Three Consumer Behavior Studies and Their Implications For Marketing Communications

George M. Zinkhan, University of Houston
ABSTRACT - All three papers in this session have potentially important implications for advertising and promotion. Suggestions are offered for improving the papers; and, in most instances, this involves deleting irrelevant issues or terminologies from the discussion. A model is proposed to synthesize these papers, and some possible directions for future research are explored.
[ to cite ]:
George M. Zinkhan (1987) ,"Three Consumer Behavior Studies and Their Implications For Marketing Communications", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 502-504.

Advances in Consumer Research Volume 14, 1987      Pages 502-504


George M. Zinkhan, University of Houston


All three papers in this session have potentially important implications for advertising and promotion. Suggestions are offered for improving the papers; and, in most instances, this involves deleting irrelevant issues or terminologies from the discussion. A model is proposed to synthesize these papers, and some possible directions for future research are explored.


One common theme which ties three studies in this session together is that all three papers have been designed to explore some aspect-of promotion, and all three have important implications for marketing communications. For example, the Schindler et al. (1987) paper has implications for point of purchase displays and packaging. The Grossbart et al. (1987) work has implications for designing claims for television commercials; and the Stern (1987) paper has several advertising implications, including stereotypes in advertising and factors contributing to advertising recall.

Individually, each of the papers has particular strengths, and each makes a contribution. However, each of the papers contains components which may not be necessary. In some cases, the papers could be strengthened if certain excess baggage were left out. Specifically, I think that the authors sometimes invoke theories or terminologies which aren't directly relevant to their purpose. This point will be elaborated as the discussion proceeds.


This paper has implications for a very important topic in promotion that often receives little research attention: point of purchase displays. At present we know surprisingly little about how consumers make quick choices under time pressure.

The method employed in this study is very ingenious. The lottery technique using a bingo drum is pioneered by these authors and may have interesting applications to other problems in consumer behavior research. The results which are obtained are in the expected direction, and consequently the authors contribute to our limited knowledge in this promotional area. Two suggestions are offered here for those who continue research on this important topic. The first suggestion concerns the experimental procedure itself, and the second concerns the relative advantages/disadvantages of laboratory versus field experiments.

Experimental Procedure and Materials

Schindler et al. show examples of their bingo cards in their figure. As shown in the figure, all cards are numbered sequentially from 1 to n (where n = 4,25, or 100). It seems to me that it might be better to randomly assign the numbers to the bingo cards, rather than numbering sequentially. This random assignment would increase the difficulty of respondents' information processing task, and this is the key experimental manipulation. It is important that respondents reach the limits of their processing ability if the theoretical model proposed by Schindler et al. is to apply. The way the numbers are currently ordered on the grid, respondents can instantly remember where all the numbers are, due to a chunking effect. Thus the typical respondent, upon being exposed to the sequentially numbered bingo card for 5 seconds, could perfectly reproduce the card.

At the beginning of the Schindler et al. paper is a fairly detailed discussion of consumer information processing limitations and the notion that consumers can store only about 7 items in an evoked set. Here, the classic Miller (1956) article is invoked concerning the magical number 7, plus or minus 2. Unfortunately, none of this discus ion is relevant given the particular experimental procedure. Because of chunking effect, a typical respondent can remember the entire bingo grid. It is not necessary for a respondent to remember every single number, since the numbers are all in order.

Thus, for this paper, it might be better to leave out the notion of information processing limitations. I'd be interested in seeing future results from this procedure when the numbers are randomly assigned to the bingo cells so as to eliminate the possibility of a chunking effect.

Laboratory Versus Field Experiments

In this particular study, Schindler et al. conduct a laboratory experiment which offers fairly high internal validity but at the cost of low external validity. The authors would, I think, agree that there is some distance between their experimental conditions and actual consumer purchases.

It wold be interesting to test the research hypotheses in a more naturalistic setting. Two possibilities come immediately to mind. First, a simulated store could be established and consumer choices observed under time pressure. Alternatively, a field experiment could be run at the check-out counter of a cooperating grocery store. The salience of various candy bar packages could be manipulated (by creating different color packages), and the effects on sales could be monitored through electronic scanner data. Another possible salience manipulation could be set up by distinguishing high salience portions of the candy display rack (e.g., some candy bars are displayed precisely at eye level. Of course it would be essential to pretest either the candy wrapper colors or the section of the display rack in order to insure that the salience manipulation was effective.

