of 6.
No two brands should receive the same
rank number.
The criteria of preference is entirely up
to you. There is no right or wrong
answer. Just try to be consistent.
Brand
Rank Order
1. Colgate
________
2. Close Up
________

124
 3. Pepsodent
________
4. Neem
________
5. Babul
________
6. Vicco
________
This type of scaling process more closely
resembles the shopping environment. It
also takes less time and if there are n
stimulus objects, only ( n – 1 ) scaling
decisions need to be made in rank order
scaling. Another advantage is that most
respondents easily understand the
instructions for ranking. The major
disadvantage is that this technique
produces only ordinal data.
6.3.1.3. Constant Sum
In constant sum scaling, respondents
allocate a constant sum of units, such as
points, among the set of stimulus objects,
with respect to some criterion. For
example respondents may be asked to
allocate 100 points of attributes of a soap
in a way that reflect the importance they
attach to each attribute. If an attribute is

125
unimportant, the respondent assigns zero
points. If an attribute is twice as important
as some other attribute, it receives twice
as many points. The sum of all points is
100. Hence the name of the scale. E.g.
Importance of Toilet Soap Attributes using
a Constant Sum Scale
Instructions
Below are eight attributes of toilet
soaps. Please allocate 100 points
among the attributes so that your
allocation reflects the relative
importance you attach to each
attribute. The more points an attributes
receives, the more important the
attribute is. If an attribute is not at all
important, assign it zero points. If an
attribute is twice as important as some
other attribute, it should receive twice
as many points.
Form

Average responses of three

segments
Attribute
Segment I Segment II
Segment III

126
1. Mildness
8 2
4
2. Lather
2 4
17
3. Shrinkage
3 9
7
4. Price
53 17
9
5. Fragrance
9 0
19
6. Packaging
7 5
9
7. Moisturizing 5
3 20
8. Cleaning power
13 60
15
Sum
100 100
100

127
The attributes are scaled by counting the
points assigned to each one by all the
respondents and dividing by the number
of respondents. The results are presented
for three groups, or segments, of
respondents in the above figure. Segment
I attaches overwhelming importance to
price. Segment II considers basic cleaning
power to be of prime importance.
Segment III values lather, fragrance,
moisturizing and cleaning power. Such
information cannot be obtained from rank
order data unless they are transformed
into interval data. However, it has two
primary disadvantages. Respondents may
allocate more or fewer units than those
specified. For example, a respondent may
allocate 108 or 94 points. Another
potential problem is the rounding error if
too few units are used. On the other hand,
the use of large number of units may be
too taxing on the respondent and cause
confusion and fatigue.

Non-Comparative (Rating) Scales

6.3.2.
In this scaling, each stimulus object is
scaled independently of the other objects

128
 in the stimulus set. Non-comparative
techniques consists of continuous and
itemized rating scales.
6.3.2.1. Continuous Rating
Continuous Rating Scale In this scaling
respondents rate the objects by placing a
mark at the appropriate position on a line
that runs from one extreme of the
criterion variable to the other. Thus the
respondents are not restricted to selecting
from marks previously set by the
researcher. The form of continuous scale
may vary considerably. For example the
line may be vertical or horizontal ; scale
points in the form of numbers or brief
descriptions, may be provided; and, if
provided, the scale points may be few or
many. Three versions of a continuous
rating scale are illustrated.
Eg. Department Store like Big
Shopper Rating
Continuous Rating Scales
How would you rate Big Shopper as a
departmental store?
Version 1
Probably the worst - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - Probably the

129
 best
Version 2
Probably the worst - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - Probably the
best
0 10 20 30 40
50 60 70 80 90 100
Version 3
Very bad
Neither bad Very good
nor good
Probably the worst - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - Probably the
best
0 10 20 30 40
50 60 70 80 90 100
Once the respondent has provided the
ratings, the researcher divides the line
into as many categories as desired and
assigns scores based on the categories
into which the rating fall. The advantage
of continuous scales is that they are easy
to construct. However scoring is
cumbersome and unreliable. Hence their
use in business research is limited.
6.3.2.2. Itemized Rating

130
A measurement scale having numbers
and/or brief descriptions associated with
each category. The categories are ordered
in terms of scale position. Commonly used
itemized rating scales are Likert,
Semantic differential and Stapel
scales.
6.3.2.2.1. Likert Scale
A widely used measurement scale with
five response categories ranging from
“strongly disagree” to “strongly agree”,
which requires the respondents to indicate
a degree of agreement or disagreement
with each of the series of statements
related to the stimulus objects.
Illustration of likert scale, in evaluating
attitudes towards Big Shopper as
Department Store.

