Content uploaded by Sanjay Mohapatra
Author content
All content in this area was uploaded by Sanjay Mohapatra on Jun 22, 2017
Content may be subject to copyright.
CHAPTER 4
SCALES AND MEASUREMENT
Once the research problem has been clearly established, the most important part of the
research, namely data collection, begins. A proper measurement system has to be
developed before actually venturing into the field to collect data. At this stage a
researcher has to address some fundamental issues relating to the variables that need to be
measured, and the different measurement scales that have to be used for measuring the
characteristic that are relevant to the research study.
The process of assigning numbers or labels to different objects under study to represent
them quantitatively or qualitatively is called measurement. Measurement thus can be
understood as a means to denote the amount of a particular attribute that a particular
object possesses. An important aspect of measurement is that there are certain rules that
will define the process of measurement; for instance, a rule might be developed which
says that number 1 should be assigned to people who are from South India and number 2
should be assigned to people who are from North India. It is however important to note
that measurement is done for the attributes of the units under study but not the units
themselves. For example the height, weight, age or other such attributes of a person are
measured but not the person himself.
In this chapter we discuss some of these issues involved in measurement, such as
determining the variables that have to be measured in a business research process, the
different types of measurement scales available and their uses. The criteria for good
measurement are also covered, and finally the chapter concludes with a discussion on the
different errors that arise during the process of Measurement.
IDENTIFYING AND DECIDING ON THE VARIABLES TO BE MEASURED
The primary step in the measurement process is to identify the area or the concept that is
of interest for the study. A concept can be understood as a general idea derived or inferred
from specific instances or occurrences. Once a concept has been identified, e.g. studying
the motivational levels of employees in an organization, the researcher can focus on
developing a construct. A construct is a general idea or an abstract inferred or derived
from specific instances. Constructs can also be considered as certain types of concepts
which exist at different levels of thought that are developed to simplify complex
situations concerning the area of study. They are developed for theoretical usage as well
as for explaining the concepts themselves.
After developing a construct the subsequent process in the research is to define the
concept constitutively and then operationally. A constitutive definition of the concept will
specify the research boundaries, and also will define the central theme of the study. The
most important use of defining the concept (under study) constitutively is to clearly
demarcate it from other concepts. The primary purpose of defining the concept
constitutively is that it will help the researchers in framing and addressing the research
question in an appropriate manner. For instance, if we just say that we want to study the
education system in India, this will not help at all in developing a research question, since
it has to be clearly defined as to what education system needs to be studied - is it primary
education or secondary or higher education, or is it related to an adult education
programme, and so on. So the constitutive definition in this example would be, say,
‘Primary education covering state government aided schools (classes I to V)’.
Once the constitutive definition is clearly defined, it becomes easier to develop an
operational definition. The operational definition defines precisely what attributes and
features of the concept are to be measured. It also specifies the process of assigning a
value to the concept. Although operational definitions can be developed for defining the
characteristics that need to be measured, it is sometimes impossible to measure certain
features which may nevertheless be crucial for the study. For instance if we want to study
the behavior of employees towards the senior management, then it is very difficult to
Exhibit 4.1
Studying Role Ambiguity
The constitutive definition of role ambiguity can be framed in the following manner. Role ambiguity
is a direct function of the discrepancy between the information available to the person and that which
is required for adequate performance of his role. Subjectively, it is the difference between his actual
state of knowledge and that which provides adequate satisfaction of his personal needs and values.
On the other hand, the operational definition can be framed as the state of uncertainty (measured on a
five point scale ranging from highly uncertain to highly certain) an employee feels regarding the
duties and responsibilities of his job relating to his co-employees and customers.
The measurement scale that has been developed consists of a 45-item scale. Each of these 45 items is
analyzed on a five point scale. The five points in the scale represent
1 = highly certain, 2 = certain, 3 = neither certain nor uncertain, 4 = uncertain, 5 = highly uncertain.
Some of the items that have been measured are given below:
What is the amount of work that I am expected to do?
What should I do to improve my chances of getting a promotion?
How vulnerable is my position in the organization?
How far will my boss go to back me?
What methods would my boss use to evaluate my performance?
What is the level of service that I should provide to my customers?
Which specific company strengths should I present to the customers?
How the top management expects me to handle ethical situations in my job?
How much information should I provide to managers from other departments?
About how much time does my family feel I should spend on the job?
Adapted from Jagdip Singh and Gary K Rhoads, “Boundary Role Ambiguity in Marketing-Oriented
Positions: A Multidimensional Mulitfaceted Operationalization,” Journal of Marketing Research,
August 1991, Vol 28, Issue 3, p 328, 10 p.
measure the behavior of these employees; however by defining behavior as the action or
reaction of an individual in response to external stimuli we can develop some scales for
measuring behavior e.g. by asking respondents some indirect questions about how they
would react to certain decisions of the top management and so on. An operational
definition therefore acts as an interface between the theoretical concepts and the live
environment. We can analyze the constitutive and operational definitions along with the
measurement scales in Exhibit 4.1. In this exhibit, the operational definition of role
ambiguity has been developed for studying salespeople and customer service people, on
the assumption that role ambiguity increases the stress factor leading to job
dissatisfaction.
Determining the variables that need to be measured is very important in business
research. In normal measurement applications, scales are usually comparable; for
instance, if we want to measure the height of a person, we measure it in centimeters or in
inches, where both scales are comparable. But in business research, we rarely find such
comparable scales of measurement. While conducting research regarding business issues,
a researcher has to initially define what is to be measured, how it will be measured, and
also the concept that needs to be measured. The concept can be measured using several
factors, but the appropriateness of the variable that has to be measured is very important.
For example if we want to measure the profitability of a particular product, then
measuring the sales of the product would be more appropriate than measuring the
productivity of the organization. A research study has even been conducted to measure
the materialism aspect in human beings. It is discussed in Exhibit 4.2.
While developing measurement variables, researchers often face the problem of construct
equivalence. This refers to the perceptions and beliefs of the measurement variables of
Exhibit 4.2
Measurement Scales for Measuring Materialism
A research study has been conducted to study a fundamental attribute affecting consumer behavior –
namely the materialism of people. Three scales have been developed to test materialistic traits.
These are: possessiveness, non-generosity and envy. Possessiveness has been defined as the
inclination and tendency to retain control or ownership of one’s possessions. Generosity is the
willingness to share with others, therefore in the context of materialism, we can understand
nongenerosity is the unwillingness to share with others. Envy has been defined as an interpersonal
attitude involving displeasure and ill-will at the superiority of another person in happiness, success,
reputation or the possession of anything desirable. The primary reason for studying the materialism
aspects of people is to understand people’s behavior as consumers, their affinity towards
possessions (possessiveness), their willingness to share the possessions with others (non-generosity)
and their feelings about objects in others’ possession (envy).
A sample of 338 members comprising people from different walks of life like students pursuing
business education, employees in an insurance firm, students at a religious institute, shop floor
workers, and so on was selected. The sample was tested for reliability, validity and their
relationships to measures of happiness. The study found that possessiveness and nongenerosity were
very similar between male and female members in the sample, but it was found that men were more
envious than women.
Adapted from Russell W. Belk, “Three Scales to Measure Constructs Related to Materialism:
Reliability, Validity and Relationships to Measures of Happiness,” Advances in Consumer
Research, 1984, Vol 11, Issue 1, p291, 6p.
different people that are related to the study. Different perceptions based on the customs,
religious aspects, culture and socioeconomic factors of different societies will affect the
development of constructs for the research study. For instance, consuming beef is not
accepted in Hindu dominant India, but in the western countries it is a common
phenomenon. As a result common questionnaires cannot be developed for both these
areas if a study on beef consumption patterns is carried out across the world.
