Chapter 14
Attitude Measurement
AT-A-GLANCE
I. Attitudes as Hypothetical Constructs
A. Importance of measuring attitudes
II. Techniques for Measuring Attitudes
III. Attitude Rating Scales
A. Simple attitude scales
B. Category scales
C. Method of Summated Ratings: The Likert scale
Reverse recoding
Composite scales
D. Semantic Differential
E. Numerical scales
F. Stapel scale
G. Constant-sum scale
H. Graphic rating scales
I. Thurstone interval scale
IV. Measuring Behavioral Intention
A. Behavioral differential
V. Ranking
A. Paired comparisons
B. Sorting
VI. Other Methods of Attitude Measurement
VII. Selecting a Measurement Scale: Some Practical Decisions
A. Ranking, sorting, rating, or choice technique?
B. Monadic or comparative scale?
C. What type of category labels, if any?
D. How many scale categories or response positions?
E. Balanced or unbalanced rating scale?
F. Even or odd number of scale points?
G. Use a scale that forces a choice among predetermined options?
H. Single measure or an index measure?
LEARNING OUTCOMES
1. Describe how business researchers think of attitudes
2. Identify basic approaches to measuring attitudes
3. Discuss the use of rating scales for measuring attitudes
4. Represent a latent construct by constructing a summated scale
5. Summarize ways to measure attitudes with ranking and sorting techniques
6. Discuss major issues involved in the selection of a measurement scale.
CHAPTER VIGNETTE: Heat and Smoke—What Keeps Them Happy?
The history of steel factories and the challenges of the furnace workers is well documented. Despite the
great advances in technology and an ever-increasing focus on safety, furnace workers still face dangerous
work environments filled with heat and smoke. A team of researchers was asked to do an assessment of
furnace employee attitudes, with the goal of identifying what aspects of their work environment
contributed to their overall satisfaction. Using a survey questionnaire, a series of statements related to the
company’s benefits, supervisory relationships, and general work-related conditions was developed. While
safety and health benefits were important, it was positive and supportive relationship with their immediate
supervisor that really made the difference on their satisfaction. Managers can focus on hiring and training
supervisors who are oriented towards positively supporting their employees.
SURVEY THIS!
One popular way to assess respondent attitudes is to use a multiattribute model. Students are asked to
compute respondents’ attitudes toward working in a business career (i.e., each of the business disciplines).
RESEARCH SNAPSHOTS
Is It Positive Emotionality, or is it LOVE?
Love is a four letter word, and it is a hypothetical construct, which is a term psychologists use to
describe or explain consistent patterns of human behavior. Love, hate, thirst, learning, and
intelligence are all examples of hypothetical constructs. They are hypothetical in that they do not
exist as physical entities; therefore, they cannot be seen, heard, felt, or measured directly.
Students Ask—Are You Responsible?
Businesses today face an increasing need to be perceived as having in interest in social
responsibility. Results of a survey of 1,554 college students indicate that 41 percent consciously
prefer products and services from companies they perceive as having a social role (e.g., Toyota,
Burt’s Bees). The implication for business leaders is that perceptions of the company itself, and
not just its products, drive purchasing decision among this important demographic.
Measuring Website Usability
How good is a website? Usability is an important factor, and there are a variety of ways to assess
website usability. Subjective evaluations can be effective, and one study compared five different
questionnaires developed to assess the respondents’ perceptions of the usability of sites. The
results indicate that the different approaches appear to be quite consistent in their ability to assess
website usability. Thus, it appears that multiple approaches can come to the same conclusion.
How Much Is a Healthy Home Worth?
Homebuilders need to know which features are extremely valued by consumers, which are nice
but not important, and which are difficult to trade off. A group of researchers at the University of
British Columbia measured attitudes toward features of “healthy houses,” and they compared the
scores with a Thurstone scale. A healthy house is one built with materials and a design affording
superior indoor air quality, lighting, and acoustics. A survey asked consumers if they would be
willing to pay extra for these attributes that were presented in every combination of pairs.
Respondents had to choose which item in each pair they considered more important, which
generated a ranking that the researchers used to create a Thurstone scale. Energy efficiency was
ranked highest followed by natural light, thicker insulation, antiallergic materials, and
airtightness. Artificial light falls noticeably below the other features.
OUTLINE
I. ATTITUDES AS HYPOTHETICAL CONSTRUCTS
An attitude is an enduring disposition to respond consistently in a given manner to various
aspects of the world, including persons, events, and objects.