131
Listed below are different opinions about
Sears. Please indicate how strongly you
agree or disagree with each by using the
following scale:
1 = Strongly disagree
2 = Disagree
3 = Neither agree nor disagree
4 = Agree
5 = Strongly agree
Neither
Agree
Strongly
nor Strongly
Disagree
Disagree Disagree Agree Agree
1. Big Shopper sells high-quality
merchandise. 1 2 3
4 5
2. Big Shopper has poor in-store service.
1 2 3 4 5
3. I like to shop at Big Shopper.
1 2 3 4 5
4. Big Shopper does not offer a good mix
of
different brands within a product
category. 1 2 3
4 5

132
5. The credit policies of Big Shopper are
terrible 1 2 3
4 5
6. Big Shopper is where Jaipur shops.
1 2 3 4 5
7. I do not like the advertising done by Big
Shopper 1 2 3 4
5
8. Big Shopper sells a wide variety of
merchandise. 1 2 3
4 5
9. Big Shopper charges fair prices.
1 2 3 4 5
To conduct the analysis, each statement is
assigned a numerical score, ranging either
from – 2 to + 2 or 1 to 5. The analysis can
be conducted on an item-by-item basis
(profile analysis), or a total (summated)
score can be calculated from each
respondent by summing across items.
Note that for a negative statement, an
agreement reflects an unfavorable
response, whereas for positive statement
agreement represents a favorable
response. In the scale shown above, if a
higher score is to denote a more favorable

133
attitude, the scoring of items 2, 4, 5, and 7
will be reversed.
The likert scale has several advantages. It
is easy to construct and administer.
Respondents readily understand how to
use the scale, making it suitable for mail,
telephone or personal interviews. The
major disadvantage of the Likert Scale is
that it takes longer to complete than other
itemized rating scales because
respondents have to read each statement.
E.g. of Likert scale in investigating Job
Satisfaction
Neither
Agree
Strongly
nor Strongly
Disagree
Disagree Disagree Agree Agree
I get a feeling of accomplishment from
the 1 2 3 4
5
work I am doing.
E.g. of Likert scale in rating the
importance of product characteristics.
Not Important
Important

134
 Reliability 1
2 3 4 5
Price 1
2 3 4 5
Range of features 1
2 3 4 5
6.3.2.2.2. Semantic Differential
The semantic differential is a seven-
point rating scale with endpoints
associated with bipolar labels that have
semantic meaning. A general semantic
differential scale for measuring self-
concepts, person concepts, and product
concepts is shown below:
A general semantic differential scale
1. Strong :–: –:–:–:–:– : –:
Delicate
2. Excitable :–: –:–:–:–:– : –:
Calm
3. Uncomfortable :–: –:–:–:–:– : –:
Comfortable
4. Dominating :–: –:–:–:–:– : –:
Submissive
5. Pleasant :–: –:–:–:–:– : –:
Unpleasant
6. Contemporary :–: –:–:–:–:– : –:
Non-contemporary

135
 7. Organized :–: –:–:–:–:– : –:
Unorganized
8. Rational :–: –:–:–:–:– : –:
Emotional
9. Youthful :–: –:–:–:–:– : –:
Mature
10.Formal :–: –:–:–:–:– : –:
Informal
11. Orthodox :–: –:–:–:–:– : –:
Liberal
12. Complex :–: –:–:–:–:– : –:
Simple
13. Colorless :–: –:–:–:–:– : –:
Colorful
The respondents mark ( X ) the blank that
best indicates how they would describe
the object being rated. Individual items on
a semantic differential scale may be
scored on either a – 3 to + 3 or a 1 to 7
scale. The resulting data are commonly
analyzed through profile analysis. In
profile analysis, means or median values
on each rating scale are calculated and
compared by plotting or statistical
analysis. This helps to determine the
overall differences and similarities among
the objects. To assess differences across

136
segments of respondents, the researcher
can compare mean responses of different
segments. Its versatility makes the
semantic differential a popular rating scale
in business research. It has been widely
used in comparing brands, product, and
company images. It has also been used to
develop advertising and promotion
strategies and in new product
development studies.
6.3.2.2.3. Stapel
The Stapel scale, is a unipolar rating
scale with 10 categories numbered from –
5 to +5, without a neutral point (zero).
This scale is usually presented vertically.
Respondents are asked to indicate how
accurately or inaccurately each term
describes the object by selecting an
appropriate numerical response category.
The higher the number, the more
accurately the term describes the object.
Example Stapel scale
Big Shopper
+5 +5
+4 +4
+3 +3
+2 +2

137
 +1 +1

High Quality Poor

Service
–1 –1
–2 –2
–3 –3
–4 –4
–5 –5
The data obtained by using a Stapel scale
can be analyzed in the same way as
semantic differential data. The Stapel
scale produces results similar to the
semantic differential. The Stapel scale’s
main advantage is that it ensures true
bipolarity of the adjectives or phrases.

