siu logo

office of scale research

Department of marketing


Main Content Area

Scale-related Pet-Peeves

Blog #11


Attitude is NOT the Same as Intention or Behavior - Part 1

We could discuss forever the meaning and importance of the theoretical construct called Attitude. For brevity sake, I will skip that discussion and focus instead on what has actually been done by scholars who have measured the construct. Basically, an analysis I conducted (Bruner 1998) as well as my continuing review of scales show that some researchers break an attitude down into its theorized components and then measure those parts separately. Other researchers view the construct as having one dimension and use a scale they think measures an attitude at a "global" level.

If you accept the tri-partite view, the parts are usually referred to as the Cognitive, Affective, and Behavioral Components. The Cognitive Component refers to what a person believes to be true about an object. It is the facts and figures the person accepts as true about an object. (Examples will be provided in the next post.)

The Affective Component refers to the feelings or feeling-related thoughts one has about an object. As I said in a previous posting, be aware that the scale items should describe an external object (e.g., an ad, product, company) rather than one's self. In other words, the Affective Component is not the same as the emotion one experiences.

The Behavioral Component is defined in various ways, but it is important to note that it definitely does not refer to the behavior itself. Instead, it is a person's evaluation of a behavior. The most typical behavior in the consumer behavior literature is the purchase or use of a product. There are several constructs that are either the same as the Behavioral Component or are related to it. That is the focus of a separate discussion.

The bothersome part for me as I review scales is seeing a measure in which a mixture of items from distinctly different components are used together. Regardless if such scales have a high reliability, the items are not measuring the same construct and should not be mixed together. If you do it anyway, it will be unclear which construct is playing the primary role in the relationship(s) you are trying to understand.

As for global scales, they are certainly used a lot, but I think it is challenging to find items that are truly global rather than measuring a particular component. Sometimes factor analysis teases the items apart into their component groups and sometimes it doesn't. When it doesn't then I suppose you can argue that you have a unidimensional scale but is it measuring the full construct or is it actually just measuring part of it. That may be best decided with a well executed analysis of the scale's content validity.

In my next posting I will provide some examples of each type as well as make some final comments.


Bruner II, Gordon C. (1998), “Standardization & Justification: Do Aad Scales Measure Up?” Journal of Current Issues & Research in Advertising, 20 (Spring), 1-18.