Perspectives on Involvement: Current Problems and Future Directions

Michael L. Rothschild, University of Wisconsin
[ to cite ]:
Michael L. Rothschild (1984) ,"Perspectives on Involvement: Current Problems and Future Directions", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 216-217.

Advances in Consumer Research Volume 11, 1984      Pages 216-217


Michael L. Rothschild, University of Wisconsin


As involvement has become a very popular construct, the literature has become replete with papers that are overly concerned with defining this hypothetical construct, organizing concepts and reviewing past work. This paper discusses problems related to an abundance of such work and suggests some other directions for researchers to take.


This is an unusual session. It is rare to see four papers on the same topic. It is a luxury to be able to discuss papers with a common theme. In my comments I would like to discuss how I feel these papers represent current dominant thinking in the involvement field, use the papers as examples of problems and suggest alternative approaches to the study of involvement. Portions of this last agenda item come from participants of a mini-conference on involvement held at New York University in June 1982.

There are a number of common points made by these and most other involvement papers:

1. Involvement is the greatest thing since sliced bread. I certainly agree with this.

2. There is no commonly accepted definition of involvement

3. There is no single direction that involvement research is taking.

I agree with both of these latter points also, but my reaction is "so what." I will discuss these two points in-depth. These papers also serve as examples of what I perceive to be problems in the involvement literature:

1. There is too much theorizing.

2. There is too little data collection.

3. There is too much complaining about lack of structure.

4. There is too much repetitive reviewing of past review papers.

I am uncomfortable with publicly denigrating the work of my colleagues. It is much easier to participate in a blind review process; I feel compelled to make these comments, though, because the papers are representative of too such current involvement work. My purpose is to try to change the direction of future papers.

The first paper is by Muncy and Hunt. This paper is a prototypical review of involvement. It shows that there are five components of involvement, and it is very similar to too many other papers beginning with Houston and Rothschild (1978) and ending with Bloch and Richins (1983). Most of these papers suggest three components; Muncy and Hunt suggest five; the five components seem similar to the three of other authors.

This constant recompartmentalization is an example of some of our problems.

Muncy and Hunt also sound the challenge to clean up the involvement mess but don't offer any new ideas to do so. Again, this is an example of what too many involvement papers try to do but never accomplish.

Antil goes beyond Muncy and Hunt. Re defines involvement as a state of perceived importance or a state of - interest evoked by the stimulus and the situation. This is a fine definition but isn't really new. Others have defined involvement as a state of motivation or a state of arousal. One problem with too many current papers is that involvement is constantly redefined but usually in very minor ways. This is of marginal usefulness.

Antil also suggests that involvement is a continuous variable, not a dichotomous one. I agree; probably most of us agree, but it is hard to do research with a continuous variable. It is manageable if the continuous variable is dependent; it is virtually impossible if this variable is independent and an experimental design is used. In addition, Antil suggests that more than two hierarchies of effect (low and high) are needed. If involvement is being used as a continuous variable, do we need an infinite number of hierarchies? For these reasons a discrete set Of values of involvement are preferred; a dichotomy is most preferred.

Finally, Antil proposes that we develop universal measures of involvement across products based on benefits. This is a nice abstract idea but how does one do this? How does one compare the benefits of a kleenex not tearing when blows one's nose versus the pulp of orange juice not sticking in one's teeth. Antil may have a good idea but I won't accept it until I see some data to support it. I won't believe it because I've tried it and it didn't work (Rothschild and Houston unfortunately unpublished). That doesn't mean it can't work, but rather the burden is on Antil to snow that it works. The point here is common to many papers. There are too many abstract ideas presented with too few data sets to support the ideas.

Stone also goes beyond Muncy and Hunt. He defines involvement as time and/or intensity of effort expended in behavior. This seems similar to what others have called response involvement and have defined in terms of complexity of processing. Again the concept is nice but not new. It has other problems also.

Is this a definition of involvement or of what results from involvement? To me, involvement is a state of interest, motivation or arousal; in turn effort is a function of the level of involvement

I am also troubled by the distinction between intensity of effort and complexity of processing. Where does one end and the other begin? If I watch a subject read a label, what am I observing? It is possible that the two concepts differ (although many psychologists feel that cognitive processing is a behavior), but Stone needs data to convince me. Again, in the absence of data there seems to be little new ground broken with this definition.

Bloch and bruce go in a direction that has more potential, I feel. They bring in a new literature and a new set of ideas, and attempt to expand the concept of enduring involvement. This has the potential to be a useful addition, but without data is, again, a minor change from past definitions of enduring involvement. The concept that leisure behavior is strongly related to enduring involvement may have some merit if new ideas from the new literature can be added to involvement.