DEVELOPMENT OF MEASUREMENT SCALES
Developing measurement scales is a critical dimension of business research. A scale can
be defined as a set of numbers or symbols developed in a manner so as to facilitate the
assigning of these numbers or symbols to the units under research following certain rules.
Generally, it is very easy to measure certain parameters such as sales of a particular
product or the profitability of a firm, or the productivity of the employees in an
organization, and so on. These are relatively easier because they can be measured
quantitatively by applying different scales for measurement. On the other hand it is
relatively difficult to measure some aspects like the motivational levels of employees in
an organization, the attitude of customers towards a particular product, or the customer
acceptance levels of a new design of a product, and so on. Measurement of such concepts
is very difficult because the respondents may be unable to put their feelings across exactly
in words, and sometimes the scales may not be capable of drawing the right response
from the respondent.
Exhibit 4.3
Measuring Customer Retention
Customer retention has become a vital ingredient in business success. Researchers adopt different
approaches for measuring the customer retention rates. One such method is the crude retention rate,
which represents the absolute percentage of customers retained. For instance if 80 out of 100
customers are retained then the retained percentage is 80. However, researchers try to adopt better
methods of measuring the customer retention rate such as weighted retention rates, where the
customers are weighted according to the volume of purchases made by them. Another useful
approach in measuring customer retention rate is the ‘lifetime value’ (LTV). Here the net present
value of the customer is analyzed by the seller. In LTV analysis, costs such as the selling and
servicing costs are considered, while costs involved in developing new customers are recorded as a
sunk cost. The LTV of a customer is calculated by considering the net value of cash flows assuming
a sustainable relationship with the customer in the future.
Although it is a better approach, LTV has some inherent disadvantages. Researchers are unsure
about which attribute to consider for measuring the LTV- should it be the age of the customer, the
working life of the product, product life cycle, or some other factor. Moreover, calculating the LTV
for each and every individual customer is a very difficult process, therefore LTV of customers is
normally carried out at an aggregate group level.
Adapted from K. Ramakrishan, (Strategic Marketing Research Team), “Customer
Retention: The Key to Business Performance,” <
http://www.etstrategicmarketing.com/smNov-Dec2/art11.html >
At times, the respondents might not be willing to reveal their opinions to the researcher.
To overcome such difficulties, a researcher’s primary objective is to seek the cooperation
of the respondent and create an environment of trust and mutual understanding. The
interviewer should try to reduce all the negative feelings of the respondent and develop a
situation wherein the respondent feels free to share all his feelings relevant to the research
with the interviewer. It is also important for the researcher to clearly specify what
information he needs and why, if the research design permits. Companies generally
develop scaling techniques to measure certain critical aspects of business, such as meas-
uring customer retention, as discussed in Exhibit 4.3.
TYPES OF MEASUREMENT SCALES
The design of a measurement scale depends on the objective of the research study, and
the mathematical or statistical calculations that a researcher expects to perform on the
data collected using the scales. The objective of the research study may be as simple as
classifying the population into various categories, or as complex as ranking the units
under study and comparing them to predict some trends. Different types of measurement
scales are given below.
•Nominal scale
•Ordinal scale
•Interval scale
•Ratio scale
Nominal Scale
A nominal scale uses numbers or letters so as to identify different objects. The scale helps
segregate data into categories that are mutually exclusive and collectively exhaustive.
This scale assigns numbers to each of these categories and these numbers do not stand for
any quantitative value, and hence they cannot be added subtracted or divided. For
example, a nominal scale designed to measure the nature of occupation (employment
status) may be given as below:
Occupation: [1] Public sector [2] Private sector [3] Self employed [4] Unemployed [5]
Others
In the above example, the numbers 1, 2, 3, 4 and 5 only serve as labels to the various
categories of employment status, and hence a researcher cannot use those numbers to
perform any type of mathematical or statistical operations on those numbers. A nominal
scale does not give any relationship between the variables, and the only quantitative
measure is the frequency of items appearing under each category i.e. the number of
people in public sector jobs, etc. One can only calculate the mode for the data collected
using nominal scale.
Ordinal Scale
An ordinal scale is used to arrange objects according to some particular order. Thus, the
variables in the ordinal scale can be ranked. For instance, if someone says that a person
came second in the exam, then we can understand that there was another person who
came first and some others were there who were ranked after him. This type of scale that
gives ranks is called an ordinal measurement scale. Ordinal variables can only give us the
information regarding relative position of the participants in the observation, but they do
not give any information regarding the absolute magnitude of the difference between the
first and the second position, or second and third position and so on.
For example, an ordinal scale used to measure the preference of customers (in Andhra
Pradesh) for various mobile telephone service providers would ask a question like
Please rank the following mobile telephone service providers from 1 to 5 with 1
representing the most preferred and 5 the least preferred.
Airtel ____
Hutch ____
Idea ____
BSNL ____
Reliance ____
A respondent may rank these players depending on his experience/perception of them. If
a respondent ranks Airtel as 1 and Idea as 2, a researcher can know that the respondent
prefers Airtel. However, the limitation is that the researcher cannot be sure as to how
strong the respondents’ liking is for Airtel when compared to Idea.
Interval Scale
Interval scales are similar to ordinal scales to the extent that they also arrange objects in a
particular order .However, in an interval scale the intervals between the points on the
scale are equal. This is the scale where there is equal distance between the two points on
the scale. Examples of interval scales are Fahrenheit and Celsius scales used to measure
temperature. In these scale the difference between the intervals is the same i.e. the
difference between 400 and 600 is the same as the difference between 250 and 450. But the
base point, freezing of water is represented by 320F and 00C. Thus there is no natural zero
(base) for these scales.
Similarly we can design an interval scale with points placed at an interval of 1 point
[10] ----- [9] ----- [8] ----- [7] ----- [6] ----- [5] ----- [4] ----- [3] ----- [2] ----- [1] and ask
the respondents to place the mobile telephone service providers on this scale of 10 to 1. If
Idea is assigned 8 and BSNL 4 we can say that the value of difference in preference is 4.
But we cannot say that the liking for Idea is twice that for BSNL because we did not
define a point of no liking i.e. 0. The only statement we can make about a respondents
preference for Idea and BSNL is ‘he likes Idea more than BSNL’ but we can’t give a ratio
of the likings as there is no base zero.
Interval scales are suitable for the calculation of an arithmetic mean, standard deviation,
and correlation coefficient.
Ratio Scale
Ratio scales have a fixed zero point and also have equal intervals. Unlike the ordinal scale
the ratio scale allows for the comparison of two variables measured on the scale. This is
possible because the numbers or units on the scale are equal at all levels of the scale. A
very good example of ratio scale is distance; for instance, not only can we say that the
difference between four miles and six miles is the same as the difference between six
miles and eight miles but we can also say that eight miles is twice as long as four miles.
In other words a ratio scale can be defined as a scale that measures in terms of equal
intervals and an absolute zero point of origin exists. This zero is common to a distance
scale using yards, meters, etc. Age, height, weight, money scales are other common
examples of ratio scales. Since their exist an absolute zero on the ratio scale the data
collected can be subjected to any type of mathematical operation say, addition,
subtraction, multiplication, and division.
CRITERIA FOR GOOD MEASUREMENT
Researchers normally develop their own scales for measuring variables for different
attributes as it is very difficult to find readily available scales. It is in this process of
developing scales that researchers have to be very careful, since the scales that they
develop should primarily stand the tests of reliability, validity, sensitivity and so on. In
the following sections, we will discuss the criteria for good measurement. There are five
major criteria for analyzing the goodness of a measurement, namely, reliability, validity,
sensitivity, generalizability and relevance.
