Advertising communication models


Published on

Published in: Business, Technology
1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Advertising communication models

  1. 1. Advances in Consumer Research Volume 12, 1985 Pages 510-524ADVERTISING COMMUNICATION MODELSJohn R. Rossiter, N.S.W. Institute of TechnologyLarry Percy, HEM/CREAMER, Inc.ABSTRACT -A general structure is proposed for constructing models of "the way advertising works"(advertising communication models). Four fundamental models with a total of eight pairedvariations are identified. These models: assist managerS to set complete advertising objectives,help creative specialists to articulate purpose, and increase the validity of advertising pre-tests.Advertising communication models are theories about "how advertising works." These theoriesor models attempt to explain and describe, at the individual buyer or consumer level, the processby which advertising communicates with and effectively persuades individuals to take action.Managers operate with these theories or models, explicitly or implicitly, whenever they create,approve, or test advertising.Most available theories or models share one of two common faults: (1) they are either singularversions of the hierarchy-of-effects notion (e.g., Colley 1961; Ehrenberg 1974; Howard andSheth 1969; Krugman 1972; Lavidge and Steiner 1961; McGuire 1976; Rogers 1962) whereas itis evident that advertising works in at least several different ways rather than via a singleprocess; (2) or else the theories acknowledge multiple processes but focus inordinately on therole or location of brand attitude as a communication objective (e.g., Ray and Webb 1974; Ray1982; Smith and Swinyard 1982; Vaughn 1981) while ignoring other necessary steps in theadvertising communication process.The purpose of the present article is to provide a new interpretation of previous approaches andto extend the context of advertising communication models to incorporate the other inputs thatadvertising managers need. Firstly, a general structure of the necessary components of anadvertising communication model is provided. Secondly, four fundamental brand attitudestrategies are described which, together with two prior types of brand awareness alternatives,produces a total of eight basic advertising communication models. Thirdly, advertising tactics forthese models are listed. Finally, major implication for the process of pre-testing advertising arediscussed.GENERAL STRUCTURE OF ADVERTISING COMMUNICATION MODELSA complete account of the overall advertising process requires at least six steps (Figure 1). Thelast two of these steps are concerned with marketing objectives, to which advertising mustcontribute, namely: sales or market share, leading to profit for the firm. These will not bediscussed further in this article. [The present article is based on several chapters from aforthcoming book by the authors to be published by McGraw-Hill. A more detailed exposition ofthe points summarized here can be found in those chapters. For the present article, the authorswould like to acknowledge the comments of Robert J. Donovan, then visiting associate processor
  2. 2. of marketing at New York University, Geraldine Fennell of Fordham University, as wellas,research personnel at Ogilvy & Mather/New York and Ogilvy & Mather/Australia.]SIX-STEP SEQUENCE OF ADVERTISING EFFECTSThe first four steps are the province of advertising as a communication process, along with thebehavioral outcome of communication. In order for advertising communication to be successful,the prospective buyers in the target audience must: (1) be exposed to an ad or series of ads in acampaign, via media, (2) process the elements of each at in the intended manner so that theadvertising results in (3) communication effects, connected to the brand, which in appropriatecircumstances produce (4) action, such as purchase of the brand. [As usual we use the term"brand" in a broad sense to include any type of product or service that the advertising is designedto promote. Also "action" can include a variety of desired target behaviors on the part ofdistributors or consumers, such as sales inquiries, visits to retail outlets, and other forms of
  3. 3. purchase-related behaviors whenever purchase is consummated by personal selling or othermarketing inputs.] Our approach postulates a "recycling" sequence of overall steps whereby abuyer may take action, then be reexposed to further advertising and go through the sequenceagain, albeit in a modified state of mind due to purchase or usage experience with the brand(Ehrenberg 1974; Smith and Swinyard 1982). The overall sequence should not be confused withthe hierarchy-of-effects notion, which is essentially a theory about the communication effectsstep. Rather, our approach postulates a "heterarchy" of effects, at both the processing step andthe communication effects step, as will be explained later.An advertising communication model should incorporate all four steps. Essentially, anadvertising communication model sets objectives for each step, and provides strategies andtactical detail on how each step is supposed to lead causally to the next one. From the managers"top down" planning perspective, an advertising communication model therefore consists ofdecisions at four levels:A. BUYER: Target audience action objectivesB. BRAND: Communication objectivesC. AD(S): Processing objectivesD. MEDIA: Exposure planA generic structural checklist for advertising communication models is given in the Appendix.The checklist asks for sufficient detail to enable a comprehensive advertising communicationmodel to be stated while at the same time attempting to be short and explicit enough so thatmanagers will use it (Little 1979). In the succeeding sections of the article, we explain how thechecklist is used to develop an advertising communication model suited to particular advertisingsituations.A further note about the checklist is that the manager is asked to indicate whether the input foreach component of the model stems from research or from judgment. Checking the research boxmeans the manager considers that adequate research supports the input; if not, the manager stillsupplies the input but checks the judgment box. Not only does this ensure that all componentsare addressed, it also highlights areas where specific types of research would be of value to yieldsounder conclusions. In the real world of advertising management, adequate research often doesnot exist nor can it afford to be done. Much planning stems from judgment and one of ourpurposes is to provide some theoretical input that will make these judgments more defensible andbetter reasoned.The checklist is not in itself an advertising communication model; it is just the generalframework. However, what goes into the framework must be based on an advertisingcommunication model, of which we will now outline the main alternatives.