One of the main contributions of the Schindler et al. work is that it brings an important theoretical perspective to bear upon the issue of consumer purchase of impulse or convenience items. If this theoretical framework could be linked to some of the important shelf space experiments conducted in previous years (see for example Cox 1964, 1970), then we might begin to understand something about how consumers make quick choices from a set of similar alternatives.


The Grossbart et al. paper examines an important question in advertising research: what type of claims are the most effective for television commercials? Here, two types of claims are considered: objective and subjective. The authors go to the trouble of creating and producing two TV commercials to fit their particular definitions of objective/subjective claims. It is no easy task in advertising research to create advertising stimuli that elicit from respondents just the manipulation that fits the research objective. Here, the authors seem to have done a complete job; and, for the most part, these objective and subjective claim manipulations seem to be successful.

One aspect of this paper which does appear troubling is the whole concept of "commitment." It's ture that the authors didn't coin the term "commitment." Commitment is a proposition developed in psychology, and there have been some marketing studies that have used the term. As discussed here, however, it's not clear how the authors intend to differentiate commitment from other related concepts, such as: different notions of involvement (e.g. task involvement, brand involvement), brand loyalty, or brand experience. It would be helpful if the authors could offer specific definitions for all of these terms. As the paper is currently written, the authors mention some of these overlapping terms but never precisely define how they would be different/similar from one another. Unfortunately, this lack of clarity about what commitment is occurs at both the theoretical level and at the measurement level.

Admittedly, this semantic and measurement confusion is not easy to clear up in consumer behavior research. At this stage of development, we don't exactly know how involvement, brand loyalty, and product experience are related. The authors do offer some discussion of these different constructs, but I don't think that the problem is solved by lumping all of these processes together and calling the group of them "brand commitment."

It is understandable that the authors would not want to use more laden words to describe their research, such as "brand loyalty" or "involvement." Both of these teras seem to have many competing definitions and measures. But again, the problem is not solved by giving the whole phenomenon a new name. To summarize, this paper might have been stronger if this whole notion of "brand commitment" were left out and more traditional terminology (e.g., brand loyalty) substituted in its place.

In terms of results and implications, the authors do a fairly good job of summarizing some limitations of their research design. Specifically, they studied a fictitious brand using only two attributes. Their subjects were all students in an atypical viewing environment. These are all common problems/limitations in advertising research.

Aside from these considerations, the results themselves are rather weak. Hypothesis two is not supported. There is only partial support for hypothesis one, and hypothesis three is supported at the .1 level of significance. In brief, there is almost no variance explained in the criterion (attitudes). Given these weak results and the limitations, it is rather difficult to draw meaningful managerial recommendations or conclusions from the study.


There are many positive aspects of the Stern paper. The review of gender research is well done, and this review provides a valuable synthesis of marketing and nonmarketing literature. The discussion of symbols and imagery is interesting, and the advertising examples add to the paper. Since this is largely a theoretical piece, it is interesting to speculate whether or not the author plans any empirical work to test the hypotheses which she so carefully develops. If so, what would be the nature of this empirical work? From the material which is presented, it's not exactly clear what empirical directions would be the most promising. Aside from this issue, there are two other unanswered questions: 1) which is more important for consumer researchers to study, the biological sex of the respondent or sex-role self-concept (the psychological world of gender)? 2) what has services marketing got to do with all of this?

Biological Versus Psychological Gender

At the outset of the paper, the author mentions that biological sex (as opposed to sex-role self-concept) has proved to be the more important predictor in marketing-oriented research. Also, the author discusses at length some of the controversies and problems which have beset psychosexual research. Despite this, however, the bulk of the paper concentrates upon the psychosexual dimension, All the hypotheses and rationales for hypotheses concentrate and depend upon psychological explanations. Thus, the reader is left with an unanswered question: which is most important/useful for consumer researchers to consider, biological or psychological gender?

Services/Product Dichotomy

The distinction between services marketing and product marketing provides a useful and interesting framework for thinking about consumer research. However, in this particular instance, I'm not certain that it is so useful or relevant to all areas of gender research. For example, hypothesis 5 essentially says that gender (and gender changes over time) can be important segmentation variables. It's not necessary to invoke the product/services dichotomy to justify that conclusion. In developing H5, Stern's logic proceeds as follows: Services managers are very interested in managing demand; gender differences cause changes in demand patterns; therefore, services managers should be very interested in gender research. Obviously, there are a large number of factors (beyond gender differences) which could potentially influence demand patterns.

The same sorts of problems appear in the development of other hypotheses. For example, there's a discussion of production variability for services, and this discussion is used to justify a hypothesis about stereotypes in advertising (H3). Again the specific connection to services marketing is unclear since sex-role stereotypes in advertising have more to do with imagery (or other factors) than with the production variability of service marketing.