Reliability
It is considered that, when the outcome of a measuring process is reproducible, then the
measuring instrument is reliable. Reliable measuring scales provide stable measures at
different times under different conditions. For example, if a coffee vending machine gives
the same quantity of coffee every time, then it can be concluded that the measurement of
the coffee vending machine is reliable. Thus reliability can be defined as the degree to
which the measurements of a particular instrument are free from errors and as a result
produce consistent results. However in certain situations, poor data collection methods
give rise to low reliability. The quality of the data collected can become poor if the
respondents do not understand the questions properly and give irrelevant answers to them.
There are three methods that can be used to evaluate the reliability of a measure. They are
test-retest reliability, equivalent forms and internal consistency.
Test-retest reliability
If the result of a research is the same, even when it is conducted for the second or third
time, it confirms the repeatability aspect. For example if 40 percent of a sample say that
they do not watch movies, and when the research is repeated after sometime and the result
is same (or almost the same) again, then the measurement process is said to be reliable.
However there are certain problems regarding the test-retest method of testing reliability,
the first and foremost issue is that it is very difficult to obtain the cooperation and locate
all the respondents for a second round of research. Apart from this, the responses of these
people may have changed on the second occasion, and sometimes environmental factors
may also influence the responses.
Equivalent form reliability
Some of the shortcomings of test-retest reliability can be overcome in this method. In
equivalent form reliability, two measurement scales of a similar nature are to be
developed. For instance, if the researcher is interested in finding out the perceptions of
consumers on recent technologically advanced products, then he can develop two
questionnaires. Each questionnaire contains different questions to measure their
perceptions, but both the questionnaires should have an approximately equal number of
questions. The two questionnaires can be administered with a time gap of about two
weeks. The reliability in this method is tested by measuring the correlation of the scores
generated by the two instruments. The major problem with equivalent form reliability is
that it is almost impossible to frame two totally equivalent questionnaires.
Internal consistency
Internal consistency of data can be established when the data give the same results even
after some manipulation. For example, after a research result is obtained for a particular
study, the result can be split into two parts and the result of one part can be tested against
the result of the other; if they are consistent, then the measure is said to be reliable. The
problem with internal consistency is that the reliability of this method is completely
dependent on the way the data is divided up or manipulated. Sometimes it so happens that
different splits give different results. To overcome such problems with split halves, many
researchers adopt a technique called as Cronbach Alpha which needs the scale items to be
at equal intervals. In case of difficulty in obtaining the data at equal intervals of time then
an alternate method called KR-20 (Kuder Richardson Formula 20) is used to calculate
how consistent subject responses are among the questions on an instrument. Items on the
instrument must be dichotomously scored (0 for incorrect and 1 for correct). All items are
compared with each other, rather than half of the items with the other half of the items. It
can be shown mathematically that the Kuder-Richardson reliability coefficient is actually
the mean of all split-half coefficients.
Validity
The ability of a scale or a measuring instrument to measure what it is intended to measure
can be termed as the validity of the measurement. For instance, students may complain
about the validity of an exam, stating that it did not measure their understanding of the
topic, but only their memorizing ability. Another example may be of a researcher who
tries to measure the morale of employees based on their absenteeism alone; in this case
too, the validity of the research may be questioned, as absenteeism cannot be purely
attributed to low morale, but also to other conditions like prolonged illness, family
reasons and so on. Validity can be measured through several methods like face validity,
content validity, criterion-related validity and construct validity.
Face validity
Face validity refers to the collective agreement of the experts and researchers on the
validity of the measurement scale. However, this form of validity is considered the
weakest form of validity. Here, experts determine whether the scale is measuring what it
is expected to measure or not.
Content validity
Content validity refers to the adequacy in the selection of relevant variables for
measurement. The scale that is selected should have the required number of variables for
measurement. For instance, if the state education department wants to measure whether
all the schools in the city have adequate facilities, and for measuring this, it develops a
scale to measure the attributes like the attractiveness of schools names, the frequency of
old students meets, the different varieties of eatables that are prepared in the school
canteen and so on. Here, it is clear that these variables considered for measurement do not
possess any content validity as they will not serve the purpose of the research. The scale
should instead be developed to measure aspects such as the number of classrooms, the
number of qualified teachers on roll, the capacity of the playground and so on. It is often
difficult to identify and include all the relevant variables that need to be studied for any
research process.
Criterion-related validity
The criterion related validity refers to the degree to which a measurement instrument can
analyze a variable that is said to have a criterion. If a new measure is developed, one has
to ensure that it correlates with other measures of the same construct. For instance, length
of an object can be measured with the help of tape measure, calipers, odometers and also
with a ruler and if a new technique of measure is developed then one has to ensure that
this new measure correlates with other measures of length. If a researcher wants to
establish criterion validity for a new measure for payment of wages, then he may want to
ensure that this measure correlates with other traditional measures of wage payment such
as total number of days worked.
Criterion validity may be categorized as predictive validity and concurrent validity.
Predictive validity is the extent to which a future level of a criterion variable can be
predicted by a current measurement on a scale. A scale for measuring the future
occupancy of an apartment complex for example may use this scale. A builder may give
preference to only those repairs that may attract new tenants in the future rather than
focusing on all the areas that need repair. Concurrent validity is related with the
relationship between the predictor variable and the criterion variable. Both the predictor
variable and the criterion variable are evaluated at the same point in time.
Construct validity
Construct validity refers to the degree to which a measurement instrument represents and
logically connects through the underlying theory. Construct validity, although it is not
directly addressed by the researcher, is extremely important. It assesses the underlying
aspects relating to behavior; it measures why a person behaved in a certain way rather
than how he has behaved. For instance, whether a particular product was purchased by a
consumer, is not the consideration, but why he has/has not purchased the product is taken
into account to judge construct validity. This helps to remove any extraneous factors that
may lead to incorrect research conclusions. For example, for a particular product, price
may not be the factor that affects a person deciding whether to buy it. If this product is
used in the measurement of a general relationship of price and quantity demanded, it does
not have construct validity, as it does not connect with the underlying theory.
There are two statistical methods for analyzing construct validity – convergent validity
and discriminant validity. Convergent validity is the extent of correlation among different
measures that are intended to measure the same concept. Discriminant validity denotes
the lack of or low correlation among the constructs that are supposed to be different.
Consider a multi-item scale that is being developed to measure the tendency to stay in
low cost hotels. This tendency has four personality variables; high level of self-
confidence, low need for status, low need for distinctiveness, and high level of
adaptability. Additionally, this tendency to stay in low cost hotels is not related to brand
loyalty or high level aggressiveness. The scale can be said to have construct, if it
correlates highly with other measures of tendency to stay in low cost hotels such as
reported hotels patronized and social class (convergent validity). Has a low correlation
with the unrelated constructs of brand loyalty and a high level of aggressiveness
(discriminant validity).
Sensitivity
Sensitivity refers to an instrument’s ability to accurately measure variability in stimuli or
responses. Sensitivity is not high in instrument’s involving ‘Agree’ or ‘disagree’ types of
response. When there is a need to be more sensitive to subtle changes, the instrument is
altered appropriately. For example strongly agree, mildly agree, mildly disagree, strongly
disagree, none of the above, are categories whose inclusion increases the scale’s
sensitivity.
Generalizability
Generalizability refers to the amount of flexibility in interpreting the data in different
research designs. The generalizability of a multiple item scale can be analyzed by its
ability to collect data from a wide variety of respondents and with a reasonable flexibility
to interpret such data.
Relevance
Relevance, as the name itself suggests, refers to the appropriateness of using a particular
scale for measuring a variable. It can be represented as,
Relevance = reliability x validity
If correlation coefficient is used to analyze both reliability and validity, then the scale can
have relevance from 0 to 1, where 0 is the low or no relevance level to 1 which is the high
relevance level. Here if either of reliability or validity is low then the scale will have little
relevance.
SOURCES OF MEASUREMENT PROBLEMS
When conducting a study, a researcher has to analyze the accuracy of the information that
has been obtained, as several types of research errors can come in during a study. Some of
the major research errors are discussed below.
