Theoretical Views of Advertising Effects: Discussant Comments

Ivan Ross, University of Minnesota
[ to cite ]:
Ivan Ross (1983) ,"Theoretical Views of Advertising Effects: Discussant Comments", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 226-228.

Advances in Consumer Research Volume 10, 1983      Pages 226-228


Ivan Ross, University of Minnesota

The Golden/Johnson, "Thinking/Feeling", paper addresses a very complex subject, and one which I think will continue to be elusive.

The main difficulty is the confounding between the stimuli and the measurement procedures. The television commercial stimulus has both aural (right brain) and visual (left brain) components, the measurement procedure (both sensory preference and advertising evaluations) is via print (left brain), the demand characteristics confronting subjects are mostly of the "sequential" task" (left brain) sort, and yet holistic perception of the stimuli and of the accompanying measurement procedure are somewhat toward the right brain. Thus, it is unclear exactly what is being measured when subjects are asked (left brain) to state a preference for auditory (right brain) stimuli, and when the authors state that the purpose deals with explaining preference for visual versus auditory verbal information, when clearly, that which is "verbal" is essentially left brain. Is "verbal" the same thing as "information", or is "verbal" only one sort of information, and indeed, not the sort of information which is contained in "emotional/feeling" advertisements?

There are many other problems. For example, consider the demand character of the task. Would we not expect that respondents would think that thinking ads are more likeable, more informative, and as having more useful information. and therefore evoking higher purchase intentions than feeling ads which, after all, are to be felt and not thought about? Are not thinking ads more "socially desirable" than feeling ads?

And is a subject whose sensory preference score on the Oakland Inventory is a 20 (auditory) versus 19 (visual) properly treated as being the same as one whose score is 10 versus 9, respectively? And is a subject who is high on both scores the same as one who is low on both? And as with subjects, can an advertisement be high in both "thinking" and "feeling"? Indeed, aren't most television commercials? And what about the subject who was exposed to an advertisement classified as "thinking" and yet for that subject, rated it higher on "feeling" than "thinking" and vice-versa? (And might a reanalysis of the data in terms of this consideration shed new light on Hypothesis 1?) Were all the subjects of the same sex and. if not, do sexes differ in either sensory preference or advertising "type" preference? And what else might be different about the two advertisements for a brand one "thinking" and one "feeling", besides this dimension. . . total information content, tone, intrusiveness, involvement, etc.?

Clarification of these issues in future research may still confirm the authors' findings that thinking and feeling ads produce different communication effects. but this study does not permit a generalization that thinking ads are more persuasive regarding purchase intentions. Were this the case, we would obviously see few "feeling" advertisements since commercial advertising research facilities such as ASI, Burke, McCullom-Speilman. Starch, Mapes and Ross, and so on, after all these years of accumulated evidence, would have long since uncovered this generalization, and they have not.

Then is the distinction which the authors address a relatively unimportant component of the magic that makes an advertisement effective? Or is the "thinking"/"feeling" distinction so inextricably confounded with the message itself, the product, and so on, that it doesn't make sense to separate these components for experimental inquiry, that is, the very separation removes any potential for external validity? Or, of course, it may be that there's something here which we just haven't seen because of these yet to be resolved measurement problems.

The Munch and Swasy, "Conceptual View of Questions", paper is good food for thought, and I think with some enrichment could lead to adequate conceptualization and theorizing that testable hypotheses could be generated.

But I think this is a much tougher topic than other "message variables" such as fear, humor, sex, passive vs. active voice, and so on! because as the author's point out, some "questions" may neither be intended by the source or understood by the receiver to be questions, and some non-questions may be intended and understood as questions. Therefore, until there is an operational definition of what message elements (thinking of advertisements) or components of exchange processes (e.g., face-to-face selling, intrachannel interactions/negotiations/selling, etc.) are "questions", it's difficult to go much further without the nagging feeling that some necessary groundwork has not been clearly laid. None of the definitions offered by the authors from the various perspectives identified satisfy my appetite for clarity.