In summary, Stern describes an interesting set of hypotheses, most of which are directly derived from a thorough review of the gender literature. It appears, however, that most of the references to services marketing could be dropped, and still we would be left with the same set of hypotheses. Thus, this paper could be improved by eliminating most allusions to the services/product dichotomy and instead concentrating upon the findings and theories forthcoming from research on gender differences.


One potential task for a discussant is to provide an overall framework to unite the studies in a particular session. Fortunately for this discussant, Robert Schindler provided such a framework in the course of the session, and his model is displayed in Figure A. This model proposes that ideas are the motive forces behind consumer behavior. As ideas come to prominence in the mind, they are evaluated by a filter (similar to Freud's notion of superego). Those ideas which are approved by the filter are then put into action (and result in a consumer behavior). Those ideas which don't pass the filter are suppressed or forgotten or else some action is taken to remove those ideas from consciousness. This framework is attractive since it is consistent with existing psychological theories (e.g., Freudian) and since it provides a simple explanation for many of the "automatic" behaviors which we study in marketing/consumer behavior. For example, Figure A provides a complete description of the manner in which a consumer responds to a point of purchase (POP) display in a grocery store. The POP display triggers an idea (brand awareness, perhaps). Nest, this idea is passed through a filter (e.g., "have I purchased this brand previously? If 60, did I like it, can I afford it?"). If the idea can pass through this screening process, it becomes the motive force for action (the consumer purchases the POP brand). Otherwise, the POP display is promptly forgotten, and the consumer moves on to the next shopping decision.

Of course, the Figure A model is not necessarily appropriate for all types of consumer behavior. It would not be adequate for explaining behavior which results from a complex choice process (e.g., purchasing a new automobile). Nevertheless, this model is quite attractive for explaining routine response behavior or impulse buying or shopping for convenience goods.

In addition, this Figure A model provides an interesting framework for thinking about the three papers in this session. As already described, the model works quite well for explaining how consumers might react to attention-getting devices in the store and, as such, provides a parsimonious explanation for the Schindler et al. results. The Grossbart et al. study concentrates on commitment (or brand loyalty). In terms of the Figure A model, brand loyalty could serve as an important consumer filtering device. For example, the consumer may ask (either subconsciously or consciously): "Is this the brand I usually buy?" y by passing this filter can the advertising message have an effect and result in purchase behavior. Likewise, consumer ideas (or even consumer evaluation filters)may vary depending upon biological or psychological gender. In this respect, the Figure A model provides a useful perspective for thinking about the Stern hypotheses.



As mentioned at the outset, the common theme which unites these three studies is that all three have implications for marketing promotions. To date, relatively little work has been done on point of purchase displays (see Schindler et al.). In contrast, there are quite a few studies on advertising effects (see Grossbart et al. and Stern). Perhaps it is time in consumer research that we begin to study how some of these promotional devices interact. For example, some interesting experiments could be designed to explore how advertising and POP displays work together to influence choice behavior. As suggested by other speakers at this conference, consumer behavior research may be able to contribute to marketing knowledge by widening its frame of reference, by considering together variables which previously have been studied only in isolation. The framework outlined in Figure A provides one approach for organizing and guiding such a comprehensive investigation of marketing communications.


Cox, Keith (1964), "The Responsiveness of Food Sales to Supermarket Shelf Space Changes," Journal of Marketing Research, 1 (Hay), 63-7.

Cox, Keith (1970), "The Effects of Shelf Space Upon Sales of Branded Products," Journal of Marketing Research, 7 (February), 55-8.

Grossbart, Sanford, James Gill, and Russell N. Laczniak (1987), "Influence of Brand Commitment and Claim Strategy on Consumer Attitudes," in Advances in Consumer Research, Vol. 14, eds. P. Anderson and M. Wallendorf, Ann Arbor, MI: Association for Consumer Research.

Miller, George A. (1956), "The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information," Psychological Review, 63, 81-97.

Schindler, Robert M., Michael Berbaum, and Donna R. Weinzimer (1987), 'Mow an Attention-getting Device Can Affect Quick Choice among Similar Alternatives," in Advances in Consumer Research, Vol. 14, eds. P. Anderson and M. Wallendorf. Ann Arbor, MI: Association for Consumer Research.

Stern, Barbara B. (1987), "Gender Research and the Services Consumer: New Insights and New Directions," In Advances in Consumer Research, Vol. 14, eds. P. Anderson and M. Wallendorf, Ann Arbor, MI: Association for Consumer Research.