There are three components of attitude:
Affective component – refers to an individual’s general feelings or emotions toward an
object.
Cognitive component – represents an individual’s awareness of and knowledge about an
object.
Behavioral component – reflects a predisposition to action by reflecting a consumer’s
buying or purchase intentions.
Attitudes as Hypothetical Constructs
Business researchers often pose questions involving psychological variables that cannot
directly be observed.
Unobserved or latent variables are known as hypothetical constructs or just constructs.
Common constructs include job satisfaction, organizational commitment, personal
feelings, role stress, and so on.
Importance of Measuring Attitudes
Most managers hold the intuitive belief that changing consumers’ or employees’ attitudes
is a major goal.
Because modifying attitudes plays a pervasive role in business strategies, the
measurement of attitudes is an important task.
II. TECHNIQUES FOR MEASURING ATTITUDES
A remarkable variety of techniques has been devised to measure attitudes, stemming in part
from lack of consensus about the exact definition of the concept.
In addition, the affective, cognitive, and behavioral components of an attitude may be
measured by different means.
These techniques range from direct to indirect, physiological to verbal, etc.
Obtaining verbal statements from respondents generally requires that the respondent perform
a task such as ranking, rating, sorting, or making choices.
Ranking tasks require the respondent to rank order a small number of objects in overall
preference on the basis of some characteristic or stimulus.
Rating asks the respondent to estimate the magnitude or the extent to which some
characteristic exists.
Sorting might present the respondent with several product concepts printed on cards and
require the respondent to classify the concepts by placing the cards into groups.
Choice between two or more alternatives is another type of attitude measurement—it is
assumed that the chosen object is preferred over the other(s).
III. ATTITUDE RATING SCALES
Perhaps the most common practice in business research is using rating scales to measure
attitudes, and this section discusses many rating scales designed to enable respondents to
report the intensity of their attitudes.
Simple Attitude Scales
In its most basic form, attitude scaling requires that an individual agree or disagree with a
statement or respond to a single question.
This type of self-rating scale merely classifies respondents into one of two categories;
thus, it has only the properties of a nominal scale.
However, such scales are used if questionnaires are extremely long, when respondents
have little education, or for other specific reasons.
A number of simplified scales are merely checklists: A respondent indicates past
experience, preference, and the like merely by checking an item.
In many cases the items are adjectives that describe a particular object.
Most attitude theorists believe that attitudes vary along continua, and the purpose of an
attitude scale is to find out an individual’s position on the continuum.
These simple scales do not allow for making fine distinctions between attitudes, but
several other scales have been developed that do provide more precise measurements.
Category Scales
The simplest rating scale contains only two response categories, such as yes/no or
agree/disagree.
Expanding the categories provides the respondent with more flexibility in the rating task.
Even more information is provided if the categories are ordered according to a particular
descriptive or evaluative dimension.
A category scale is a more sensitive measure than a scale that has only two response
categories—it provides more information.
Question wording is an extremely important factor in the usefulness of these scales.
Exhibit 14.1 shows some common wordings for category scales.
Method of Summated Ratings: The Likert Scale
The Likert scale is an extremely popular means for measuring attitudes.
With the Likert scale, respondents indicate their attitudes by checking how strongly they
agree or disagree with carefully constructed statements, ranging from very positive to
very negative attitudes toward some object.
Individuals generally choose from approximately five (although alternatives may range
from three to nine) response alternatives: “strongly agree,” “agree,” “uncertain,”
“disagree,” and “strongly disagree.”
Researchers assign scores, or weights, to each possible response (e.g., 1, 2, 3, 4, and 5
assigned to each level of agreement).
Reverse Recoding
If a statement is framed negatively, the numerical scores would need to be reversed.
This is done by reverse coding the negative item so that a strong agreement really
indicates an unfavorable response rather than a favorable attitude.
In the case of a 5-point scale, recoding would make the value 1 equal to a new value
of 5, 2 equal to 4, 3 equal to 3, 4 equal to 2, and 5 equal to 1.
SPSS has a recode function.
Alternatively, for a 5-point scale, a simple mathematical formula can be entered:
Xnew value = 6 – Xold value
Composite Scales
A Likert scale may include several scale items to form a composite scale.
Each statement is assumed to represent an aspect of a common attitudinal domain.
The total score is the summation of the numerical scores assigned to an individual’s
responses.
In Likert’s original procedure, a large number of statements are generated, and an
item analysis is performed to eliminate those that are poor because they lack clarity
or elicit mixed response patterns.
Scales that use multiple items can be analyzed for reliability and validity.