For example, defining a question as an attempt by the questioner to elicit information from a respondent is too general. Most "non-questions" are put for the same reason. The same with the psycholinguistic definition (subordinating self to other, "craving" a response) and the cognitive psychology/psycholinguistic perspective. where a question is gathering knowledge/information.

Whether Bales' or some other codification scheme can be reliably employed to score exchanges given the broader view of questions suggested by the authors in Figure 1 remains to be seen. I suspect that such a codification scheme would be very difficult to develop given the facts that: 1) information external to the exchange itself may be necessary to accurately evaluate what goes on in the exchange, 5) much of the cause and effect of exchange is covert, and therefore may never be known (or knowable) by an observer, and 3) facial expressions, pauses, intonation, and other such non-verbally structured material is crucial, and indeed, may be primary. in understanding the exchange.

And beyond these considerations there are other conceptual problems requiring clarification and or enrichment. For example, and as the authors' literature review touches upon, there are many different kinds of "marketing communication" contexts, and the role and meaning of "questions" may be sufficiently different in each that a taxonomy reflecting these situations may be necessary.

For example, there are face-to-face interactions. But because they do occur in many different contexts, the "meaning" of questions exchanged therein may be very different. In typical customer-salesperson interaction where neither of the participants in the negotiation have knowledge about one another, it would presumably be true that many questions initiated by the salesperson are intended to 'size-up" or otherwise qualify the prospect. The customer, on the other hand, may devote most "questioning" to the product, to the terms of sale... rather than to the salesperson's motives. Now shift the face-to-face interaction to negotiations between a buyer and a vendor. They have interacted before. The focus of the interaction would more apt directly move to the functional aspects of the exchange. Now put the interaction on the telephone. Would there be differences in the nature and purpose of questions? Now put the interaction in the form of written exchanges...telegrams. letters, and so on. How does that interaction mode influence the meaning of questions? Now when you move to questions in print or electronic advertising media. I think you again have a very different situation. Because the communication is one-way only. it must be true that the way in which questions are used and what they mean are different.

Not only do each of these marketing communications contexts have these obviously different structures. so too is it obvious that in each of these situations the involvement and motivation character and level is different. The hypotheses advanced by the authors for the different way in which open versus closed-end questions might work in controlling or directing the perceptions or beliefs of the recipient may make sense in one but not another of these contexts...personal selling versus media advertising, for example. A question in print advertising may be most usefully viewed as a hook to gain attention (whether open or closed) but have little or nothing to do with -the receiver's belief in or being persuaded by the message. In personal selling, the primary role of the question may be in funneling the customer's perceptions to the advantages of the particular product, i.e., in persuasion, and have little to do with attentional issues.

In sum, there needs to be a stronger conceptual foundation developed before research in this area can easily progress. Much of the literature cited by the authors may have relatively little theoretical bearing on marketing communication issues. For example. criminal proceeding interactions, casual (non-sales oriented) conversations, and educational/learning/teaching contexts, may have their own logic regarding the meaning of questions, and this logic may not be correspondent with that required in marketing communications.

Klein and Wolfson's "Paraproxemic", paper shares the goal of the Golden/Johnson, "Thinking/Feeling". paper in that it too seeks to discover stimulus properties relating to information processing concepts that would hold across situations.

Toward the end it would be useful for consideration to be given to the types of communication situations (or advertisements, if you will) represented in the research. The study does not make clear what the stimulus message was. but whatever it was, these situations need to be varied in order for generalizations to emerge. Similarly, the actor in this study was male. Are we to assume the same pattern of findings would emerge with a female spokesperson? Or would we expect an opposite sex interaction as in the present study?

And other factors may interact with the paraproxemic issue. For example, is not the perceived attractiveness or credibility of the spokesperson him or herself an issue, regardless of distance from the viewer? Or is that question ultimately confounded since credibility must always be measured at one or another "distance"?

What about the information content of the message? Is the impact of paraproxemics on perceived credibility different in a commercial which is factually oriented versus mood/emotional oriented? (Here is an opportunity for Klein and Wolfson to join with Golden and Johnson.)

Is what the actor is saying during the paraproxemic manipulation important, and does the absolute or relative amount of time given to the actor's distance manipulation impact effects?

