Processing Advertising Information

John R. Rossiter, N.S.W. Institute of Technology
ABSTRACT - Three papers on "processing advertising information" are discussed. This review contends that the relevant types of processing responses depend on the particular advertising communication model underlying brand choice. The communication model, not the form of the advertisement, determines which processing responses and which brand responses should be measured when assessing advertising effects.
[ to cite ]:
John R. Rossiter (1984) ,"Processing Advertising Information", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 26-28.

Advances in Consumer Research Volume 11, 1984      Pages 26-28


John R. Rossiter, N.S.W. Institute of Technology


Three papers on "processing advertising information" are discussed. This review contends that the relevant types of processing responses depend on the particular advertising communication model underlying brand choice. The communication model, not the form of the advertisement, determines which processing responses and which brand responses should be measured when assessing advertising effects.


My comments on the papers in this session will be facilitated if I begin by outlining an advertising communication model framework developed by Rossiter and Percy (1983). The framework is of major relevance to the three papers.

he Rossiter and Percy framework distinguishes advertising communication models in terms of two universal communication objectives: brand awareness and brand attitude: brand awareness options center on whether the advertising communication objective is to create brand recall or brand recognition. Brand recall is clearly the relevant brand awareness objective for the Federal Express TV commercials, for example, when the brand (here, courier) selection decision is made from an evoked set of recallable brands. Brand recognition, in contrast, is the relevant brand awareness objective far the low-calorie frozen food brand, which must be recognized at the Point-of-Purchase,

The advertising tactics for brand recall and brand recognition differ. Advertisements designed to generate brand recall tend to use repetition of the brand name and, sometimes, mnemonic devices, such as jingles, to increase the

probability of subsequent brand recall. Advertisements designed to produce brand recognition tend to use sustained close-ups of the package, so that it will be easily recognized later by color and name cues.

The main options that concern us in this session are with regard to brand attitude. The options are an two dimensions, forming four quadrants (Figure 1).



One dimension of brand attitude is involvement, or perceived risk, associated with the brand choice. Low involvement brand choice is based on a low involvement brand attitude which is weakly or tentatively held, but polarized positively enough to induce trial. Since there is little cost in making a poor decision, processing of pre-purchase information is not deep (it's basically rote learning), because behavioral (trial) experience is the easiest way to find out whether the tentatively held (advocated) brand attitude holds good.

Apparently, all the commercials in the three studies Federal Express, the low-calorie food, and the nationally advertised brands and, perhaps, even the political commercials in the Shavitt and Brock study - would be categorized as low involvement brand attitude, as indicated in the first figure. This categorization is important because in our theory, it is hypothesized that cognitive responses are not relevant to successful processing of advertisements based on low involvement brand attitude (Figure 2). Hypothetically, it is only in high involvement brand decisions, where brand choice entails substantial economic or psycho-social risk, that cognitive responses (especially personal acceptance of the message claims prior to purchase) are relevant causal mediators of brand attitude



The other dimension of brand attitude is the motivating need that drives or energizes the brand choice into action. Whereas a detailed typology of motivating needs is available in Rossiter and Percy (1983), the summary labels suggested by Wills (1981) will suffice here. Brands bought either for negatively reinforcing motivations in which "informational" (reason why) advertising suggests solutions to consumer problems; or for positively reinforcing motivations in which "transformational" (brand user image) advertising enhances sensory, intellectual, or social appeal of the brand.

Federal Express and the low-calorie frozen food brand are apparently informational. Federal Express touts their speed and reliability as informational attributes that solve urgent small package delivery problems. The low-calorie frozen food brand's attributes offer to solve the problem of gaining weight while eating tasty foods.

This categorization is important because, hypothetically, it is only in transformational (positively reinforcing) advertising that attitude toward the ad per se (Aad) should be relevant. By contrast, in informational advertising, what counts is not whether consumers like or dislike the ad, but rather whether the advertisement contains a usefuL reason why consumers should try the brand. Note that the "hated" commercials for Ajax (white tornado) and Wisk (ring around the collar) mentioned in the Lautman and Percy paper were successful campaigns for informaitonally or negatively motivated brand choices where attitude toward the ads themselves would be theoretically irrelevant.