Respondent Associated Errors
A majority of research studies rely on eliciting information from respondents. If the
researchers are able to obtain the cooperation of respondents and elicit truthful responses
from them, the survey can easily achieve its targets. However, two respondent associated
errors arise when researchers do not obtain the information as stated above. These
respondent errors are non-response error and response bias.
Non-response errors
Non-response errors arise when the survey does not include one or more pieces of
information from a unit that has to be part of the study. The research results will have
some bias to the extent that those not responding are different from those who respond.
Non-response errors include failure to respond completely or even failure to respond to
one or more questions of the surveyor. Unit non-response occurs when a person or a
household that exists in the data set does not respond. Item non-response is one where a
person selectively responds to only certain questions of the survey and will not respond to
one or more questions of the survey. The reasons for not responding to some questions
may be: lack of knowledge or it may be that the respondent doesn’t want to answer. Non-
respondent error may become an important source of bias in the result of the survey if a
large number of the potential respondents do not respond, and if the non-respondents are
significantly different from the respondents on some of the characteristics that are
important for the study.
Response bias
When the respondents consciously or unconsciously misrepresent the truth then it
amounts to response bias. Sometimes respondents deliberately mislead researchers by
giving in false answers so as not to reveal their ignorance or to avoid embarrassment and
so on.
Instrument Associated Errors
Instrument associated errors can surface due to poor questionnaire design, improper
selection of samples, etc. Even a simple thing like lack of adequate space in the
questionnaire for registering the answers of the respondent can result in errors of this sort.
Another type of instrument errors occurs if the questionnaire is complex or ambiguous as
this can result in a lot of confusion for the respondent. If the questions in the
questionnaire use complicated words and sentences, they will inadvertently lead to errors
due to the misinterpretation of such questions by the respondents.
Situational Errors
Plenty of errors arise due to the situational factors. The respondent may not provide
proper responses if a third person is present during the interview, or sometimes the third
person might himself participate in the interview process without any invitation leading to
inappropriate responses. Other factors such as the location of the interview also play a
crucial part; for instance, if the researcher is conducting intercept interviews in public
places then the respondents may not respond as properly as they would if they were
interviewed in their homes. If the researcher does not assure the respondent that the data
provided will be kept confidential, the respondent may not part with certain information
that may be crucial for the research.
Measurer as Error Source
The measurer may be a source of error because of some of the common mistakes
committed by interviewers. During the process of the interview the interviewer might
encourage or discourage the respondent while giving responses to certain questions,
through body language and gestures - smiles to encourage certain responses, frowning to
discourage certain responses and so on. After the collection of the data, the interviewer
might reword or rephrase the responses which may lead to errors. Failing to record the
full response of the respondent, inappropriate coding and tabulations, and application of
irrelevant statistical tools for measurement will also lead to errors.
ATTITUDE MEASUREMENT
Attitudes have been understood as learned predispositions that project a positive or
negative behavior consistently towards various objects of the world. Attitudes are
generally formed on a permanent basis and they develop as a combination of several
interrelated beliefs. People in society have different attitudes towards different aspects of
the world. Attitudes play a major role in a person’s good or bad behavior, based on the
standards set by society. A person may have a negative attitude towards society and go
against its customs and beliefs. On the other hand, a person with a positive attitude will
not go against standards set by the society. There will be some people, who take the mid-
path, conforming in some things and rebelling in others. Further, attitudes are not just
confined to one aspect but a predisposition towards several features, a world-view as
such. To give a small example, a person may have a positive attitude towards a particular
hotel, based on its clean and hygienic environment and the tasty food it serves.
In any company or industry, it is crucial to measure customer attitudes to understand their
behavior towards products and services. Although it is difficult to measure attitudes
qualitatively, attempts have been made to do this with a certain degree of accuracy.
Having an accurate measure of consumer attitudes towards various business situations
and marketing mix variables saves companies from committing huge sums to business
activities that do not add value to customers or stakeholders.
In this chapter, we will discuss the components of attitude, the definition of scaling and
different types of single item scales and multi-item scales. The chapter concludes with a
discussion on considerations for selecting an appropriate scale.
COMPONENTS OF ATTITUDE
Attitude has three components – the cognitive, the affective and the behavioural
components. If a person says that he loves Britannia biscuits because they are tastier and
will always eat them, the statement comprises all these three components of an attitude.
Firms usually study components of attitudes in consumers to improve their marketing
communications to attract customers and to develop a competitive advantage. Let us now
elaborate these components.
Cognitive Component
The knowledge and perceptions acquired by a combination of direct experience with the
attitude object and related information from various sources are based on cognition of an
individual. This is termed cognitive component. Such knowledge in a person commonly
leads to a belief that a particular type of behavior leads to a particular outcome. The
cognitive component of attitude consists of beliefs, opinions, knowledge and information
held by a person regarding an object or an issue. The knowledge comprises awareness
about the existence of the object, belief about its different characteristics and features of
the product, apart from the relative importance the person gives to each characteristic.
Let us give an illustration. Anand, a businessman, is planning to travel to Delhi from
Hyderabad by air. He remembers the names of several carriers, which he can use, such as
Indian Airlines, Jet Airways, Air Deccan and so on. This is his knowledge about the
existence of an object. This knowledge is not just confined to awareness. Anand will have
certain beliefs about each airline based on his personal experience or through experiences
of relatives and friends, knowledge gained through advertisements, books, magazines etc.
This constitutes beliefs about different features of the object. Anand may also feel that the
service in Indian Airlines is superior to Air Deccan. This is called placing relative
importance. Anand’s beliefs may not be entirely accurate, but to him, they are facts. Once
these positive beliefs increase, they give rise to a favourable cognitive component
towards an object. Marketers use various marketing mix variables to attract the customer
and overtime try to nurture positive beliefs about their products and services in the minds
of customers.
Affective Component
A person’s emotions or feelings towards an object comprise the affective component of
an attitude. Researchers treat such feelings of individuals as their favourable or
unfavourable assessment of an object. Such feelings, which are called the affective
component of attitude, may transform themselves into emotionally charged states such as
anger, happiness, shame, distress, guilt and so on. These types of experiences will
influence one’s perception of an object and that person’s later behavior. For instance, a
woman might say that she loves shopping at Lifestyle and that Shoppers’ Stop does not
have as wide a range of apparel as Lifestyle does. The woman’s overall emotional
feelings form the affective component. It is important to note that two persons may share
a cognitive component, but when it comes to the affective component, one may have a
positive affective component and the other a negative affective component towards the
same object.
Behavioural Component
The behavioural component comprises a person’s future actions and intentions. It is
concerned with the likelihood or tendency that an individual will behave in a particular
fashion with regard to an attitude object. Going back to our previous example, if Anand
wants to fly Indian Airlines in the future too or the lady wants to buy clothes from
Lifestyle in her next shopping excursion too, these are the behavioral components of
attitude. These intentions, however, have limited timeframes. Sometimes suggestions
become a behavioral component. For instance, when an individual suggests that a friend
travel by Indian Airlines, it is a behavioral component of that individuals’ attitude.
RELATIONSHIP BETWEEN ATTITUDES AND BEHAVIOR
It is difficult to analyze the relationship between attitudes and behavior. Analyzing the
future behavior of a group of people is relatively easier than analysis for a single
individual. Researchers have discovered that there are certain critical aspects governing
the attitudes and behavior of consumers. They are:
1. A product or service usage will be maximum if the person develops
a positive attitude towards it. The converse is also true.
2. Attitudes of consumers towards products that they have never tried
will be neutral.
3. When attitudes are developed based on actual trial and experience
of a product, attitudes predict behavior effectively. On the other
hand, when attitudes are based on advertising, consistency in
attitude and behavior is considerably reduced.