6.4. Rating Scales Decisions

The researcher must make six major
decisions when constructing any of these
scales.
1. The number of scale categories to
use : The greater the number of scale
categories, the finer the discrimination
among stimulus objects that is
possible. On the other hand most
respondents cannot handle more than

138
 a few categories. . The appropriate
number of categories should be seven
plus or minus two: between five and
nine.
2. Balanced verses unbalanced scale: -
In a balanced scale , the number of
favorable and unfavorable categories
are equal; in an unbalanced scale, they
are unequal. Example of balanced and
unbalanced scales are given below.
Balanced Scale
Unbalanced Scale
Jovan Musk for Men is
Jovan Musk for Men is
Extremely good __________
Extremely good __________
Very good __________ Very
good __________
Good __________ Good
__________
Bad __________
Somewhat good __________
Very bad __________ Bad
__________
Extremely bad __________ Very
bad __________

139
In general, the scale should be balanced in
order to obtain objective data.
3. Odd or even number of categories:
With an odd number of categories, the
middle scale position is generally
designated as neutral or impartial. The
Likert scale is a balanced rating scale
with an odd number of categories and
a neutral point.
4. Forced versus non-forced choice:
On forced rating scales, the
respondents are forced to express an
opinion, because a “no opinion” option
is not provided. In such a case,
respondents without an opinion may
mark the middle scale position. In
situations where the respondents are
expected to have no opinion, as
opposed to simply being reluctant to
disclose it, the accuracy of data may
be improved by a non-forced scale that
includes a “no opinion” category.
5. Nature and the degree of verbal
description
6. Scale categories may have verbal,
numerical, or even pictorial
descriptions. Furthermore, the

140
 researcher must decide whether to
label every scale category, some scale
categories, or only extreme scale
categories to reduce ambiguity. The
category descriptions should be
located as close to the response
categories as possible.
7. Physical form or Configuration: A
number of options are available with
respect to scale form or configuration.
Scales can be presented vertically or
horizontally. Categories can be
expressed by boxes, discrete lines, or
units on a continuum and may or may
not have numbers assigned to them. If
the numerical values are used, they
may be positive, negative or both.
Illustration
A variety of scale configurations may be
employed to measure the gentleness of a
detergent.
The detergent is:
1. Very ___ ___ ___
___ ___ ___ ___ Very
harsh
gentle.

141
 2. Very 1 2 3
4 5 6 7 Very
harsh
gentle.
3. Very harsh

Neither harsh nor gentle

Very gentle
4. ___ ___ ___ ___
___ ___ ___
Very
Very
harsh
gentle.
–3 –2 –1 0 +1 +2 +3
5.
Very
Very
harsh
gentle
Thermometer scale
In the thermometer scale, the higher the
temperature, the more favorable the

142
evaluation. This scale is particularly used
for children.

Like Very
Much
100o
75o
50o
25o Dislike Very
0o Much

Smiling Face scale

In smiling face scale, happier faces
indicate more favorable evaluations. This
scale is particularly used for children.

1 2 3 4 5

143
6.5. Scale Evaluation
A scale is evaluated for its reliability and
validity and generalizability. Approaches
to assessing reliability include test-retest
reliability, alternative forms reliability, and
internal consistency reliability. Validity can
be assessed by examining content validity
and construct validity.
Scale Evaluation

Reliability Validity

Internal
Test/ Alternativ
Consistenc Content Construct
Retest e Forms
y

144
 Fig. 17.3 Evaluation of a
Scale

6.5.1. Reliability
Reliability refers the extent to which a
scale produces consistent results if
repeated measurements are made.
Potential sources of errors in
measurement include factors like health,
emotions, fatigue, presence of other
people, noise, social desirability,
intelligence of an individual, lack of clarity
of the scale, mechanical factors such as
poor printing, overcrowding of items in the
questionnaire and poor design. The errors
lead in lowering the reliability. Reliability is
defined as the extent to which measures
are free from errors.
Approaches to assessing reliability include
test-retest reliability, alternative forms
reliability, and internal consistency
reliability.
 Test-Retest Reliability
In test-retest reliability, respondents are
administered identical sets of scale items
at two different times under as nearly
equivalent conditions as possible. The