The two dimensions of brand attitude intersect in the quadrants. Aad should be a relatively large mediator of brand attitude far low involvement brand choices, following Petty and Cacioppo's (1983) distinction between peripheral and central routes to persuasion. Peripheral cues assume major weight in low involvement brand attitude formation and change. Peripheral cues assume minor weight, relative to the consumer's initial attitude and the message's advocated attitude, in high involvement brand attitude formation and change. Note, however, that the quadrant formulation hypothesizes that Aad will assume major weight only for low involvement/transformational brand attitude, where likability of the advertisement is relevant.

With this theoretical background, I will now discuss the papers themselves.


Belch and Belch begin their paper with a comprehensive and interesting review of possible and often contradictory theoretical relationships between humor and processing. One is that humor may increase attention, but not necessarily attention to the main message points. This could be problematic for each of the brand attitude quadrants except low involvement/transformational advertising where attention to "irrelevant" humor could increase a transformational brand attitude through classical conditioning. In this way, relationship one is just a forerunner to relationship three, which postulates humor as a UCS+ in classical conditioning, with the brand as CS.

The second theoretical relationship is the distraction effect. Processing of humorous stimuli may inhibit counterarguing. However, this should be problematic only with, as the authors note, high involvement decisions (where there is motivation to counterargue) for established brands (where the familiar audience has prior information with which to mount counterarguments). Federal Express would not fit this high involvement/informational quadrant - unless the campaign were aimed at high volume courier users (such as corporate users) with negative attitudes toward Federal ExPress.

The final theoretical relationship mystifies me. That is that humor may enhance source credibility. I could see humor enhancing source attractiveness (following Kelman's 1960 distinction between the source characteristics of credibility, attractiveness, and power) and specifically via the likability subcomponent of attractiveness, but not credibility. How might this work? Are funny presenters o funny companies more credible? Nevertheless, for some strange reason, the investigators did observe an increase in source credibility with the humorous commercial, specifically on the objectivity subcomponent of credibility, but not the expertise subcomponent. In Rossiter and Percy (1983), it is argued that source objectivity should only be relevant for high involvement/informational brand attitude so I'm not surprised that this phenomenon did not change brand attitude or brand purchase intention.

The main finding is that a humorous advertisement does not influence brand attitude and brand purchase intention any differently than a series or straight advertisement. This reinforces the conclusion that humor is not a separate communication or persuasion strategy, but is simply an alternative tactical execution that can be employed with any of the brand attitude (quadrant) strategies.

One limitation of the study, as the authors note, is the student sample, who would not be in the usual target audience for Federal Express. However, to the authors' credit, the study measured purchase (usage) intention conditional on having the category need (even though few students would have such a need). Another limitation is the small "n" of humorous and serious ads. I would like to see not just a larger sample of ads, to increase generalizability, but a sample of the use of humor across the four brand attitude quadrants.

If it is granted that Federal Express commercials fall into the low involvement/informational quadrant, then the relevant processing measures would be attention and (rote) learning - not cognitive responses. The relevant communication effects (the more permanent, post-processing responses connected to the brand that guide later brand choice) would be brand recall, brand attitude (versus competing courier services) and brand purchase intention (deduced from brand attitude). I am trying to make the point, here, that effectiveness measures derive their validity from the prior advertising communication model. You can't just use any set of measures and focus on those that significant effects Post hoc. These effects may not mean anything if they're irrelevant to the model and are therEfore irrelevant to the brand choice.

The other ma m finding, that variations on a theme can forestall wearout, is reassuring and consistent with practitioners' conclusions. However, I would contend that wearout of cognitive responses is irrelevant in the low involvement/informational model (see Figure 2 earlier). But wearout of attention or learning - of the brand name, or of the learned attitude that Federal ExPress is the best courier - would be cause for concern if low involvement processing is based solely on attention and learning.