  4. 4. A. Target Audience Action Objectives (Buyer)Step A-1: Target Audience. The first step in constructing an advertising communication modelfor a particular brand and advertising situation is to identify the target audience. In our approach,a target audience is defined behaviorally and attitudinally as the group of people (or households,companies or retailers) from whom sales are expected to come. A target audience consists ofthose people who will be most responsive to advertising.The concept of target audience differs from the broader concept of market segments. Marketsegments are based on the other "4 Ps" in the marketing mix, such as product segments fordifferent end uses, price segments for high and low priced brands, geographic markets fordistribution, or customer sales potential segments for personal selling. A target audience foradvertising may be drawn from people within a market segment or across market segments. It istheir vulnerable behavior and attitude toward the brand that draws them together as a targetaudience for advertising.Increased sales through advertising can come from one or more of four prospective targetaudiences: (1) new category users -- who can be induced to try the product category via ourbrand, e.g., IBM personal computers; (2) brand loyals, who can be induced via new users to usemore of our brand than they use at present, e.g., Arm & Hammer baking soda; (3) brandswitchers -- who can be induced to switch to our brand more frequently than they do at present,e.g., the Coke-Pepsi battle; and (4) other-brand loyals - who can be converted to our brand fromloyalty to another, e.g., Ralph Laurens Polo shirts inroads on Izod Lacostes previously loyalbuyers.A particular advertising campaign rarely addresses more than one target audience. To do sowould be requiring too much of what should be a tailored communication effort. However,managers will sometimes establish a primary target audience, to whom the communicationcontent is mainly tailored, and one or more secondary target audiences, who will be affected butto a lesser degree. A campaign targeted to other-brand loyals, for example, is often undertakenwith the provision that the campaign does not alienate but rather maintains the behavior ofcurrent brand loyals.Step A-2: Decision-Maker. Unless purchase of the brand is an entirely personal decision, themanager must then identify the decision-maker or decision-makers within target audiencehouseholds or companies to whom the advertising is directed. In our checklist the manager isasked to nominate the decision-maker by role and by action (Webster and Wind 1972) as towhether the target audience individual should: propose the brand for consideration (initiator),recommend it (influencer), make the final decision (decider), order or buy it (purchaser), or useor consumer it (user). The communication content of the advertising will differ according to thedecision-maker target, e.g., mens shirts such as Hathaway being advertised to women asinfluencers, or childrens products such as Fisher-Price toys being advertised to parents asdeciders. For entirely personal purchases, the individual occupies all five roles and is the solitarydecision-making and action target.
  5. 5. Step A-3: Personal Profile. Target audiences are behaviorally and attitudinally defined withregard to the brand. (A better term for the latter would be communication effect-defined, since itis not only attitude that determines action, as we shall see in the next section.) In this last sectionof Part A of the checklist, the manager is asked to specify the target audience decision-makerscurrent rate of behavior in terms of frequency and volume, as well as the future target ratedesired as an action objective for the advertising.Also requested are several personal profile variables: media exposure patterns, to help mediaplanners reach decision-makers directly; demographics, to help copywriters portray the decision-maker; psychographics, to further help copywriters in writing "to" the decision-maker; and anestimate of the decision-makers likely "mental state" during media exposure, which can beuseful to copywriters to determine the style of ads, e.g., for tired late-night TV viewers or harriedcommuters reading newspapers.Again it should be emphasized that although research may not be available for all these inputs,they will be tacitly assumed anyway in the process of advertising creation. That they areidentified as judgments forces these aspects to be considered and can highlight points at whichaudience research may be needed.Already, therefore, we see emerging the alternative content decisions that need to go into theparticular advertising communication model via the general checklist. However, it is not yetreasonable to refer to these as alternative advertising communication models, since thesedecisions mainly refer to alternative targets of the communication rather than to alternativecommunication processes. The next section of the checklist, Part B, differentiates thefundamental advertising communication models via communication processes.B. Communication Objectives (Brand)Our approach utilizes five advertising communication effects (see Table 1 for definitions). Inorder to take action such as purchase of a brand, a target audience individual must: (1) have thecategory need, i.e., be "in the market" for the product class; (2) be aware of the brand as anoption within the class; (3) have at least a tentatively favorable brand attitude toward it; (4)intend to buy it, although this intention may be quite latent or subconscious until the individual isin the purchase situation; and (5) experience no barriers to purchase facilitation, such asdistribution unavailability or inability to meet the price or pricing terms.