CHANGING ATTITUDES
Changing customer attitudes, and changing them positively towards a company and its
products, is the most important activity of businesses across the world today. Whenever
sales of a product fall or market share declines, it becomes imperative for marketers to
identify ways and means to overcome the downturn. Changing attitudes of stakeholders
becomes the top priority in company’s development efforts. Companies can attempt to
change the attitudes of customers (an important stakeholder) towards a product in three
ways: •Altering existing beliefs about a product
•Changing attitudes by changing the importance of beliefs
•Adding new beliefs
Altering Existing Beliefs about a Product
A marketer’s fundamental responsibility here is to convert the neutral or negative belief
that a customer holds about the product into a positive belief. For this, the marketer may
attempt to change consumer perceptions about the product or service. Several tactics can
be used. For instance petrol stations have long been viewed as dusty, poorly lit places
where weary petrol pump service personnel dressed shabbily serve. This perception or
belief has been entirely changed by BPCL, which, by branding petroleum products and
developing petrol stations into shopping malls (like in Western countries) has changed
existing beliefs about petrol stations. BPCL’s strategy is discussed in exhibit 4.4.
However, marketers need to understand that customer beliefs cannot be changed by
advertising alone. Any change achieved cannot be sustained if there is no tangible quality
in the product to support advertising claims. Second, marketers trying to change
consumer beliefs should ensure that the change is incremental rather than drastic. For
example, an aggressive advertising campaign aimed at changing traditional beliefs of a
community may meet with customer resistance. Therefore, the change process should be
slow and preferably take the customer through all the stages in the learning process.
Changing Attitudes by Changing the Importance of Beliefs
Another strategy is to change customer attitudes by changing the importance of beliefs
that a customer holds about a particular product feature. For example, when Kelloggs
entered India, it faced a lot of problems initially to sell its products, as a consequence of
which it tried to change the people’s attitude towards its products. This resulted in
significantly improving its sales. Exhibit 4.5 discusses the repositioning strategy of
Kelloggs’ for changing customers’ attitude towards the product.
Adding New Beliefs
An altogether different strategy adopted by marketers for changing customer attitudes is
to develop new beliefs in customers about products. Adding new beliefs is an important
job for marketers. Once such new beliefs are clearly communicated to consumers, there is
a likelihood of higher sales since customers who previously did not bother to buy a
product may now choose to buy it. For instance, traditionally salt has been promoted on
the taste attribute. Tata promoted salt with iodine content as essential for health (Iodine
Exhibit 4.4
Bharat Petroleum’s Efforts to Change Face
Petrol pumps in India have come a long way from being dusty, poorly lit places manned by shabbily
clothed and indifferent personnel, to the shopping malls of the early 21st century. Bharat Petroleum
Corporation Ltd. (BPCL), a leading player in the Indian petroleum industry, has got wide acclaim
for having brought about this change in the fuel retailing business.
With the deregulation of the oil industry in April 2002, Indian players realized the need to become
more customer focussed. BPCL’s pioneering efforts to create brand awareness for its products were
thus a welcome. BPCL’s first foray into petrol pump retailing was through Bharat Shell Ltd. (Shell).
The store, offering eatables, soft drinks, stationery, newspapers, magazines, frozen foods, light
bulbs, audio- cassettes and CDs, came as a pleasant surprise to Indian consumers.
By July 1999, 35 of BPCL’s retail outlets across the country had ‘Bazaar’ stores running
successfully. In October 2000, BPCL introduced another revolutionary concept by launching a
McDonald’s fast food outlet at a petrol pump near Mathura (UP) on the Delhi-Agra highway. The
4,000 sq.ft., 180-seat outlet was set up at a cost of Rs 40 million. McDonald’s paid a fixed rent,
besides a percentage of its sales to BPCL for using the facility. The outlet was expected to pull in
foreign and domestic tourists headed to and from Agra, besides residents of surrounding areas.
In January 2001, BPCL further upgraded the ‘Bazaar’ stores and, a month later, launched the ‘In &
Out’ stores at around 40 outlets in Bangalore, Mumbai, Delhi, Kolkata and Chennai. To offer
enhanced customer service, BPCL tied up with various companies from different industries. These
included fast food, photography, music, financial services, ISPs, e-commerce portals, document
centers, ticketing, greeting cards, ATMs, and courier services. All these efforts have helped BPCL
significantly in changing customer beliefs.
helps the growth, development and functioning of the thyroid gland), thus completely
changing common beliefs about salt.
ASSOCIATION BETWEEN MEASUREMENT OF BELIEFS AND SITUATION
It is generally found that there is a lesser relationship between researchers’ measurements
and the actual prevailing situation. In other words, the match between what the researcher
finds and what actually happens is low owing to a number of reasons. For projecting a
favourable attitude towards a product, a respondent should first have a felt need for the
product. For instance, a respondent should need a car; only then might he or she display a
positive or negative attitude towards a particular brand of car. A person may have a
favourable attitude towards a car, but this is not sufficient and has to be backed with the
ability to purchase the product. Often, certain parameters of the purchase process are
neglected while measuring attitudes. For example, a person might have decided to buy a
product and may have decided on a particular brand, but on the day of purchase, the
person may be lured away by a competitor’s better promotional campaign. Or the person
may decide to buy a brand that is cheaper and use the money saved for some other
purpose. Sometimes the person’s attitude may change or be influenced by other family
Exhibit 4.5
Kellogg’s Repositioning Strategy
Kellogg’s started its India operations in September 1995. Its initial offerings were cornflakes,
wheat flakes and Basmati rice flakes. Later, it introduced the breakfast cereals Chocos and
Frosties. Despite good quality products and support from the parent company’s technical,
managerial and financial resources, Kellogg’s products failed in the Indian market. Kellogg
realized that it was a tough task to get Indian consumers to accept its products. Kellogg banked
heavily on the quality of its crispy flakes. But, pouring hot milk on the flakes made them soggy. As
Indians boil their milk before drinking, unlike in the West, the milk was usually warm or
lukewarm. Americans often use cold milk for cereals, retaining their crispness. Indians also like to
add sugar to their milk. When Kellogg flakes were put in hot milk, they became soggy and did not
taste good. In cold milk, it was not sweet enough for Indian tastes because the sugar did not
dissolve easily in cold milk. The rice and wheat versions too did not do well. In fact, some
consumers even referred to the rice flakes as rice corn flakes.
Kellogg’s began working towards better positioning for its products. The company’s research
showed that the average Indian consumer did not give much importance to the level of iron and
vitamin intake, and looked at quantity, rather than quality, of food consumed. Kellogg’s worked
towards changing the positioning of Chocos and Frosties – which had not been placed on the health
platform but had been projected instead as ‘fun-filled’ brands.
In 1995, Kellogg had a 53% share of the Rs 150 million breakfast cereal market, which had been
growing at 4-5% per annum till then. By 2000, the market size was Rs 600 million, and Kellogg’s
share increased to 65%. Analysts claimed that Kellogg’ entry was responsible for this growth. The
company’s improved prospects were clearly attributed to the shift in positioning, increased
promotion and an enhanced media budget. The effort to develop products specifically for the
Indian market helped Kellogg make significant inroads into the Indian market.
members while making a purchase. These are some issues that will reduce the intensity of
association between measurement of beliefs and the actual situation.
ATTITUDE SCALES
We have understood how important it is for marketers to measure attitudes. Marketers try
to understand these attitudes and influence them to gain an advantage in the market.
Measuring attitudes is a highly difficult process and unlike measurement scales in the
physical sciences like measuring height, weight etc, measurement scales for attitude are
less precise. In the following sections, we will discuss the different attitude scales that can
be used to measure attitudes as precisely as possible.