These four papers provide examples of issues which pervade the involvement literature. How should we proceed to remedy the problems? This issue was briefly addressed by participants at the NYU involvement conference. The majority (or at least most vocal) feeling seemed to be that at this time we should not try for a perfect definition of involvement. Rather, we should decide on a generally acceptable generic definition which will allow us to comfortably proceed to other matters. Since we are dealing with a hypothetical construct, we will probably never achieve agreement.

The proposed generic definition was that involvement is a state of motivation, arousal or interest. This state exists in a process. It is driven by current external variables (the situation; the product; the communications) and past internal variables (enduring; ego; central values). Its consequents are types of searching, processing and decision making.

We don't need more at this time. Let's call a ten year moratorium on definitions of involvement; let's go collect data on interesting aspects of involvement and then, in ten years, see if we can (or need to) devise a better definition.

What directions should we pursue? To ne the value of involvement is in recognizing that not all decisions are important and that there are low level learning processes to examine. At this moment there are many people doing consumer research who are implicitly assuming a high level of processing, learning and motivation. we need to explore alternative processes. Leave high involvement to others; those of us interested in involvement need to explore low involvement.

Those who wish to pursue this avenue should start with Leavitt, Greenwald and Obermiller (1981). This paper proposes a hierarchy of responses; at the top are cognitive responses. Next comes recall, then recognition and at the bottom there are physiologic responses. The top two require a high level of motivation in order for a subject to respond. The bottom two can be tapped even though the subject may have low involvement. We need to use measures which gain responses in a spite of low motivation.

I would like to suggest some examples of such work. Singh (Singh and Rothschild 1983) has developed methods to tap learning via recognition. These measures discriminate between stimuli and show that learning has occurred where recall shows no learning.

In the physiologic areas there is exploratory work being done by Olson and Ray (brain waves), Bagozzi (heart rate), Cacioppo and Petty (electromyogram), Mizerski (multiple measures), Kroeber-Riel (multiple measures) and Rothschild, Reeves and Thorson (brain waves) among others. In each case the response can be obtained in a low involvement case without tampering with the level of involvement.

Low involvement is important in developing advertising strategies. Thorson (1983) is studying the psycholinguistic structure of messages and has found a relation between structure and learning. What type of structure is most appropriate when consumers have low levels of involvement? Similarly, Cacioppo and Petty (1983) show learning differences as a function of informational versus inferential message cues. What type of cues are most appropriate when consumers have low levels of involvement?

In the promotions area Smith and Swinyard (1983) show that the attitude-behavior relation is lowest when involvement is low, but they also show that the relation is strengthened by product trial. This supports the heavy use of promotions in low involvement cases. Peter and Nord (1982) and Rothschild and Gaidis (1981) have developed models of low involvement based on operant conditioning. These are potentially important in developing promotions strategies for low involvement cases.

These are just a few examples; there are more. In each case the work deals with low involvement and shows that there are alternatives to the current models that imply high involvement. These are examples of the direction that I feel involvement research should take. We don't need more theoretical papers, literature reviews or definitions of involvement at this time.


Bloch, P. H. and M. L. Richins (1983), "A Theoretical Model for the Study of Product Importance Perceptions," Journal of Marketing, 47, 3 (Summer), 69-81.

Houston, M. J. and M. L. Rothschild (1978), "A Paradigm for Research on Consumer Involvement," unpublished working paper. University of Wisconsin-Madison.

Leavitt, C., A. G. Greenwald and C. Obermiller (1981), "What is Low Involvement Low In?," in K. Monroe (ed.), Advances in Consumer Research, Vol. VIII, Ann Arbor: Association for Consumer Research.

Peter, J. P. and W. R. Nord (1982), A Clarification and Extension of Operant Conditioning Principles in Marketing," Journal of Marketing, 46, 2 (Summer), 102-107.

Petty, R. E., J. T. Cacioppo and D. Schumann (1983), "Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement," Journal of Consumer Research, 10, 2 (September), 135-146.

Rothschild, M. L. and W. C. Gaidis (1981), "Behavioral Learning Theory: Its Relevance to Marketing and Promotions," Journal of Marketing, 45 (Spring), 70-78.

Singh, S. N. and M. L. Rothschild (1983), "Recognition as a Measure of Learning From Television Commercials," Journal of Marketing Research, 20, 3 (August), 235-248.

Smith, R. E. and W. R. Swinyard (1983), "Attitude-Behavior Consistency: The Impact of Product Trial Versus Advertising," Journal of Marketing Research, 20, 3 (August), 257-267.

Thorson, E. (1983, Forthcoming), "Propositional Determinants of Memory for Television Commercials," in C. Martin and J. Leigh (eds.), Current Issues in Advertising, Ann Arbor: University of Michigan Press.