We might also wonder about what we wish "credibility" to mean. The two item "competence" factor ("qualified/unqualified" and "reliable/unreliable") seems to ..,e to be in more strict accord with what consumer and advertising researchers mean by the construct than the other factors reflected. What would the results have looked like with just these two items as the dependent variable measures

A final point, is the issue the relative amount of space in the ad given to the close-up, or is it the actual size of the face in the visual field? For example, when projected on a 4' x 6' screen, where a face may be "larger than life" even though "small" relative to other components of the visual field, would we have the same or different paraproxemic situation as would the same visual image when presented on a '5" television screen?

Regarding the final paper I believe that the issues which Calcich and Blair have chosen to address are very important, but I'm less certain that the empirical work reported sheds as much light on them as the authors suggest.

The introduction sets out these important issues; "...understanding the nature of the perceptual task (in acquiring and using package information), the extent to which individuals differ in their abilities to perform i :, and the correlates of these individual differences...", and the data reported are to address the first two issues.

But the research reported really does not address what ; this or the title implies directly. If the question is, "What is the perceptual task confronting the consumer in the acquisition of package information?", the study does not celL us. Rather, it assumes that what is going on in such a situation is best explained by a "disembedding" construct, and since there is a strong relationship between EFT scores and the particular task imposed, the authors suggest that this assumption is supported.

But it's not clear to me that a food package is the same sort of "organized visual field" that is imposed or measured by the EFT task. Is the information on a food package an organized Gestalt out of which some target component must be identified? Or has not much of this information already been disembedded both from a physical point of view (different types of information are given different colors, type sizes/faces, etc.) and from a perceptual/learning point of view (some consumers may know where to look on the package for certain kinds of information, thus, do not "disembed" what they are looking for from all else on the package)?

Moreover, and as the authors note, that the EFT correlates highly with other measures of cognitive spatial ability makes it impossible to know whether individual differences in this acquisition task are really best conceptually explained by the disembedding skill. A high correlation may simply reflect differences in intelligence, education, product experience, verbal comprehension, visual acuity, or many other variables common to both measures obviously coo numerous to mention. And it is likely that especially in a speeded cask, the correlation between any of these predictors and the dependent variable employed would be heightened. Thus, the external validity of the study turns largely on the question of whether or not the typical consumer's set in package information acquisition is speeded. I don't think we know very much about that, nor about the more general question of the basic task confronting the consumer: how much of what kind of information is being sought, and what format of providing this information is most apt to result in rapid and accurate acquisition, although on the latter point there has been some research as, for example, in the use of graphics and pictorial representations as an aid to the perception of magnitude.

And the "apparent accessibility" of price vs. protein vs. iron vs. servings as a basis from implying construct validity for the results is less persuasive than had some objective (or subjective) measure been employed regarding the accessibility (and embeddedness?) of specific content. The article's conclusion suggests that both position and prominence of information explain its accessibility, yet neither were directly measured. Policy implications would be very different if position and prominence played different roles here.

In sum, the research area addressed here is important and should be continued, especially given that the Feds continue to add to the kind of information which should be printed on the package (and analogously, disclosed in advertising). We would provide a valuable service if we could integrate the construct employed here. disembedding, along with others which undoubtedly play a role. many of which the authors identify, but also including information overload, for example.

Information processing is a central theme in all four papers we have heard. Yet, in the last year or so we have heard from several leaders in our field that information processing research topics are in the decline phase of their product life cycle. So are we hearing the last dying gasps, or are we seeing a new growth phase, or were the leaders wrong in their assessment of the field? Or is the question by definition silly to ask at an ACR Conference?

Whatever the case, I found these four papers to be well worth the reading, and each of them-of the pioneering rather than the "son of Sam" sort which many of us have tired of.

And I think the papers have a good deal in common beyond the general "information processing" label. For example, one could envision an advertisement being created with an eve towards ease of disembedding the information contained therein, presented in a "thinking" mode by a spokesperson shown in extreme close-up, who begins and ends his or her pitch with the right kind of questions.