Exhibit 5.4
Types of Attitude Scales
Attitude Scales
Single Items Scales Continuous Scales Multi Item Scales
Itemized
Category
Scales
Rank
Order
Scales
Comparative
Scales
Constant
Sum Scales
Q Sort
Scales
Paired
Comparison
Scales
Pictorial
Scales
Likert
Scales
Semantic
Differential
Scales
Thurstone
Scales
Associative
Scales
Stapel
Scales
Source: ICMR Centre for Management Research
Exhibit 4.6
Type of Attitude Scales
DEFINITION OF ATTITUDE SCALE AND SCALING
An Attitude Scale is a set of items (questions or statements) that probe a single aspect of
human behavior, attitudes, or feelings. Scaling is the process of measuring quantitative
aspects of subjective or abstract concepts. It is a method to assign numbers or symbols to
some attributes of an object. Scaling involves developing a continuum, based on which
measured objects are located. For instance, if we want to measure the satisfaction levels
of customers using a product, we might develop a six-point scale where respondents can
choose 1 if they are least satisfied, 3 if they are moderately satisfied and 6 if they are
highly satisfied. Scales are uni-dimensional or multi-dimensional. The former, as the
name suggests, are used to measure one particular attribute of an object. For instance,
while using a scale for measuring customer satisfaction levels, although several items will
contribute to the measurement, these will be combined to give the final rating on
customer satisfaction. A multi-dimensional scale is used to measure several attributes of
an object, for example work environment, interpersonal relations, and behaviour of top
management and compensation procedures. These will all be measured to determine the
motivational levels of employees in an organization.
TYPES OF ATTITUDE SCALES
There are two major types of scales used to measure the attitudes of respondents. They
are single item scales and multi- item scales. The different types are shown in exhibit 4.6
Single Item Scales
Single item scales are those with only one item to measure. Itemized category scales are
most commonly used under single item scales. Besides, itemized category scales, there
are several other scales such as comparative scales, rank order scales and so on, which are
used for attitude measurement. We will discuss each of these in the following sections.
Itemized category scales
Itemized category scales are those in which respondents have to select an answer from a
limited number of ordered categories. Respondents are given the scale that contains a
number or a brief description about a particular category. These categories are ordered in
terms of position of the scale and respondents have to select one category that they feel
best describes the object. It is easy to develop itemized category scales. Exhibit 4.7 gives
an itemized category scale where a hotel customer is asked to indicate the level of
satisfaction for service provided.
Rank – order scales
Rank order scales are comparative scales, where the respondent is asked to rate an item in
comparison with another item or a group of items against each other based on a common
criterion. For instance, a respondent may be asked to rank three motorcycle brands on
attributes such as cost, mileage, style, pick-up and so on. Although it is easy to develop a
rank-order scale, it has some disadvantages. It is very difficult to include every possible
brand or attribute on a scale. Therefore, a respondent may rate a brand as number one, but
it might not be his first choice as the brand he prefers may not have been included in the
list at all. Sometimes, respondents may feel that the attributes used to construct the scale
are not relevant to judging the subject under research. One major shortcoming is that the
researcher will not have any clue about why the respondent has given a particular rating
for items listed on the scale. Exhibit 4.8 shows a rank order scale for ranking different
brands on motorcycle on specified attributes.
Q-sort scales
When the number of objects or characteristics to be rated is very large in number, it
becomes difficult and tedious for respondents to rank order. In such cases, Q-sort scaling
is used. Here, respondents are asked to sort out various characteristics or objects that are
being compared into various groups so that the distribution of the number of objects or
characteristics in each group follows a normal pattern. For instance, let us consider that
Highly Satisfied Considerably Satisfied Reasonably Satisfied Unsatisfied Highly Unsatisfied
Exhibit 5.5
Itemized Category Scale
Given below is a itemized category scale ranging from highly satisfied to highly unsatisfied.
Please select one of the following options based on your satisfaction levels of the hotel service.
Adapted from Donald S. Tull and Del I. Hawkins, “Marketing Research – Measurement and
Method,” Prentice Hall of India, Sixth Edition, New Delhi, 1998, p380
Exhibit 4.7
Exhibit 4.8
Rank Order Scale used for Analyzing Motorcycles
Please rank the following brands of motorcycles with 1 being the brand that best meets the
characteristic being evaluated and 7 being the worst on the characteristic being evaluated. Let us
now start rating these brands basing on their affordability, first. Which brand has the highest
affordability? Which is second? (Record the answers below).
Brand of MotorcycleAffordable CostHigh MileageStylishGreat PickupHero HondaTVSBajaj
Adapted from Donald S. Tull and Del I. Hawkins, “Marketing Research – Measurement and Method,”
Prentice Hall of India, Sixth Edition, New Delhi, 1998, p384
the designing team of a toy manufacturing company has come out with hundreds of new
product ideas with slight variations. The research team’s task is to find out from
customers which combination of features is the best and will generate maximum sales. To
accomplish this, Q-sort scaling is the best method. The procedure followed is:
Respondents are given a set of cards, usually varying from 80 to 120 cards, containing
different categories of items to be selected from. For instance, if respondents have to rate
100 different products according to their tastes and preferences, each respondent will be
given about 100 cards containing a product and its features. Respondents are then asked
to segregate the cards into 10 stacks so that the 1st stack contains a set of cards that are
highly preferred by respondents. The 10th stack will contains a set of cards that are least
preferred by them. The individual stacks in between (2nd to 9th) should be prepared by
the respondent in such a way that they range from higher preference to lower preference.
Once the stacks are ready, the cards in each stack should be arranged in the respondents’
order of preference, based on criteria like features of a product, communication processes
and customer service. This gives the best and the worst product in each stack. The
disadvantage of this process is that it asks a lot of time and effort on the part of
respondents.
Comparative scales
In the itemized category scale, respondents select a category that they feel best describes
a product. The problem here is that respondents may select a category based on their own
perceptions. For instance respondent A might select a category based on his or her view
of an ideal brand, respondent B may pick a brand based on knowledge of an existing
brand and respondent C might choose based on some other criteria. Ultimately, the
selection process lacks uniformity. To overcome this, comparative scales have been
developed, where the researcher provides a point of comparison for respondents to
provide answers. Therefore, all respondents will have a uniform point of comparison for
selecting answers. For instance, rather than asking a person to evaluate the quality of
sweets in one sweet shop in Hyderabad, the respondents will be asked to evaluate the
quality of that sweet shop in comparison to another sweet shop in Hyderabad. Exhibit 4.9
gives a comparative rating scale.
Exhibit 5.7
Comparative Scales
Excellent Very Good Good Both are same Poor Very poor
Given below is the scale ranging from excellent to very poor. If you were asked to
rate the sweet shop ‘X’ in comparison to sweet shop ‘Y’ in Hyderabad. Which one
will you choose. If you choose excellent then select the first option.
Exhibit 4.9
Paired comparison scales
In paired comparison scales, respondents are asked to select one of two items in a pair
based on pre-set criteria. As each item is compared with all other items, the number of
times an item is selected from a pair gives its rank. The higher the number, the better is
the rank. In this method, the shortcoming of rank order scaling is overcome, as it is easy
for respondents to select one item from two rather than ranking a long list of items.
Another advantage is that the problem of order bias is eliminated as no set pattern is
followed while providing respondents the pairs. A typical paired comparison scale for
toothpaste is shown in Exhibit 4.10.
Constant sum scales
In constant sum scales, respondents are asked to divide a given number of points, usually
100, among two or more attributes based on the importance they attach to each attribute.
These scales are often used in place of paired comparison scales to eliminate the long lists
in paired comparisons. Here, respondents have to rate an item in relation with all other
items. Ranking for each item is based on the points assigned by the respondent to the
items. The disadvantage of this approach is that the researcher is limited to giving 10
items for the respondent as a higher number of items will confuse the respondent. Exhibit
4.11 shows the constant sum scale where respondents are asked to rate 10 characteristics
of a supermarket for a total sum of 100 points.