145
time interval between the tests or
administrations is generally two to four
weeks, so that the respondents may not
remember the answers they gave the first
time. The degree of similarity between two
measurements is determined by
computing a correlation coefficient. The
higher the correlation coefficient, the
greater is the reliability.
 Alternative-Forms Reliability
In alternative-forms reliability, two
equivalent forms of the scale are
constructed. The same respondents are
measured at two different times, usually
two to four weeks apart, with a different
scale form being administered each time.
The scores from the administration of
alternative-scale forms are correlated to
assess reliability.
 Internal Consistency Reliability
The internal consistency of a scale is
measured by evaluating the split-half
reliability. The items on the scale are
divided into two halves and the resulting
half scores (summated to form total score)
are correlated. High correlations between
the halves indicate high internal

146
 consistency. The scale items can be split
into halves based on odd-and even-
numbered items or randomly. Cronbach's
alpha is the most common form of internal
consistency reliability coefficient. By
convention, a lenient cut-off of .60 is
common in exploratory research; alpha
should be at least .70 or higher to retain
an item in an "adequate" scale; and many
researchers require a cut-off of .80 for a
"good scale."
6.5.2. Valadity
 Content Validity
It is the measure of the
representativeness of the content of the
scale for the measurement task at hand.
This includes examining whether the scale
items adequately cover the entire domain
of the construct being measured. For
example, a scale designed to measure
store image would be considered
inadequate if it omitted any of the major
dimension (quality, variety and
assortment of merchandise, etc.).
 Construct Validity
It addresses the question of what
characteristic the scale is measuring.

147
 When assessing construct validity, the
researcher attempts to answer theoretical
questions of why the scale works and what
deductions can be made concerning the
theory underlying the scale.
6.6. Internal consistency of the Scale
Internal consistency measures estimate
how consistently individuals respond to
the items within a scale.
The measure of internal consistency
in a scale
Cronbach's alpha also called as the
reliability coefficient is based on work in
the 1940s by Guttman and others, is the
most common estimate of internal
consistency of items in a scale. Alpha
measures the extent to which item
responses obtained at the same time
correlate highly with each other. Alpha is a
measure of level of mean intercorrelation
weighted by variances or a measure of
mean intercorrelation for standardized
data, stepped up for number of items. A
set of items can have a high alpha and still
be multidimensional. This happens when
there are separate clusters of items
(separate dimensions) which intercorrelate

148
highly, even though the clusters
themselves do not intercorrelate hightly.
Also, a set of items can have a low alpha
even when unidimensional if there is high
random error.
In addition to estimating internal
consistency from the average correlation,
the formula for alpha also takes into
account the number of items on the
theory that the more items, the more
reliable a scale will be. That is, when the
number of items in a scale is higher, alpha
will be higher even when the estimated
average correlations are equal. As the
number of items rises, alpha rises.
Also, the more consistent within-subject
responses are, and the greater the
variability between subjects in the sample,
the higher Cronbach's alpha will be.
Finally, alpha will be higher when there is
homogeneity of variances among items
than when there is not.
The widely-accepted social science cut-off
is that alpha should be .70 or higher for a
set of items to be considered a scale, but
some use .75 or .80 while others are as
lenient as .60. That .70 is as low as one

149
may wish to go is reflected in the fact that
when alpha is .70, the standard error of
measurement will be over half (0.55) a
standard deviation.
In SPSS, Cronbach's alpha is found under
Analyze, Scale and Reliability Analysis.
Then in the Statistics button, select Scale
to get alpha.
QUESTIONS

Q 1. Discuss the various types of rating

scales used in business research. Give
suitable examples. MBA Part
I - Exam, 2004.

150
 Chapter 7

Sociometry

7.1. Sociometry

151
Sociometry is the science and art of
measuring relationships between social
configurations or structures and
psychological well-being. Sociogram, is a
systematic method for graphically
representing individuals as points/nodes
and the relationships between them as
lines/arcs.
Sociometry is a broad term indicating a
number of methods of gathering and
analyzing data on the choice,
communication and interaction patterns of
the people in groups. In other words,
sociometry is the study and measurement
of social choice. It is the study through
which we come to know about the
attractions and repulsions of members of
groups.
Sociometry is a study of group behaviour
in several social activities, particularly
with regard to revealing individual
choices. This has served as a useful
technique in analyzing the choices of
social groups with regard to acceptance or
otherwise of a group leader or even in
explaining individual choices with regard
to each other.