Semantic Differential
The semantic differential is actually a series of attitude scales.
This popular attitude measurement technique consists of getting respondents to react to
some concept using a series of 7-point bipolar rating scales.
Bipolar adjectives—such as “good” and “bad,” “modern” and “old-fashioned,” or “clean”
and “dirty”—anchor both ends (or poles) of the scale.
The subject makes repeated judgments of the concept under investigation on each of the
scales.
Business researchers have found the semantic differential versatile and useful in business
applications.
The validity of the semantic differential depends on finding scale anchors that are
semantic opposites, which can sometimes prove difficult.
For scoring purposes, a numerical score is assigned to each position on the rating scale
(i.e., 1, 2, 3, 4, 5, 6, 7 or -3, -2, -1, 0, +1, +2, +3).
Many find it useful to assume that the semantic differential provides interval data, but
critics argue that the data have only ordinal properties because the numerical scores are
arbitrary.
Practically, business researchers treat them as metric (at least interval) because the
amount of error introduced by assuming the intervals between choices are equal is fairly
small.
Exhibit 14.4 illustrates a typical image profile based on semantic differential data.
Numerical Scales
In numerical scales, numbers, rather than semantic space or verbal descriptions, serve as
the response options to identify categories (response positions).
If the scale items have five response positions, the scale is called a five-point numerical
scale, six options is called a six-point scale, etc.
The numerical scale utilizes bipolar adjectives in the same manner as the semantic
differential.
Stapel Scale
The Stapel scale was originally developed in the 1950s to measure simultaneously the
direction and intensity of an attitude.
Modern versions of the scale, with a single adjective, are used as a substitute for the
semantic differential when it is difficult to create pairs of bipolar adjectives.
The modified scale places a single adjective in the center of an even number of numerical
values (ranging, perhaps, from +3 to -3).
The scale measures how close to or distant from the adjective a given stimulus is
perceived to be.
The advantages/disadvantages are similar to those of the semantic differential, but the
Stapel scale is easier to administer, especially over the telephone.
Constant-Sum Scale
With a constant-sum scale, respondents are asked to divide a fixed number of points
(e.g., 100) among several attributes to indicate their relative importance.
Works best with respondents who have high educational levels.
If respondents follow the instructions correctly, the results will approximate interval
measures.
As the number of stimuli increase, this technique becomes more complex.
Although widely used, strictly speaking, the scale is flawed because the last response is
completely determined by the way the respondent has scored the other choices.
Graphic Rating Scales
A graphic rating scale presents respondents with a graphic continuum.
Respondents are allowed to choose any point on the continuum to indicate their attitude.
Typically, a respondent’s score is determined by measuring the length (in millimeters)
from one end of the continuum to the point marked by the respondent.
Alternatively, the researcher may divide the line into predetermined scoring categories
(lengths) and record respondents’ marks accordingly.
The graphic scale has the advantage of allowing the researcher to choose any interval
desired for purposes of scoring.
The disadvantage of the graphic scale is that there are no standard answers.
A variation on the graphic scale design is the scale ladder; this and other picture response
options enhance communication with respondents.
Thurstone Interval Scale
The construction of the Thurstone scale is a fairly complex process that requires two
stages.
The first stage is a ranking operation, performed by judges who assign scale values to
attitudinal statements.
The second stage consists of asking subjects to respond to the attitudinal statements.
This method is time consuming and costly and is rarely used in applied business research.
IV. MEASURING BEHAVIORAL INTENTION
The behavioral component of an attitude involves the behavioral expectations of an
individual toward an attitudinal object.
The component of interest to researchers may be turnover intentions, a tendency to make
business decisions in a certain way, or plans to expand operations or product offerings.
The wording of statements used in these cases often includes phrases such as “I would
recommend,” “I would write,” or “I would buy,” to indicate action tendencies.
Behavioral Differential
The behavioral differential instrument is used for measuring the behavioral intentions of
subjects towards any object or category of objects.
A description of the object to be judged is followed by a series of scales on which
subjects indicate their behavioral intentions toward this object.
V. RANKING
Consumers often rank order their preferences.
An ordinal scale may be developed by asking respondents to rank order (from most preferred
to least preferred) a set of objects or attributes.
This task is easily understood by respondents.
Like the constant sum scale, technically the ranking scale also suffers from inflexibility in
that if we know how someone ranked five out of six alternatives, we know the answer to the
sixth.
Paired Comparisons
In paired comparisons the respondents are presented with two objects at a time and
asked to pick the one they prefer.