  6. 6. THE FIVE BASIC COMMUNICATION EFFECTS DEFINEDThe five communication effects may appear to resemble and perhaps to extend the notion of ahierarchy-of-effects, and it would be surprising if they didnt, given the widely acclaimed facevalidity of the hierarchy notion. However, there is no assumption that they occur in anyhierarchical order, and indeed they may be generated simultaneously or at different times andwith varying degrees of strength in a prospective buyers mind. For example, an individual mayknow all about Preparation-H, but not experience the first communication effect, category need,for a hemorrhoid remedy until later in life. Similarly, an individual may have the category needand experience no barriers to purchase facilitation, but make an "impulse" selection on the brandwherein brand awareness, brand attitude, and brand purchase intention are created by point-of-
  7. 7. purchase advertising at the last minute. There is no hierarchical necessity although thecommunication effects may in some cases be experienced at full strength in the numerical ordershown.Not all communication effects are necessarily communication objectives for a particularadvertising campaign. There are, however, two effects that are universal objectives -- brandawareness, and brand attitude. [One could argue that brand awareness is the single universalcommunication objective. For example, the one-word ad "Perrier" may effectively increase ormaintain brand awareness and cause purchase among target audiences who already have a fullylearned favorable attitude toward the Perrier brand. However, most managers would concedethat, although the brand attitude content is by implication only, the brand attitude objective ofeven this one-word ad is to maintain the target audiences favorable attitude toward Perrier byreminding the target audience of their attitude and thus protecting against competitive attitudeinterference. It is therefore meaningful to regard both brand awareness and brand attitude asuniversal communication objectives.] It is these two communication objectives that differentiateadvertising communication models. All advertising campaigns are aimed at maintaining brandawareness (if not to increase it) and at maintaining brand attitude (if not to change it). The otherthree communication effects are optional as objectives.Step B-1: Category Need. Category Need is an optional communication objective for aparticular campaign. In the checklist, as with all the communication effects, the manager is askedto mike this decision explicitly. Category need can be ignored as an objective if thiscommunication effect is at full strength in the prospective buyers mind. For example, Coca-Colaprobably does not have to address the cola category need in advertising Coke; whereas inadvertising for Diet Coke, the category need for diet cola may require reminding, or selling, theother two options in the checklist. Discontinuous innovations (Robertson 1971) invariably haveto "sell" the category need in their advertising; new brand entries in a well known category mayhave to remind the target audience of the category to which the brand is aspiring; but establishedbrands rarely have to address category need unless, as Campbells Soup did recently, they aretrying to stimulate category sales of which they reap a large share.Step B-2: Brand Awareness. Brand Awareness is a necessary communication objective. Indeed,without brand awareness being experienced at some point prior to the purchase decision, thebrand cannot be bought. Brand awareness is poorly conceptualized in most advertising plans.Only Bettman (1979) and a few others have come close to identifying what is required to setvalid brand awareness objectives. Many advertising agencies and their client companies continueto rely blindly on top-of-mind brand recall, when many brands are in fact chosen by brandrecognition at the point-of-purchase, not by recall prior to the purchase situation.The main decision concerning the brand awareness objective (and thus the brand awarenesscomponent of the advertising communication model) is whether the target audiencepredominantly enters the brand purchase decision via brand recall or brand recognition. This maydiffer by target audience for the same brand. For instance, R-C Cola brand loyals maypredominantly plan to buy that brand, by recall; whereas R-C Cola brand switchers maypredominantly notice it at the point-of-purchase as one of the alternative cola brands that theyswitch between, by recognition. The checklist does provide an option of "both," to be selected
  8. 8. only when a substantial proportion of the target audience is known or judged to use thealternative method. This should not be used as a "cop-out" option by managers because, as weshall see, the advertising communication tactics differ markedly depending on whether theobjective is brand recall or brand recognition. Complexity is increased if both brand awarenesscommunication objectives must be addressed in a single advertisement, and furthermore, theexposure (media) schedule differs too.Table 2 presents a summary of the advertising tactics recommended for the respective types ofbrand awareness. [It should be clearly stated that the tactics recommended for brand awarenessand brand attitude are hypothetical at this point. Most were synthesized from a thorough readingof various advertising sources, although there is a good deal of original speculation. If these apriori tactical recommendations stimulate research to test or challenge them, they will haveserved their present purpose.] A full rationale for each tactic is given in Rossiter and Percy(1983) and the rationales can only be summarized here. Brand recognition is a much easierresponse to learn than brand recall. The main brand recognition tactic is to emphasize the
  9. 9. package and the name visually in the advertising. This prescription is often ignored or slighted,and it may be noted that it renders radio a very poor medium for generating brand recognition.Media weight can also be reduced after initial learning of brand recognition, since it is arelatively easy response to maintain (see also Krugman 1972). However, the reduced mediaweight tactic may be overruled by the brand attitude strategy, as explained in the next section.Brand recall is a considerably more difficult form of brand awareness to achieve. As suggested inthe table, the key is not simply repetition of the brand name, but repetition of the association ofthe brand name with the category need. Brand recall does not occur in a vacuum. Rather, brandrecall is a "response" to a category need "cue" and it must be learned in association with that cue.(Note that in brand recognition, the process is reversed: brand recognition is the cue and categoryneed is the response. For example, you see the Fab package and remember that you needdetergent.)The category-brand association can be made in what we have called the "main copy," whichincludes headlines, tag lines, and also the copy claims themselves, to ensure repetition. The otherbrand recall tactics are explained further in Rossiter and Percy (1983) where it is shown thatpersonal reference increases brand recall by personalizing the association; that bizarre executionsare a very effective associative vehicle as long as they do not detract from the brands "image";and that jingles, if they catch on with the target audience and elicit spontaneous rehearsal, are avery effective mnemonic device for increasing brand recall because music offers greateropportunity of unique encoding than words heard o; read in cows unaccompanied by music.SteP B-3: Brand Attitude. Brand Attitude is the second necessary communication objective.Brand "attitude" as a communication effect is defined in our approach a little differently from theusual academic definition and more in line with the way most practitioners use the term (Figure2). Academic definitions tend to follow the Fishbein type of definition (e.g., Fishbein and Ajzen1975) in which attitude is conceptualized as overall affect toward the act of buying the brand.However, as Wyer (1974) has argued, overall affect is simply one class of beliefs about the brand-- that "the brand (or the act of purchasing the brand) is likable." Evaluation or likability in thissense has very limited motivational status; only a few products or brands are bought merelybecause they are liked. Rather, as Fennell (1975, 1978) has cogently pointed out, people buybrands to fulfill one or several of a relatively finite set of motivations. It should be noted thatthese motivations are not Just "benefits" but rather underlying energizing mechanisms of humanaction to which benefits contribute in a secondary manner.Our approach to brand attitude builds on Fennells and identifies eight basic motivations to whicha brand may be attitudinally connected: problem removal, problem avoidance, and normaldepletion (the negatively originated motivations); and sensory gratification, intellectualstimulation, and social approval [Social approval is meant in the sense of social rewards, whichare positively motivating. If social approval is sought because of personal anxiety, it comesunder problem removal and is negatively motivating.] (the positively originated motivations).Brand attitude is conceptualized as a summary belief (an overall evaluation) linking the brand toa motivation.