152
 Sociometry has been defined as “ a
method for discovering, describing and
evaluating social status, structure and
development through measuring the
extent of acceptance or rejection between
individuals in groups.” Sociometric test is
undertaken to determine the degree to
which individuals are accepted in a group
or for discovering the relationships which
exist among these individuals and for
disclosing the structure of the group itself.
Sociometry is a method used for
discovering and manipulation of social
configurations by measuring the
attractions and repulsions between
individuals in a group.
7.2. Socio Metric Scale
7.2.1. Method of sociometric analysis
The basic technique in sociometry is the
‘sociometric test’ which consists in having
each member in a group chooses from all
other members those with whom he
prefers to associate in specific situations.
A person is asked to give his preference to
one or more other persons according to
one or more criteria supplied to him by the
researcher. Preferences and rejections are

153
 always indicated and evaluated in terms
of particular criteria around which the
group is organized such as roaming
together, eating together, working
together or otherwise associating with
persons in specific situations. The initial
steps in sociometry, therefore, involve first
ascertaining the fundamental criteria
around which the group activities take
place and second, testing for the
attraction, repulsion patterns in terms of
these criteria.
The main forms of sociometric analysis are
the sociometric matrices and sociograms.
7.2.1.1. Sociometric Matrices
An important technique for portraying and
analyzing sociometric data is the matrix.
In sociometric studies, only square or n x
n matrices are used, n being equal to the
number of persons in a group. Rows of
matrix are numbered i ; columns are
labeled j; i and j can stand for any number
and any person in a group. a would mean
ij

any number in the i row and j column of

th th

the matrix. The various entries, frequently

represented by symbols, indicate each

154
 person’s feeling towards every other
member of the group.
Suppose we take an example that there is
group of six member from six Indian
communities (sections) viz. Hinduism,
Buddhism, Christianity, Islam, Jainism &
Sikhism. and are asked to respond to the
sociometric question, “ With whom would
you like to associate among the group
members for work?”. Please give three
names in order of preference. If a group
member chooses another member, his
choice in the matrix is represented by the
subscript 1. On the contrary, if a group
member does not choose another
member, the negative choice is
represented by subscript 0. The S matrix
of this hypothetical example has been
depicted below in the table.

Sociometric Choice Matrix: Six

Member group,
Three Choice question.
Rows Hind Buddhis Christia Musli Jai Sik
→
u t n m n h
Colum
ns
↓

155
Hindu - 1 0 0 1 0
Buddhis 0 - 1 0 0 1
t
Christia 1 0 - 0 1 0
n
Muslim 0 1 1 - 0 0
Jain 1 0 1 0 - 0
Sikh 1 0 1 0 0 -
Total 3 2 4 0 2 1
In the above table there are three kind of
choices, simple or one way, mutual or two
way and no choice. A simple one way
choice is made when i choose j but j does
not choose i. In the above table Hindu
chooses Buddhist but Buddhist does not
choose Hindu. We write this kind of choice
as i → j or Hindu → Buddhist . In case of mutual
choice i chose j and j choose i. This type of
choice is written as i ↔ j . In the above table
Hindu choose Jain and Jain choose Hindu
or Hindu ↔ Jain .
The extent to which any member is
popular in the group is seen by adding the
columns of the matrix. In the above table,
Christian is most popular. He is chosen by
4 other group members. Evidently Muslim
is not at all popular. This can be a socio-

156
 metric scale for Indian society, showing
how close its different sections
(communities) with each other are.
7.2.1.2. Sociogram
Sociogram is a technique by which the
sociometric data is diagrammatically
presented. In this technique, individuals
are placed on paper and their choices are
marked with arrows. Target sociograms is
drawn in circle shape and is most widely
drawn for small and medium size groups
having members not more than 30. While
preparing such sociogram, the most
chosen individual is placed in the middle
and the least chosen or the isolated
individual on the outer circle. Other
members get their positions in between
according to their sociometric status. The
target sociogram of the above matrix of
the group of six members from six Indian
communities is shown below.

Hindu Muslim

Jain

Buddhist
Christian

Sikh
Choice 157
 Fig. 19.1 Circular type of sociogram

QUESTIONS

Q 1. Design a detailed socio-metric scale for

Indian society, showing how close are
its different sections (communities) with
each other. MBA Part – I 2006

158
 Chapter 8

Case-Studies
8.1. Case Study – RTU MBA 2007
(Solved)
The administrators of Soni Hospital, Jaipur
one of the city's largest, and interested to
know the attitudes of the patients towards
various aspects of the hospital after a stay
of approximately one week or longer. The
hospital administrators are specifically
interested in the attitudes of those
patients who have been treated in the
hospital from 01, April 2006 onwards till
today and on average 2500 patients are
treated in the hospital. Out of this almost
35 per cent patients stay beyond one
week in the hospital for treatment.
They have authorized the undertaking of
the marketing research study to gather
this, and have directed that it covers the
following areas: All administrative
functions with which patients come into
contact (such as admitting, billing and so
on); the treatment and the attitudes of
159
interns and doctors; the food service and
the general housekeeping; the attention of
nurses and nurse's aides; the courtesy and
care of personnel in such departments as
physical therapy and x-ray; the availability
of religious services; and a generally quiet
and clean environment.
( a ) What research design (exploratory,
descriptive, experimental) would you
recommend for this study? and why?
Ans.: This proposed study is
exploratory in nature with the intention
of discovering the attitudes of patients
treated within an year.
( b ) What would be the best sample
method and why? What sample size
would you recommend.
Ans.: A random sampling method can
be used as we have the data about
complete population (i.e. 35% of 2500
= 875 patients). Out of finite
population size is 875, 100 patients
can be chosen randomly by the use of
Random Numbers.
( c ) Design a questionnaire that can be
used for this study.