Exhibit 4 .10
Paired Comparison Scale for a Toothpaste
Please select one item each from the following pairs that is most important to you
for selecting a toothpaste.
a. Fights decay b. Affordable
a. Affordable b. Longer germ protection
a. Longer germ protection b. Fights decay
Pictorial scales
Here, the different types of scales are represented pictorially. The respondents are asked
to rate a concept or statement based on their intensity of agreement or disagreement, on a
pictorial scale. Pictorial scales have to be developed carefully so that respondents will not
have problems selecting appropriate responses. These scales are generally used for
respondents who cannot analyze complex scales, such as young children or illiterates.
Typical pictorial scales are a thermometer scale or a scale depicting a smiling face.
Exhibit 4.12 shows the smiling face scale for measuring the effectiveness of an
advertisement campaign for a chocolate.
Continuous scales
Exhibit 4.11
Constant Sum Scale used for a Supermarket
Given below are the ten characteristics of a supermarket. Please give each characteristic some
point(s) based on your assessment, so that the total points add up to 100. The higher number of
points allocated to a particular characteristic, the higher its importance to you. However, if any
particular characteristic is unimportant to you, then you need not assign any points to it. However, it
is essential that all points given add up to 100.
Characteristics of a Supermarket Number of Points
The supermarket is conveniently located _______
The supermarket has enough range of products _______
All the items in the store are conveniently located _______
Sales persons are cooperative _______
Aisle space is comfortable _______
Prices are very much affordable _______
The ambiance in the store is pleasing _______
Soft music played in the store is entertaining _______
Billing counters are sufficient _______
Parking facility is adequate _______
________
100 Points
Exhibit 5.10
Smiling Face Scale
(Verbal Instructions)
Face 1 should be selected, if you did not like the chocolate ad at all and face 5 should
be selected if you liked it very much, now how did you like it?
Adapted from J. P. Neelankavil, J. V. O’Brien, and R. Tashjian, “Techniques to
obtain market-related information from very young children,” Journal of advertising
research June/July 1985.
Exhibit 4.12
Continuous scales are those where respondents are asked to rate items being studied by
marking at an appropriate place on a line drawn from one extreme of the scale to the
other. These scales are rarely used in marketing research as they do not give accurate
results and the scoring process is complicated. This scale’s only advantage is that it is
very easy to develop. For instance, if a fast food outlet such as Pizza Hut wants to find out
whether customers are satisfied with its overall service, then a continuous scale can be
developed as shown in Exhibit 4.13
Multi- Item Scales
Let us move now to multi-item scales. These scales are used when it is difficult to
measure people’s attitudes based only on one attribute. For instance, to measure
respondents’ attitudes towards the Indian Railways, if you ask them only whether they are
satisfied with Indian Railways or not, it will not suffice. People may say that they are
satisfied on an overall basis, but there might be number of factors that they find
unsatisfactory. Thus, it is impossible to capture the complete picture with one overall
attitude-scale question. To measure individual attributes, a number of scales have been
developed that can measure a respondent’s attitude on several issues on a scale ranging
from most favourable to least favourable. The Semantic, Likert, Thurstone and
differential scales are some examples that follow such measurement techniques.
Developing multi-item scale involves certain crucial steps that have been discussed in
Exhibit 4.14.
Exhibit 4.13
Continuous Rating Scale
Given below is a continuous scale ranging from 0 to 100 points. You have to indicate a point that
best describes how you rate the overall service. If you rate it the best then it would be 100.
How would you rate the overall service of Pizza hut?
Best -------------------I-------------------------------------------------------------------------------- Worst
100 90 80 70 60 50 40 30 20 0
Exhibit 5.12
Developing a Multi-item Scale
Initial Development of a theory
Generating a number of Items
Selecting a final set of Items with the Help of
experts
Adding some items for improving the validity
of the scale
Administering the items to an initial set of
respondents
Analyzing the Items
Optimizing the Final Scale Length
Identifying the appropriate attitudinal scale to
be used
Exhibit 4.14
Semantic differential scales
Semantic differential scales are used to describe a set of beliefs that underline a person’s
attitude towards an organization, product or brand. This scale is based on the principle
that individuals think dichotomously or in terms of polar opposites such as reliable-
unreliable, modern-old fashioned, cold-warm.
The respondents are asked to rate an attitude object on a set of itemized, seven- point
rating scale, bounded by bipolar phrases or adjectives. The initial process of developing a
semantic differential scale starts with determining the object to be rated. Once this object
is determined, respondents are asked to express their attitudes towards the object, using
the dichotomous pair on a scale. Such points are then plotted on a graph. This is the most
efficient technique for determining the strengths and shortcomings of a product/service or
a company in the market.
Exhibit 5.13
A Semantic Differential Scale for Measuring the Attitudes of
Respondents for a Newspaper.
1 2 3 4 5 6 7
Contemporary Old Fashioned
Comprehensive
News Coverage
News coverage is very
limited
Interesting
Supplements Uninteresting Supplements
Balanced coverage of
news
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . . Unbalanced coverage of news
Quality of language is
high
. . . . . . . Quality of language is
poor
Too much
international
coverage
. . . . . . . Too less international
coverage
Excellent Editorial Articles
. . . . . . . Worst Editorial Articles
Appealing to youth . . . . . . . Not at all appealing to
youth
Not inclined towards any
political party
. . . . . . . Inclined towards a political
party
Given below is a semantic differential scale, where a respondent has chosen his
options on a 7 point scale based on 9 bi-polar categories.
Exhibit 5.15
While designing the scale, care should be taken that all negative or positive adjectives or
phrases do not appear on one side. This avoids a person from picking either only positive
or negative phrases.
Another problem that should be addressed while developing a seven- point semantic scale
is the response of 4. If the respondent selects 4 for all items, then it becomes neutral
without indicating any specific direction. Exhibit 4.15 represents the semantic differential
scale administered for measuring the attitudes of respondents towards a newspaper. It can
also be used for comparing the products with that of the competition. Consider four
brands of cars being rated on the same scales, as shown in the Exhibit 4.16 below.
Below given is the semantic differential scale rated by a responding by comparing 4
brands of cars.
Exhibit 5.14
Semantic Differential Scale for Comparing Four Brands of Cars
Mitsubishi (L)
Lancer
Hyundai (E)
Elantra
Skoda
Octavia (O)
Honda (C)
City
Fast EL O C Slow
Large L E O C Small
Plain E L O C Stylish
Inexpensive C L E O Expensive
Adapted from Donald S. Tull and Del I Hawkins. “Marketing Research-Measurement
and Method. Prentice-Hall of India Sixth Edition. New Delhi, 1998 pg 390.
Exhibit 4.16
Stapel scales
A Stapel scale is an attitude measure that places a single adjective or an attribute
describing an object in the centre of an even number of numerical values. In general,
staple scales are constructed on a scale of 10 ranging from –5 to +5, without a neutral
point (zero). The respondent is asked to rate attributes on this scale.
Stapel scales are similar to semantic differential scales but here there is only one pole
(single adjective) rather than bipolar adjectives. This scale is useful for researchers to
understand the positive and negative intensity of attitudes of respondents. The numeric
value assigned to an adjective shows how well it describes the object. The higher the
positive value, the better it describes the object.
One big disadvantage is that the respondent might select all attributes on a positive or
negative range. A Stapel scale that is designed to measure the attitude of passengers
towards an airline is shown in Exhibit 4.17.