  10. 10. The checklist item 3-3 for brand attitude is divided into two sections. Section 3(a) is quitestraight-forward as it simply asks the manager whether the brand attitude objective is to: create anew attitude from zero; increase a currently favorable attitude; modify an existing attitude(connect the brand to a new motivation); maintain a current attitude; or change a currentlynegative attitude across the neutral point into the positive zone, which is generally a much hardertask than an increase within the positive zone.Section 3(b) for brand attitude identifies the brand attitude strategy that will meet the brandattitude objective. It is here that advertising communication models become truly differentiated.Referring to Figure 2, it can be seen that brand attitude -- from an advertising communicationstandpoint -- has two strategic components: (A) correct emotional portrayal of the motivation,and (B) adequate logical support for perceived brand delivery on the motivation. These are ofcourse the affective and cognitive components of attitude. In advertising terms, to borrowpsychologist George Mandlers words (1979), the crux of the attitudinal approach is "heat" and
  11. 11. "light." An effective advertisement engages the prospective buyers emotions and enlightens himor her about the brand.The emotional (motivational and energizing) and cognitive (directional) components of brandattitude form the basis for a four-fold typology of brand attitude strategies (Figure 3). On theemotional dimension, we have borrowed the terms suggested by Wells (1981) to categorize thepredominant type of motivation governing purchase of the brand. Brand attitude strategies can beclassified as relying primarily on either an "informational" (reason why) motivation or a"transformational" (brand-user image) motivation. In our approach, these are not just nominaldistinctions. Information strategies apply when the brand is linked to one of the five negativelyoriginated motivations: problem removal, problem avoidance, incomplete satisfaction, mixedapproach-avoidance, or normal depletion. Transformational strategies, in contrast, apply when
  12. 12. the brand is linked to one of the positively originated motivations: sensory gratification,intellectual stimulation, or social approval.On the cognitive dimension, we utilize the concept of involvement or perceived risk associatedwith buying the brand. Involvement is categorized according to the economic theory developedby Nelson (1970), which classifies the brand purchase decision as either "low involvement"Or"high involvement" (search/conviction required). Related developments of this conceptualizationof involvement can be seen most directly in the theory advanced by Ehrenberg (1974) and alsoLutz and Reilly (1974), Smith and Swinyard (1982) and Finn (1982). The perceived risk leadingto involvement can be either economic, especially for products sold informationally: orpsychosocial, especially for products sold transformationally (Bauer 1967; Peter and Tarpey1975).The involvement classification applies to a particular brand, for a particular target audience. Asindicated in Figure 3, brand purchase decisions in some product categories tend to involve solittle economic and psychosocial risk that it is meaningful to speak of a "product" as being lowinvolvement. However, the classification depends on the target audience for the brand. Forexample, the brand loyal buyer of a Rolls-Royce automobile, an ostensibly high involvementproduct, is essentially making a low involvement purchase decision; likewise, the other-brandloyal buyer of Tylenol, an ostensibly low involvement product, would be making a highinvolvement decision in switching to the aspirin-containing Bayer brand. Involvement, and thusthe cognitive classification of brand attitude, must be determined for the brand and for theparticular target audience.The brand attitude strategy classification produces four fundamental advertising communicationmodels which, when combined with the two brand awareness alternatives described earlier,produce a total of eight models. The tactical recommendations for the four brand attitude strategyvariations of these models are summarized in Table 3. Space limitations again preclude adetailed exposition of these tactics (Rossiter and Percy 1983) but several important distinctionsare reviewed next.Authentic emotional portrayal of the motivation. In the transformational models, emotionalauthenticity is of paramount importance. Indeed, in the low involvement/ transformational motel,positive emotion is the sole "benefit" associated with the brand, e.g., the exuberant portrayal ofsensory gratification in the "Coke is it" commercials. In the informational motels, correctemotional portrayal, which usually follows a negative emotion to positive emotion problem-solution format, is also important, but less so relative to the cognitive component. For instance,the efficacy with which Wisk detergent is shown (and perceived) to remove shirt collar stains isrelatively more important than the particular Portrayal of the stain problem itself.