160
 Ans.: A five point likert scale can be
used to measure variables like service
level in administrative functions,
accuracy of treatment received, food
service and courtesy of hospital staff
etc. See chapter 3 sample
questionnaires.
8.2. Case Study – RU MBA 2005
(Solved)
One of the most costly aspects of living in
America involves hospitalization and
medical costs that are incurred when a
person becomes ill or is hospitalized. A
hospital stay can cost over $2,000 per day
in addition to the money spent on various
tests and the medical bills for visits to the
doctors and the cost of medicines which
can be very high. That is one reason why
it is very important to have health
insurance coverage and most people,
when they enter the job market, make
sure that the job that they accept carries
good health benefits. However, most
people, who are in their own businesses,
have to carry their own insurance and that
can be very costly. Those who are not
insured and cannot afford to pay the

161
medical bills are subsidized by
Government.
In a recent study, the researchers have
found that the number of New York City
residents, under the age of 65, who are
without health insurance have jumped
from 20.9% to 24.8% in the last five years.
As the cost of heath care has gone up, the
companies have cut down on heath
benefits or have increased premiums to be
paid by the workers towards heath
insurance, making it difficult for low
income families to keep health insurance.
While the government health insurance
programmes have continued to cover the
elderly and the very poor, it is the low and
lower middle class in economic terms, who
have become increasingly vulnerable.
Most uninsured families earn between
$15,000 and $45,000 per year. The
number of uninsured Americans under the
age of 65 has risen to 40.3 million by the
year 1995 or nearly 17.4% of the
population. The comparative figures for
New York State are 17.2%, California
22.7% and Texas 27%. Both California and
Texas have large immigrant populations.

162
A survey report on the uninsured in New
York City issued the following
conclusions:-
( a ) Fewer than half of New York City
residents now have private insurance; it
has come down from 60% only five
years ago.
( b ) The proportion of children who have no
insurance has risen from 14% to 20% in
these five years (1990-1995).
( c ) About 22% of the uninsured work for
comparatively large companies which
have more than 1000 employees. This
is contradictory to the conventional
wisdom that large companies offer
comprehensive benefits.
( d ) 41% of men and 28.4% of women
between the ages of 18 and 24 are
uninsured as of 1995. The younger
group probably feels that they are less
likely to get sick and hence it is not
worth spending a lot of money for
health insurance.
Many health professionals are of the
opinion that the situation could worsen
unless concrete steps are taken by the

163
Government to provide affordable health
insurance to the lower income workers.
(i) How do you think these figures as
described in the case study were
obtained?
Suggested Solution
Statistical surveys can be used to collect
quantitative information about items in a
population. Surveys of human populations
and institutions are common in political
polling and government, health, social
science and marketing research. A survey
may focus on opinions or factual
information depending on its purpose, and
many surveys involve administering
questions to individuals.
When the questions are administered by a
researcher, the survey is called a
structured interview or a researcher-
administered survey. When the questions
are administered by the respondent, the
survey is referred to as a questionnaire or
a self-administered survey.
Data collection methods that can be
employed – Telephonic survey, Mailing
questionnaire, Online Surveys, Personal in-
home survey etc.

164
( ii ) How reliable are these figures?
Suggested Solution
The reliability of data collected through
survey depends on subjects’ motivation,
honesty, memory, and ability to respond.
Subjects may not be aware of their
reasons for any given action. They may
have forgotten their reasons. They may
not be motivated to give accurate
answers, in fact, they may be motivated to
give answers that present themselves in a
favorable light. The reliability of the data
collected can depend on following factors:
 Although the chosen survey individuals
are often a random sample, the
respondents are usually self-selected,
and therefore a non-probability samples
from which the characteristics of the
population sampled cannot be inferred.
 The bias associated with self-selection
makes them scientifically worthless
unless response rates are high.
( iii ) Would you like to interpret
conclusions differently?