Likert scales
Likert scales consist of a series of statements where the respondent provides answers in
the form of degree of agreement or disagreement. This expresses attitude towards the
concept under study. The respondent selects a numerical score for each statement to
indicate the degree of agreement or otherwise. Each such score is finally added up to
measure the respondent’s attitude. The various steps involved in developing a Likert scale
are given below. •Identify the concept that needs to be measured
•Develop a series of statements (say, 100) that articulate
respondents’ feelings towards the concept
•Every test item is categorized by the respondent as generally
favourable or unfavourable based on the attitude that needs to
be measured
•A pre-test is conducted to measure the intensity of the
favourable or unfavourable attitude of respondents towards
each test item. The scale would have intensity descriptors
like, highly favourable, favourable, neutral, unfavourable,
and highly unfavourable. These responses are given a
numerical weight.
The total attitude score is represented by the algebraic sum of the weights of the items. To
make the measuring process uniform, the weights are consistently assigned. For instance,
if 5 were assigned to reflect strong agreement with a favourable situation, then 5 should
be assigned to show strong disagreement with an unfavourable situation too.
After the results have been obtained, the researcher selects items that reveal a clear
discrimination between high and low total scorers by identifying the highest and lowest
quartiles based on total scores. Subsequently, mean differences are computed for these
high and low groups.
Finally, a set of items is chosen that represent the greatest difference between the highest
and the lowest mean values.
Likert scales are very popular among researchers for measuring the attitudes of people.
But, in practical situations, commercial researchers are more concerned with finding the
respondents attitudes towards individual components, rather than overall positive or
Exhibit 5.15
Stapel Scale for Measuring the Attitudes of Flight Passengers
+5
+4
+3
+2
+1
+5
+4
+3
+2
+1
+5
+4
+3
+2
+1
Friendly Cabin Crew
-1
-2
-3
-4
-5
-1
-2
-3
-4
-5
-1
-2
-3
-4
-5
Comfortable Interiors Accurate Timings
Below given is a staple scale designed to measure your attitude on three attributes.
Please circle one number from the following three columns that best describes your
attitude towards them.
Exhibit 4.17
negative attitudes of respondents. For, instance the manufacturer of a brand of shoes will
be more interested in finding out why people are not buying the brand rather than
respondents’ attitudes towards shoes in general. A typical Likert scale is discussed in the
exhibit 5.18
Thurstone scales
In Thurstone scales, researchers select a group of 80 to 100 items indicating the different
degrees of favourable attitude towards a concept under study. Once items are selected,
they are given to a group of judges, who are asked to categorize them according to how
much they favour or disfavour them. The judges are asked to treat intervals between
categories as equal and analyse each item without expressing their own attitudes. Once
the results are obtained, all those items that have a consensus from the judges are selected
and items where there was no consensus are eliminated. These results are then distributed
uniformly on a scale of favourability. This scale is then administered to a set of
respondents for measuring their attitude towards a particular concept. Although the
thurstone method is time consuming as it involves a two- stage procedure, it is easy to
administer. This method comes under criticism because the scale values are developed
based on the attitudes of the judges.
Profile analysis
Profile analysis is a process where two or more objects are rated by respondents on a
scale. Profile analysis can be considered as an application of the semantic differential
scale. Comparing different objects visually, based on different attributes, is possible in
this approach. The major disadvantage is that it is very difficult to interpret the profiles as
the number of objects increases. The profile analysis is used in exhibit 4.19 to compare
the three jeans brands.
CONSIDERATIONS IN SELECTING A SCALE
Researchers tend to use those scales that are easy to administer and develop. For instance,
a rank order scale can be quickly developed, while a semantic differential scale takes
Exhibit 5.18
Likert Scale
A Likert scale for evaluating the attitudes of customers, who have not used a vacuum cleaner, but
are aware of its existence, is given below.
Here are some statements that describe how customers might feel about vacuum cleaners.
Please indicate your agreement or disagreement. For each statement given below, please circle the
appropriate number to indicate whether you: 1- strongly agree, 2- agree, 3- neutral, 4- disagree and
5- strongly disagree
Strongly
Agree
Agree Neutral Disagree Strongly
Disagree
The product is costlier 1 2 3 4 5
I don’t find time to use a vacuum
cleaner
1 2 3 4 5
Advertising of the product is not
convincing enough
1 2 3 4 5
I have never used a vacuum cleaner 1 2 3 4 5
I am satisfied with the way I am
cleaning my house right now
1 2 3 4 5
Using a vacuum cleaner is
cumbersome
1 2 3 4 5
Competitor’s vacuum cleaner has
better features
1 2 3 4 5
The initial enthusiasm to use a
vacuum cleaner dies down at a later
stage and it is permanently kept in
the store room
1 2 3 4 5
The demonstration of the product
given by the salesperson is not
effective
1 2 3 4 5
longer time and is cumbersome. It is also important to consider the client’s requirements
before selecting a scale. The type of data that is needed also plays an important role in
selection. Some other factors that have to be considered while selecting a scale are:
balanced versus unbalanced, the numbers of categories to be used, whether the categories
should be odd or even and whether the choice be forced or unforced. Let us give some
details.
Balanced Versus Unbalanced Scales
A balanced scale is one, which has the same number of positive and negative categories,
while an unbalanced scale is weighted towards one or the other end. A balanced scale is
used in situations where a broad range of responses are expected. An unbalanced scale is
used where the results of preliminary research lean more towards one side of the scale
than the other. For instance, if preliminary research conducted to measure the
performance of a new solar powered car indicates that it is generally liked by the people,
then a scale with categories such as 1) excellent, 2) very good 3) good, 4) fair, 5) poor is
developed.
Number of Categories
The number of categories that have to be included in the scale should be decided based on
the research concept. If a scale is developed with very few categories (say, 2. good/bad)
then it does not reveal the respondents’ true attitudes. At the same time, if a scale contains
10 or more categories, the respondent might get confused and will not be able to
Exhibit 5.17
A Profile Analysis for Measuring the Attitudes of Respondents towards
Jeans Brands
1 2 3 4 5
Stylish . . . . . Old Fashioned
Affordable . . . . . Expensive
High Brand Valve . . .
Exhibit 5.4
Types of
Attitude Scales
Exhibit 5.4
Types of
Attitude Scales
. .
Low Brand Valve
High Range . . . . . Low Range
More Number of Outlets . . . . . Lesser Number of Outlets
Levis Lee Wrangler
Given below is a profile analysis of three jeans brands where a respondent has
indicated his attitude towards them on a five point scale based on five bi-polar
categories.
4.19
accurately assign items to the different categories. Therefore, it is always better to
develop a scale that contains between five and nine categories.
Odd or Even Number of Scale Categories
If a scale has an even number of categories, it means that it does not have a neutral point.
This restricts the respondents and forces them to choose a negative or a positive aspect of
a scale. So, respondents who are actually neutral cannot express this feeling. Adding a
neutral point in the scale helps respondents. Some feel that respondents can take the easy
way out by saying that they are neutral and need not concentrate on their inner and real
feelings. Deciding whether to have an odd number or an even number of categories on the
scale is dependent on the nature of research to be conducted. For instance, if a company
has recently changed the design and is attempting to study whether the customers have
liked it, it cannot expect the respondents to be highly emotional towards the package
design and therefore it needs to have neutral category (odd number). While if a company
only wants to know how strongly the consumers like or dislike a product then adding a
neutral category will not serve the purpose.
Forced Versus Unforced Choice
If respondents are given adequate choice for selecting a response, it becomes an unforced
choice. If they are not given any choice for selecting a response, then it becomes a forced
choice. An unforced decision can be either in the form of ‘Neutral’ (which a respondent
chooses if he is not inclined towards either object) or “Don’t know’ (which a respondent
can choose if he lacks the knowledge to answer the question). When these two categories
are included in the scale, it becomes an unforced choice, as the respondents do not have to
select a positive or negative opinion when they don’t have any opinion. When a neutral or
don’t know category is not included in the scale, it obviously becomes a forced choice.
Restricting the choice of respondents, although essential in some research studies, should
be avoided as a rule.