  13. 13. At likability (attitude toward the ad). A second distinction, also related to the informational-transformational advertising, is that in transformational advertising, it is essential that the targetaudience like the ad itself, regardless of its opinion of the brand. This is not true forinformational campaigns, where sometimes "irritating" commercials get the point across moreeffectively, e.g., the notorious "Mr. Wipple" campaign for P&Gs Charmin which effectivelycommunicated that Charmin toilet tissue is squeezably soft. It is noteworthy that all publishedstudies in which attitude toward the ad has been shown to contribute significantly to attitudetoward the brand have employed what appear to be low involvement/transformational products:beer (Rossiter and Percy 1980); facial tissues (Mitchell and Olsen 1981); and soda (Shimp andYokum 1982).Following the low involvement route to persuasion demonstrated by Petty and Cacioppo (1979)and summarized by Petty and Cacioppo (1983), extraneous elements of the message such asexecutional likability assume much greater weight in low involvement attitude formation andchange than they do in high involvement attitude formation and change. This is also reflected inthe models.Adequate logical support for perceived brand delivery on the motivation. A third distinctionturns now to the cognitive component of brand attitude strategy in the models. In the lowinvolvement motels, claims stated (informational) or visually implied (transformational) aboutthe brand need only be tentatively believed, to a degree that is adequate to prompt trial of thebrand (cf. Maloneys 1962 concept of "curious disbelief"). This means that claims in low
  14. 14. involvement advertisements should be stated or implied as extremely as possible. [Extremeclaims are of course subject to legal substantiation. However, on the verbal site, there are plentyof psycholinguistic devices for delivering perceptually extreme claims (Harris et al. 1980) suchas, "You can t beat Crest for fighting cavities" (but you can equal it), and on the visual side,perceptually extreme claims are often made effectively by implication rather than explicitly andmay avoid legal challenge (Rossiter and Percy 1981).]Claims in low involvement advertising have to be learned, rather than fully believed andaccepted, and a more extreme claim will be more likely attended to and learned, and more likelyto induce trial to test the claim via usage experience. [Smith and Swinyard (1982) attempt toaddress the low involvement trial induction phenomenon with their concept of lower order andhigher order beliefs. However, this concept confounds the degree of belief with the confidencewith which it is held (see also Wyer 1974). Beliefs, in low involvement attitude formation, areextremely polarized buy only weakly or tentatively held, subject to post-trial usage experience.]This has been labelled the "ask more, get more" tactic of attitude formation and change(McGuire 1969).In contrast, it is only in high involvement models that the textbook attitude principles of "latitudeof acceptance/rejection" and the careful tailoring of claims to the target audiences prior or initialattitude are truly applicable (Houston and Rothschild 1977). High involvement purchasedecisions dictate that advertising claims must be believed and accepted before purchase actionwill be considered. Accordingly, the cognitive tactics for the high involvement models(especially the high involvement/informational model) are much more detailed than for the lowinvolvement models.Media exposure schedule. A fourth distinction in the models relates to the exposure schedule fordifferent types of advertising (Wells 1981). Generally speaking, informational campaigns have towork immediately -- the reason or reasons why the target audience should buy the brand shouldbe evident and fully learned (low involvement) in one or two exposures. In contrast, intransformational campaigns, brand attitude continues to build with multiple repeated exposures(Zajonc 1980) until a peak or asymptote is reached; thereafter, continued exposures are needed toreinforce the attitude, especially in low involvement/transformational advertising, where attitudetoward the advertising is a significant contributor to attitude toward the brand.Quadrant dimensions as dichotomies in practice. The four brand attitude strategy "quadrants" arepostulated to represent functionally distinct models. Although it may be contended that the twodimensions that form them are continual, thus rendering the classification artificially extreme, inpractice it is comparatively easy to determine, especially through research, whether most of thetarget audience regards the brand purchase decision as within the realm of "try-it-and-see" (lowinvolvement) or whether they would have to be convinced first before buying (highinvolvement). In rare cases when the target audience exhibits mixed levels of involvement orwhen a single ad is aimed at two target audiences, one of which is high-involved and the otherlow-involved, then high involvement tactics should be assumed. High involvement models arethe more conservative or "safer" models buy they also require more complex and carefulexecution and, according to the theory, a high involvement model will be less effective if usedwhen the brand decision is actually low involvement.
  15. 15. Similarly, by identifying the main motivation, it is comparatively straightforward to decidewhether the predominant executional focus of the ad should be informational, ortransformational. Truly mixed cases on the informational-transformational dimension most oftenoccur with high involvement/transformational advertising that has to provide information so thatthe prospective buyer can "rationalize" before accepting the transformation. Mercedes-Benzautomobiles, for example, are purchased primarily because of social approval motivation, buttheir ads provide plenty of performance information with which the prospective buyer canrationalize the brand choice. Mixed cases of informational and transformational strategies can behandled by including both sets of tactics, although this necessarily poses a more complex task forthe advertising to accomplish.Step B-4: Brand Purchase Intention. Brand purchase intention is usually a purchase intention butcan refer to any intended action targeted by the advertising. Industrial advertising, for instance,often targets sales inquiries as the intended action rather than purchase directly. Brand purchaseintention is an optional communication objective. The option refers to the difference between"hard sell" (intention-induced) advertising and "soft-sell" (intention-deduced) advertising.In hard-sell advertising, the target audience should form a conscious, immediate intention to actat the next purchase opportunity. In the tactical extreme, promotion offers and exhortations to"act now" are included in hard-sell advertising. The hard-sell approach, intended to generate animmediate purchase action intention, mainly is used with informational advertising. However,high involvement/transformational advertising, e.g., for vacation packages, may also employhard-sell to try to stimulate immediate intentions. The manager has to decide which approach ismore suitable even though there is a reasonable correlation with the communication modelselected.In soft-sell advertising, the target audience does not form an immediate conscious intention topurchase or take action with regard to the brand. An intention to act is only experiencedconsciously, later, when it is triggered for a fraction of a second in the choice situation (Krugman1965). At this time, the prospective buyer "deduces," from brand awareness and brand attitude, amonetary but effective intention to act. We suggest that low involvement/transformationaladvertising frequently operates in this way to trigger purchase. If so, it is pointless to try tomeasure purchase intention prior to the event.We do not, however, suggest that the hard-sell and soft-sell options represent a sufficientconceptual difference to warrant these being designated as separate advertising communicationmodels. Rather, they are seen as close complements to the brand attitude model incorporated inthe advertising.Step B-5: Purchase Facilitation. Purchase Facilitation, the fifth communication effect, is alsooptional as an objective. It applies when the advertising must overcome a barrier to purchaseresulting from the remainder of the marketing mix. Barriers may take the form of a perceivedproduct problem, e.g., Tylenol and tamper-proof containers; a perceived high price problem,addressed in purchase facilitation by the inclusion of easy payment terms or a high quality appealto offset the price via perceived value; or an actual distribution problem, overcome byexclusivity appeals such as "not available everywhere" or by offering home delivery.