Lautman and Percy's is a finely analyzed study; it is rare that researchers take the care to measure communication effects among different target audiences - in this case, behaviorally (dieting, usage) and attitudinally (brand dispositionally) defined. However, I have several doubts about the theoretical basis of the study.

I want to make a point that perhaps was not obvious when I introduced the quadrant models: this is that advertising effects depend on the advertising communication model underlying brand choice, not on the type of advertisement. A transformational advertisement should not work successfully if the brand choice is informational. This seems to be exactly what the investigators found: ads that focused on the end-benefit (transformational) were less effective than ads that focused on the attributes (informational) of the low-calorie food product.

Besides being informational, the brand choice in this case is low involvement. Thus I would contend that the cognitive response results are interesting but irrelevant from a validity standpoint. The affective Aad measures should only be applicable to transformational brand choice. It may be noted that all previous studies in which Aad contributed to brand attitude have been for transformational brand choices: luxury toilet tissue, soda, and beer. The cognitive Aad measures (cognitive responses) should only be relevant for high involvement brand choice. Low-calorie frozen food would seem to best fit a brand recognition/low involvement/informational advertising communication model Cognitive responses for this model are of questionable validity.


The Shavitt and Brock study on individual versus group viewing of TV commercials is very important. Research an this issue is long overdue. The study is notable, too, for its excellent design and experimental controls.

I have two comments on the study. One comment is theoretical and that is - again - that cognitive responses should not matter in low involvement brand choice, the model an which most of the commercials employed in the study were most probably based. Only attention and (rote) learning during processing should be relevant to brand choice. Attention and learning are sufficient for the acquisition of brand awareness and also low involvement brand attitude.

The final comment is more pragmatic. The "group" condition, in which subjects talked before but not during viewing, would simulate typical family TV viewing. But it in no way simulates focus groups, in which people do talk and otherwise interact during viewing. The finding that cognitive responses (relevant for high involvement brand choice) do not differ in individual vs. "passive" group viewing cannot be interpreted to mean that you can validly test commercials in focus groups. In focus groups, commercials are shown and re-shown, hashed and rehashed, by participants and by the moderator. This simulates nothing in the real world. Focus groups are fine for showing advertisements to develop advertising strategy but they are not the place to test ads once the strategy has been decided.

All in all, this was an excellent session on the processing of advertisements. I would yet issue the call for more thought about the types of processing responses that matter in different advertising situations. I hope that the advertising communication models approach is a start in the right direction.


Belch, G.E. and Belch, M.A. (1984), "An Investigation of The Effects of Repetition on Cognitive and Affective Reactions to Humorous and Serious Television Commercials," in T.Kinnear (ed.), Advances in Consumer Research, Vol. II, Ann Arbor: Association for Consumer Research.

Lautman, M.R. and Percy, L. (1984), "Cognitive and Affective Responses in Attribute-Based versus End-Benefit Oriented Advertising," in T. Kinnear (ed.), Advances in Consumer Research, Vol. II, Ann Arbor: Association for Consumer Research.

Petty, R.E. and Cacioppo, J.T. (1983), "Central and Peripheral Routes to Persuasion: Application to Advertising," in L. Percy and A. Woodside, Advertising and Consumer Psychology, Lexington, Mass.: Lexington Books, 1-24.

Rossiter, J.R. and Percy, L. (1983), Advertising and Promotion Management, Chapter 5 (mimeo), Sydney, Australia: The Sew South Wales Institute of Technology.

Shavitt, S. and Brock, T.C. (1984), "Consumer Research Validity: The Effect of Social Settings on Cognitive Responses to Television Commercials," in T. Kinnear (ed.), Advances in Consumer Research, Vol. II, Ann Arbor: Association for Consumer Research.

Wells, W.D. (1981), "How Advertising Works," paper presented to the 12 annual conference of the Association for Consumer Research.