Main Content Area
Attitude is NOT the Same as Intention or Behavior - Part 2
In the previous posting I described two main ways that attitudes have been measured in consumer research (overall vs. per component). In this one, I will provide a few more observations, suggestions, and illustrations.
First, if understanding attitude is extremely important to your study then measure the components separately. Work conducted by Petty (Crites, Fabrigar, and Petty 1994; Petty, Wegener, and Fabrigar 1997) supports the notion of separately measuring the components. The benefit is being able to determine the relative effect of each on something else such as purchase intention, another attitude, or on behavior itself. For example, imagine if you were trying to determine if the affective component of an attitude regarding a new piece of consumer technology better predicted adoption than did the cognitive component. In other words, you would be checking to see whether the new product's adoption was more related to its usefulness and ease of use (the cognitive component) or by how fun and exciting it was (affective component). Determining this would not be possible if a global measure were used.
Second, behavioral intention is considered to be the construct in closest proximity to actual behavior (e.g., Fishbein and Ajzen’s 1975). Since it may be measured more easily than actual behavior, it is no wonder that it is among the most measured constructs in scholarly marketing research using multi-item scales. Another thing to strongly consider when measuring intention is that its predictive value is best when the time period is specified and temporally near rather than far. People are more accurate in predicting what they will do in the next few days than what they will do weeks and months from now. Given that, instead of asking something open-ended like “Do you plan to download coupons” it is better to ask “Do you plan to download coupons in the next week?” See Ittersum, Koert Van and Fred M. Feinberg (2011) for more discussion of this issue.
Examples of each of these measures as they relate to an ad are provided below.
Cognitive component. The most precise way to measure beliefs is by statements that express what respondents think is true about an object. Here are some example statements: online coupons are easy to use, it is a hassle to print out online coupons, coupons are complicated to download, and online coupons are easy to use. However, phrasing scale items that way and using a Likert-type response format (agree/disagree) may make the scales idiosyncratic. A popular alternative to measuring the cognitive component while being more malleable for a variety of objects and contexts would be using semantic differentials. Note in the following example that the scale stem refers to an ad but the items themselves could just as easily be used with respect to a spokesperson, a book, a news program, or many other things:
The ad is:
- informative / uninformative
- believable / unbelievable
- convincing / unconvincing
Affective component. Measuring this part of an attitude involves describing the object in emotional terms. As explained in the previous posting, the scale is phrased such that the stimulus (e.g., the product, the ad, the company) is the object of the statement or descriptor, not self. A Likert-type scale could be used but semantic differentials are quite popular when measuring the affective component as shown in this example:
The ad is:
- likeable / dislikeable
- enjoyable / unenjoyable
- exciting / boring
Behavioral Component: Items used to measure this part of an attitude express evaluation of action with respect to an object but do not go so far as to state the degree to which a person is personally committed to engaging in the action. An example is:
Buying the product featured in the ad would be:
- good / bad
- wise / foolish
- pleasant / unpleasant
Behavioral Intention: Instead of evaluating the propriety of an action in a hypothetical sense, a proper measure of intentions has people specify the degree to which they plan to personally engage in the behavior. An example is:
What is the probability that you will buy the product featured in the ad during the next week?
- unlikely / likely
- improbable / probable
- uncertain / certain
For more discussion of measuring purchase intention and related constructs, see another of my pet-peeve posts (Bruner 2017).
Bruner II, Gordon C. (2017), “Loosey-Goosey Measurement of Purchase Intention,” Scale related Pet-Peeves, Blog #30.
Crites, Stephen L. Jr., Leandre R. Fabrigar, and Richard E. Petty (1994), “Measuring the Affective and Cognitive Properties of Attitudes: Conceptual and Methodological Issues,” Personality and Social Psychology Bulletin, 20 (December), 619-634.
Fishbein, Martin and Icek Ajzen (1975), Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research, Reading, MA: Addison-Wesley Publishing Company.
Ittersum, Koert Van and Fred M. Feinberg (2011), “Cumulative Timed Intent: A New Predictive Tool for Technology Adoption,” Journal of Marketing Research, 47 (Oct.), 808-822.