165
8.3. Case Study – RU MBA 2004
(Solved)
Suppose you are asked to carry out
research to find out knowledge, attitude
and practices of customers towards soft
drinks after knowing pesticides residues
being used.
How shall you carry out the research?
Explain with suitable research
instruments, illustrations assumptions and
examples.
Suggested Explanation
The two global majors PepsiCo and Coca-
Cola Co.dominate the soft drink market in
India. Coca- Cola, which had winded up its
India operations during the introduction of
the FERA regime, re-entered India 16
years later in 1993. Coca-Cola bought
local brands-Thumps Up, Limca and Gold
Spot from Parle Beverages and soft drink
brands Crush, Canada Dry and Sport Cola
from Cadbury Schweppes in early 1999.
Pepsi started a couple of years before
Coca Cola in 1991 has bought over
Mumbai based Duke ’s range of soft drink
brands.

166
According to NCAER survey 91%of the
total consumption of soft drinks in the
country is by lower, lower middle and
upper middle class people. Per capita
consumption in India is among the lowest
in the world at 6 bottles per annum
compared to 80 bottles in Thailand and
800 bottles in USA. Delhi market has
highest per capita consumption in the
country with 50 bottles per annum.(India
Info line Sector Report,2002).
An Exploratory Research has to be
carried out to know the consumers
attitude towards pesticides being found in
soft drinks. It can be used in formulating
problems more precisely, clarifying
concepts, gathering explanations, gaining
insight, eliminating impractical ideas, and
forming hypotheses. Exploratory research
can be performed using
 A literature search
 Surveying certain people about
their experiences:
When surveying people, exploratory
research studies would not try to acquire a
representative sample, but rather, seek to
interview those who are knowledgeable

167
and who might be able to provide insight
concerning the relationship among
variables.
Sampling
A Survey has to be conducted for to find
out knowledge, attitude and practices of
customers towards soft drinks after
knowing pesticides residues being used.
The cities that can be selected for survey:
Delhi, Mumbai, Kolkata, Hyderabad,
Chennai, Bangalore, Ahmedabad,
Lucknow, Jaipur, Patna, Pune, Indore,
Ludhiana, Chandigarh and Kochi.
Approx. number of respondents from each
city can be 500, divided in male and
female respondents. The age profile has to
taken care of.
Data Collection
There can be several ways of
administering the survey, including ,
Telephone , Mail , Online Surveys ,
Personal in-home survey , Personal mall
intercept survey etc. The major brands
that can be included in the research:
Pepsi, Coke, Thump Up, Maaza, Limca,
Fanta, Sprite, Miranda and 7 up.
Data Analysis

168
 Following data analysis techniques can be
used: Univariate analysis like mean,
frequency charts, histograms. Bivariate
analysis like correlation, regression
analysis. ANOVA to find differences
between means, Chi Square Test to find
any difference between observations and
expectations, Factor Analysis to identify
factors affecting attitude and practices,
Multidimentional Scaling can be used to
understand the consumer perceptions.
8.4. Case Study – Market Research
A toy manufacturer is developing a new
toy for 3-5 year olds for which they now
have to finalize the product design and
introduce it to the market. Design a
research project, taking into consideration
any likely problems that may be
encountered, that will allow them to:
1. Identify the factors that influence
the decision to buy a toy;
2. Determine what parents and
children think of the toy; and,
3. Assess the effectiveness of three
alternative advertisements for
promoting the toy.
Suggested Solution

169
 1. Identify the factors that influence
the decision to buy a toy
Overall, this case calls for an exploratory
research design. The population of interest
is the people who buy the toys, that is,
parents (or other buyers such as
grandparents, aunts/uncles) or the 3-5
year olds, and the data analysis will be
qualitative.
Research methods: as this section of the
research is exploratory, appropriate
primary research methods would include
focus groups and individual in-depth
interviews, however, before primary
research is conducted, the manufacturer
should check to see whether any
secondary data is available. Sources on
secondary data might include previous
research conducted by the manufacturer
or its distributors (e.g. toy retailers), and
information available from trade
association or published information
available from trade association or
published information in trade journals.
Secondary data should be considered first
as it is cheaper and quicker than primary
data. However, if no secondary data

170
exists, then the manufacturer should
conduct 4-6 focus groups. The interaction
available with this method of research
makes it preferable to individual in-depth
interviews, especially as the topic under
discussion is unlikely to be sensitive. In
addition, conducting group interviews will
make it possible to collect the data faster
and more cheaply and will facilitate the
exchange of views among the
participants, than if interviews were
conducted with individuals.
Sampling: As this is an exploratory
research, a non-probability sampling
method is likely to be used – a judgment
sample obtained by using a sampling
frame from local primary schools or
nurseries may be available, however, if
this is not possible, a convenience sample
may be recruited from the general
population using a screening question to
check that the potential participant buys
for a child of 3-5. This type of sampling is
possible because there is a relatively high
incidence rate of potential participants in
the population. Overall, the sample size
will be relatively small with 8-12