  16. 16. Purchase facilitation is not usually a communication objective for national brand advertisingalthough it frequently occurs in retail and direct mail advertising. If purchase facilitation is notrequired, the manager so indicates on the checklist and omits this objective.C. Executional Processing (Ad)Communication models can, if desired, cease detailed exposition with the establishment ofcommunication objectives. The creative department and the media department, respectively, canbe left to construct specific advertisements and a media plan that will meet the communicationobjectives. However, the eight advertising communication models (differentiated by the twotypes of brand awareness and the four brand attitude strategies) require concomitant variations inthe way ads are processed and the schedule on which they are best delivered. Accordingly, a fulladvertising communication model will specify the two remaining levels: processing andexposure.Processing refers to the prospective buyers immediate responses to elements of theadvertisement. Following the popular cognitive metaphor, processing responses occur in short-term or "active" memory, and are "ad-centered." The intended outcome of processing is toproduce communication effects in long-term or semipermanent memory, which are "brand-centered." Processing is not exclusively cognitive, however. In our approach, processingincludes: (1) attention to all elements of the advertisement relevant to respective communicationobjectives; (2) physiological responses to emotional elements of advertisements pertaining to themotivation portrayal; (3) rote (often subconscious) learning of associations between elementspertaining to brand awareness and, if applicable, low involvement brand attitude; and (4)cognitive acceptance of elements pertaining to, again when applicable, high involvement brandattitude, category need, induced brand intention, and purchase facilitation.In our work with the checklist, we have found it both highly instructive, and very useful as ameans of guiding later diagnostic searches following ad testing, to have the creative personnel onthe account complete the processing section of the model. Here is what they are asked to do. Itshould be noted that the checklist skips the attention phase of processing, as this would be verytedious to complete and is implicit in the points listed in the remaining three steps of processing.Step C-1: Emotional Portrayal. The copywriter and the art director are asked to describe theexact emotion, or sequence of emotions, that they are trying to elicit in the audience inconjunction with the brand motivation. For example, an informational advertisement addressingthe problem removal motivation might employ the sequence: disappointment + hope + relief. Atransformational advertisement addressing the sensory gratification motivation might simplynominate elation as the emotion or "tone" to be portrayed.Including emotional descriptions in the advertising communication model used for a brandmakes explicit an aspect of advertising effectiveness that is almost always neglected bymanagers who focus only on approving written copy. As we have seen in the definition of brandattitude, the emotional component is a necessary complement to the cognitive or brand benefitdelivery component. For transformational campaigns in particular, it is the emotional executionof advertisements that differentiates the great from the average campaign. Furthermore, an
  17. 17. explicit listing of emotions to be portrayed makes the task of ad testing easier becauseappropriate emotional adjectives can then be selected with which to gauge audience reactions tothe ad.Step C-2: Points to be Learned. In our approach (Rossiter and Percy 1983), two communicationeffects require only rote learning during processing. These are brand awareness, and lowinvolvement brand attitude. In brand awareness processing, the target audience must learn theassociation between the category need and the brand, producing a subsequent recall response orrecognition response as appropriate. [In brand recall, the prospective buyer experiences thecategory need first then recalls the brand(s) associated with it. In brand recognition, theprospective buyer recognizes the brand first, e.g., in-store, then associates (actually recalls) thecategory need whereupon a mental check of category need status is made. Recall and recognitionthus reverse the stimulus and response roles of the associative link between category need andbrand awareness.] The copywriter is asked to indicate those elements of the ad that tie in thecategory need with the brand - following the tactics for brand recognition and brand recalldescribed earlier.In brand attitude processing, the target audience must learn further associations between thebrand and specific benefits (related to the motivation). In low involvement attitude formation andchange, rote learning is all that is required. The copywriter is asked to identify the specificmessage points to be learned -- be these in the verbal copy or implied in the visual portrayal.Step C-3: Points to be Accepted. For all other communication effects apart from brand awarenessand low involvement brand attitude (and note that this means that many low involvementcampaigns are based strictly on learning), the in-ad elements pertaining to these effects have tobe consciously accepted by the target audience. The perspective buyer of the IBM PersonalComputer, for example, has to : accept the category need for a personal computer; accept that theIBM Personal Computer is a good brand; accept an action intention, such as visiting an IBMstore or calling for a demonstration; and accept that the price or payment terms aid purchasefacilitation.Acceptance signifies personal agreement with the relevant advertising elements or messagepoints. Acceptance is manifest consciously in the phenomenon of "cognitive responses"(Cialdini, Petty and Cacioppo 1981; Hovland, Janis and Kelly 1953; Perloff and Brock 1980;Wright 1980) which seem to be prerequisite for shifts in high involvement attitudes and, wewould argue, for shifts in these other non-rote communication effects.The copywriter does not, of course, have to be knowledgeable about acceptance or cognitiveresponse theory. All that is required is a careful listing of the message points -- either stated orimplied visually or verbally in the ad -- that the target audience is supposed to accept. It isstraightforward to do this, especially if the points are categorized in terms of the communicationobjectives they are intended to address. In working with the checklist, categorization of messagepoints in terms of specific communication objectives to which they relate has been found toilluminate the copywriters purpose rather than hinder it. The results of the exercise are ofimmense value to managers and researchers because they help to reduce the mystery (but not themagic) of what creatives are doing.