Ranking objects with respect to one attribute is not difficult if only a few products are
compared, but as the number of items increases, the number of comparisons increases
geometrically [(n)(n – 1)/2].
If the number of comparisons is too large, respondents may fatigue and no longer
carefully discriminate among them.
Sorting
Sorting tasks require that respondents indicate their attitudes or beliefs by arranging items
on the basis of perceived similarity or some other attribute.
A variation of the constant-sum technique uses physical counters (e.g., poker chips or
coins) to be divided among the items being tested.
VI. OTHER METHODS OF ATTITUDE MEASUREMENT
Attitudes, as hypothetical constructs, cannot be observed directly, but they can be inferred by
the way one responds to multiple attitude indicators.
A summated rating scale can be made up of three indicators of attitude (e.g., very good/very
bad, very unfavorable/very favorable, very positive/very negative).
The terminology is such that now attitude would be represented as a latent (unobservable)
construct indicated by a consumer’s response to these items.
VII. SELECTING A MEASUREMENT SCALE: SOME PRACTICAL DECISIONS
There is no best scale that applies to all research projects.
The choice of scale will be a function of the nature of the attitudinal object to be measured,
the manager’s problem definition, and the backward and forward linkages to other choices
that have already been made (e.g., telephone survey versus mail survey).
There are several issues that will be helpful to consider:
1. Is a ranking, sorting, rating, or choice technique best?
2. Should a monadic or a comparative scale be used?
3. What type of category labels, if any, will be used for the rating scale?
4. How many scale categories or response positions are needed to accurately measure an
attitude?
5. Should a balanced or unbalance rating scale be chosen?
6. Should an even or odd number of response categories be provided?
7. Should a scale that forces a choice among predetermined options be used?
8. Should a single measure or an index measure be used?
Ranking, Sorting, Rating, or Choice Technique?
The answer to this question is determined largely by the problem definition and especially
by the type of statistical analysis that is desired.
Monadic or Comparative Scale?
If a scale is other than a ratio scale, the researcher must decide whether to include a
standard of comparison in the verbal portion of the scale.
A monadic rating scale uses no such comparison; it asks a respondent to rate a
single concept in isolation.
A comparative rating scale asks a respondent to rate a concept in comparison with
a benchmark—perhaps another similar concept, such as a competing brand—
explicitly used as a frame of reference.
In many cases, the comparative rating scale presents an ideal situation as a
reference point for comparison with the actual situation.
What Type of Category Labels, if Any?
We have discussed verbal labels, numerical labels, and unlisted choices.
The maturity and educational levels of the respondents will influence this decision.
Unlabeled response categories often are selected because the researcher wishes to
assume interval-scale data.
How Many Scale Categories or Response Positions?
This issue is basically a matter of sensitivity, but at the operational rather than the
conceptual level.
Balanced or Unbalanced Rating Scale?
The fixed-alternative format may be balanced or unbalanced.
A balanced rating scale has a neutral point, or point of indifference, at the center of
the scale.
Unbalanced rating scales may be used when the responses are expected to be
distributed at one end of the scale; an unbalanced scale may eliminate this type of “end
piling.”
Choice generally depends on the nature of the concept or the researcher’s knowledge
about attitudes toward the stimulus to be measured.
Even or Odd Number of Scale Points?
Should there be a “neutral” point in the scale?
If the researcher feels that respondents can truly hold a neutral attitude, an odd number
of scale points is appropriate.
If issues tend to evoke strong opinions, these are likely better captured by an even
number of scale points where a respondent is forced to come down on one side of the
issue.
Use a Scale That Forces a Choice Among Predetermined Options?
In many situations, a respondent has not formed an attitude towards a concept and
simply cannot provide an answer.
If a forced-choice rating scale compels the respondent to answer, the response is
merely a function of the question.
If answers are not forced, the midpoint of the scale may be used by the respondent to
indicate unawareness as well as indifference.
If many respondents are expected to be unaware of the attitudinal object, this problem
may be eliminated by using a non-forced-choice scale that provides a “no opinion”
category.
The argument for forced choice is that people really do have attitudes, even if they are
unfamiliar with the objects and should be required to answer the question.
Still, the use of forced-choice questions is associated with higher incidences of item
non-response (no answer).
Single Measure or an Index Measure?
Depends on:
the complexity of the issue to be investigated
the number of dimensions the issue contains
whether individual attributes of the stimulus are part of a holistic attitude or are
seen as separate items.
The researcher’s conceptual definition will be helpful in making this choice.