171
respondents required for each focus
group. (min 24, max 72).
Data analysis: As the data collected is
qualitative, the data analysis will be
subjective. The researcher will have to
interpret the data to determine which
factors have an impact – common factors
may emerge from all of the focus groups,
or a certain aspects in individual focus
groups, however, all factors should be
considered of equal importance at this
stage as exploratory research is not
quantifiable and as such it is not possible
to rank the importance of individual
factors.
2. Determine what parents and
children think of the toy
Determine options is descriptive
research, as such quantitative data will
be collected. As the opinions of both
parents and children are needed, two
research designs are required.
(i) Parents
Research methods: A survey using a
structured questionnaire with
standardized response formats needs to
be conducted for the parents to evaluate

172
the toy. As the toy will also have to be
available, face-to-face contact of some
kind is needed for this research design. As
such, door-to-door interviewing or mail
intercept methods would be appropriate.
Measurement should ideally be interval or
ratio, and questions concerning liking and
purchase intent should be included on the
questionnaire, alongside demographic
information.
Sampling: As this is a descriptive
research a sample size of at least 30 is
required. As mail intercept survey method
is being used, non-probability sampling
will have to be employed. A quota sample
that covers all the relevant demographic
groups is likely to be appropriate.
Data analysis: After the questionnaires
have been edited, coded and data
cleaned, parametric data analysis will be
used to determine whether parents like
the toy and how likely parents are to buy
the toy (means).
3-5 year olds
Research Methods: As the survey is not
appropriate for this age group a structured
observation using a standardized form will

173
be used to determine whether children
like the toy. An observation room will be
set up to resemble a playroom with a
variety of toys. Trained researchers will
observe the children at play and record
how the toy is played with, and for how
long. In addition, a specialist researcher
will help to supervise the children and ask
them what they like about it in an informal
way.
Sampling: As this is a descriptive
research, a sample size of at least 30 is
required, however, this will have to be
broken down into smaller groups in order
to make the observation feasible.
Obtaining a sample of 3-5 year olds will
require the cooperation and consent of the
main carer. However, as a convenience
sample will be used, it might be
appropriate to approach entry level
teachers or nurseries for possible
respondents.
Data analysis: As the sample size will be
at least 30, it may be possible to conduct
parametric analysis for measures such as
relative time spent playing with the toy
(ratio measurement – mean available),

174
however, some of the measurement will
be nominal i.e. what they did with the toy,
and as such tests such as Chi-Square
might be used to expose any differences
between how children played with the
subject toy and other alternatives. The
responses to the informal questioning
should help the researchers to interpret
the qualitative data they have collected.
3. Access the effectiveness of three
alternative advertisements for
promoting the toy.
This aspect of the research problem is
causal; as such quantitative data will be
collected via experiments. As the
opinions of both the parents and
children are needed, the experimental
design will have to take this into account.
Research method: An experimental
design that allows all three
advertisements to be evaluated needs to
be used, however, the effectiveness of no-
advertisement also needs to be
considered as such a control group needs
to be used. The ‘effectiveness’ measure
also needs to be decided; it must be
awareness, desire for the toy, intention to

175
buy or actual sales or change in any of
these measures. As field experiments are
very expensive and expose the new
product to the competition, laboratory
experiments, with their greater internal
validity are the more appropriate choice.
The experimental design will consist of
three experimental groups and one control
group for the parents and the same for the
children. While the measurement of the
parents’ response to the advertisements
may be achieved via a structured
questionnaire, the measurement of the
children’s response may need to occur via
observation.
Sampling
The sampling frame can be same as
above, but a greater number of sample
size is required (at least 120 of each, 30
for each experimental group), for this
experiment. Once the non-probability
sample has been drawn, each pair should
be randomly assigned to the experimental
group and exposed to the advertisement
concerned. Overall the research design
will be as below:

176
 Sample Pre-treatment *
Treatment Post-treatment
Ad1 (R) 30 O1 X1 O2
Ad2 (R) 30 O2 X2 O4
Ad3 (R) 30 O3 X3 O6
Control(R) 30 O7 O8
* if change in measurement
of interest required
Data Analysis: Parametric data analysis
that distinguished between groups is
needed to determine which advertisement
is most effective. An ANOVA that looks at
the differences in the post-observation
measures would be an appropriate test
(separately for the measures for adults
and children). If the pre-observation
measurements were considered
important, then ANCOVA could be used (or
the parameter compared could be the
difference between the post and the pre-
treatment measures in an ANOVA).

***

177