  18. 18. The four types of responses in processing -- attention, emotional responses, learning, and (ifappropriate) acceptance -- are also "heterarchical." For an ad to meet a communication objective,each element relevant to the communication effect concerned must be processed. Because adsconsist of multiple elements, this means that a person processing the ad may be making, forexample, an acceptance response to one element while simultaneously seeking attention tofurther elements. Just as in the communication effects step earlier, there is no set hierarchs ofeffects in the processing step.Nor, except in the rare case of a new brand being launched into a virtual mental vacuum, is thereany necessity for the four overall steps themselves to form a hierarchy, even though they areshown this way for convenience (Figure 1 earlier). For example, communication effects (Step 3)may be salient in the audiences mind before the ad is processed (Step 2). Similarly, behavioralaction (Step 4) may lead to deliberate reexposure to advertising for the brand (Step 1) as in thewell-known dissonance reduction sequence. The whole series of steps should be regarded as apotentially inter-looping mental heterarchy, punctuated by occasional behavioral acts such aspurchase.Returning now to the steps in the checklist, there is another frequently used tactic or set ofelements that affects processing -- the use of a presenter.Step C-4: Use of a Presenter The use of a presenter (or endorser) in ads is another decision thatoften confronts managers. In our approach, endorsement is not a separate strategy; ratherpresenters or endorsers can be used, with any of the advertising communication models, toincrease Processing of specific communication effects that need strengthening. Presenters shouldbe considered to "boost: communication effects when a "standard" advertising execution fallsshort of attaining the communication objectives.The processing checklist for presenters identifies presenter characteristics that relate to variouscommunication effects and particularly to the four brand attitude models. The VisCAP Acronym(an extension by McGuire 1969 and Percy and Rossiter 1980 of Kelmans 1958 apProach)summarizes the major presenter characteristics.Visibility. Visibility or recognizability, the strong characteristic of celebrity presenters, is likelyto heighten brand awareness -- notably brand recall, although the advertiser must be careful thatthe presenter does not obscure the presentation of the brand itself. The processing mechanismassociated with visibility, of course, is the expectation that a highly visible presenter will drawattention to the ad and thus make the brand more visible; that is, the presenter will increase brandawareness.Credibility. Credibility consists of two characteristics, expertise and objectivity. A presenter canbe perceived as expert without being objective, and vice versa. Expertise is relevant toinformational communication models, both low and high involvement, because perceivedexpertise enhances attention to and learning of (low involvement) or acceptance of (highinvolvement) information presented in support of brand attitude. Objectivity, on the other hand,is mainly relevant to the high involvement/informational model. This is because high
  19. 19. involvement claims have to be believed (accepted) whereas low involvement claims are moreeffective if they stretch credibility and are stated more extremelY and thus less effectively.Attraction. Attraction or attractiveness as a presenter characteristic also consists of twocomponents, likability and similarity. The attraction the presenter holds for the target audience isof primary importance for the transformational motels, where the advertising content most offerspositive stimuli to enhance the positive motivation. Likability is mainly relevant to the lowinvolvement/transformational motel, where everything about the ad must be likable, includingthe presenter. SimiLarity (to the target audience) is a high involvement/ transformational factor,where the target audience must not like the at, but identify with the brand presentationpersonally.Power. Power, or perceived authority, is not a widely employed presenter characteristic.However, it is relevant in hard-sell campaigns where the purpose is to induce immediateintention to act. Public service campaigns on safety and health (problem avoidance motivation)frequently use powerful, authoritative presenters to good effect so that the message will beaccepted almost as a commandment or a duty rather than as a message the audience can freelyaccept or reject. Pharmaceutical ads, too, can imply medical or clinical authority by usingpowerful presenters.The use of a presenter, therefore, is not an arbitrary decision. Presenters must be selected so thattheir salient personal characteristics are those which amplify audience processing of elementsrelevant to the particular communication model through which the ad is designed to operate.D. Exposure (Media)A full advertising communication model also addresses exposure of advertisements via themedia plan. Advertising communication models have inherent implications for media selectionand media scheduling. As such, they dictate the overall media strategy of a campaign whileleaving tactical details and specific considerations to the media specialists.Step D-1: Media Selection. Media planners usually select a primary medium for a campaign,then supplement this with one or more secondary media to reinforce particular communicationobjectives or to reach prospects omitted in primary media coverage. The strategy checklist inSection D-1 asks the media planner to indicate these choices. As shown in Table 4, it is essentialthat the media planner selects media that are compatible with the particular communicationmodel upon which the campaign is based. Particular models eliminate or severely limit the use ofcertain models. [Rossiter and Percy (1983) discuss creative solutions for overcoming many ofthese limitations. However, the solutions are typically elaborate and expensive. and should onlybe used when other considerations strongly dictate use of the medium. The discussion here isconfined to conventional use of the respective media.]
  20. 20. The respective models eliminate or restrict media selection according to: (i) visual capacity(radio is eliminated for brand recognition models where packages must be shown): (ii) colorquality and cost (newspapers are limited for new package brand recognition and fortransformational campaigns); and (iii) verbal message capacity (the "short" media of TV, ratio,and outdoor are virtually eliminated for high involvement/informational campaigns). A furtherlimitation is imposed by (iv) high effective frequency, which is discussed in connection withmedia scheduling.Step D-2: Media Scheduling. Media planners are increasingly recognizing the importance ofbasing media schedules on effective frequency calculations. Effective frequency (e.g., Naples1979) is based on the estimated minimum number of times an individual target audience membermust be exposed -- within a purchase cycle -- in order to induce purchase of the brand (see D-2).
  21. 21. The effective frequency estimate will vary with the target audience; for instance, brand loyalswill require less frequency than brand switchers, other-brand loyals, and new category users,usually in this order (Rossiter and Percy 1983).Effective frequency also depends on the advertising communication model. Table 4 (earlier)indicates limitations (in parentheses) when high effective frequency is required. Brand recallcampaigns typically demand high effective frequency to instill the brand or to protect againstcompetitive brand learning. Hence, in the checklist, the media planner is asked to estimate thefrequency per purchase cycle used by the leading competitor, so that a "dominance" schedule canbe planned. Low involvement/transformational campaigns also demand high effective frequency(Wells 1981) to generate and maintain brand attitude or "image."When all factors inherent in the advertising communication models are considered, the mediaplanners choice of media for a particular campaign becomes quite circumscribed. The finalchoices can be inferred from the table. To provide examples from just two of the eight basicmodels: brand recall/low involvement/informational campaigns are best suited to TV, radio, ornewspapers; whereas brand recognition/high involvement/transformational campaigns are bestsuited to TV, magazines, outdoor. or direct mail.To complete the exposure plan checklist, the media planner is asked to list other factors such ascontinuity, seasonality, and geographic market considerations that will affect the plan (Step D-3).These do not depend on advertising communication models and are not discussed hereIMPLICATIONS FOR AD TESTINGThe eight advertising communication models have crucial implications for ad testing. There is noone way in which ads work; therefore, there is no single procedure that can validly test all typesof advertising. Failure to appreciate this fact has led to endless, fruitless debates among variousproponents (such as syndicated test service providers) of particular advertising testing models.[Preston (1982) has made a heroic effort in proposing an advertising communication model thatattempts to integrate most of the major syndicated measures. However, this seems backwardsoperationally, in that correct concepts should surely precede consideration of extant measures.Moreover, Prestons is yet another singular model and, as we have seen, singular models cannotaccount for the different ways in which advertising works.]The eight advertising communication models differ on three basic dimensions: (1) brandrecognition versus brand recall; (2) low involvement versus high involvement brand attitudestrategy; and (3) informational versus transformational brand attitude strategy. These threedichotomies require different ad testing measures, as summarized in Table 5. When thesemeasures are recombined into the eight models, it can be seen that eight different testingprocedures emerge. Rather than discuss all eight, which are relatively easy to piece together, wewill examine just the three dimensions of difference.
  22. 22. Brand Recognition Versus Brand RecallIt should be obvious that if the brand is chosen in the real world through a brand recognitionprocess, as with many supermarket products, then the standard brand recall measure usuallyobtained in conjunction with half-hour delayed advertising recall (e.g., ASI, McCollum-Spielman procedures) is irrelevant. The brand awareness objective with brand recognitionmodels is to get the package recognized in a typical competing package display setting. Anappropriate visual measure, such as a tachistoscope test of a multi-brand display, should be used.Brand recognition is not correlated with brand recall (Thorson and Rothschild 1983) and so thesubstitute use of a brand recall measure for a brand recognition objective is inappropriate andmisleading.If the correct model is, however, brand recall, then there are two main brand awarenessmeasurement details to be considered. One is the cue to be used to elicit recall; this shouldcorrespond with the category need that prompts brand recall in the real world, and it depends onhow consumers define the category and not necessarily on how the marketer defines it. The otheris the number of recall responses that are allowed; this number should correspond with thetypical evoked set size from which consumers select the brand in the real world.Also, because recall usually declines with time (mainly due to interference), brand recall in an attest situation can provide only a relative measure of brand recall awareness. It can provide anabsolute measure only when the exposure-to-test interval is similar to the average real worldinterval. Brand communication effects are conditional on brand awareness, and a brand recallmeasure (or a brand recognition measure) taken very soon after exposure in a test situation canproduct a spuriously inflated estimate of advertising effectiveness.Low Involvement Versus High InvolvementThe testing of low involvement ads differs from the testing of high involvement ads in severalways. Firstly, the nature of processing is different. Low involvement brand attitude registrationdepends on the perceived (learned) rather than the believed (accepted) message. Thus, the correcttype of processing measure for the low involvement models centers on correct learning of theintended message. A suitable measure would be along the lines of: "In this ad, What do you thinkthe advertiser is trying to tell you about the brand?" Open-ended responses are then coded ascorrect if they mirror or closely paraphrase the advertisers intended message, i.e. the benefitbeliefs to be learnedAn overall measure of brand attitude is taken, usually relative to other brands in the evoked set(except in the low involvement/transformational model as explained below). Finally, aided ordiagnostic measures of the benefit beliefs presumed to lead to the overall brand attitude can betaken. For low involvement beliefs, it is recommended that the beliefs be measured on a 0-1(yes-no basis: for low involvement brand purchase decisions, the brand either has thecharacteristic or it does notHigh involvement processing is quite different. Here it is the believed message that counts. Thecorrect type of processing measure for high involvement ads is of the cognitive response variety
  23. 23. (e.g., Wright 1980; Belch 1982; Lutz and MacKenzie 1981; Petty, Ostrom and Brock 1981).Measurement focuses on the target audiences own responses rather than the advertisers intendedresponses. A question such as, "List all the thoughts, visual images, or feelings that you hadwhen watching/listening to/looking at this ad," is used to prompt reproduction of actualprocessing reactions, then these are coded as positive, negative, or more specifically in terms ofreactions to elements of the ads.An overall brand attitude measure is taken, followed by measure of specific benefit beliefs. Inhigh involvement models, beliefs are most likely to be graduated rather than all-or-none, and somulti-step scales (either Likert or semantic differential, see below) are appropriate.Informational Versus TransformationalA further set of differences occurs in testing informational ads as distinct from transformationalads. For ads based on the informational brand attitude strategy, a rough execution is sufficient fortest purposes, because the informational (reason why) message should be apparent regardless ofthe executional quality of the ad presented to consumers. [Schlinger and Green (1980) havepresented detailed comparisons of the test results from rough versus finished ads. Thecorrespondence is noticeably higher for informational measures than for transformational or"image" measures.] Likewise, the message should be apparent immediately, so that one exposure(print ads) or two exposures (broadcast; cf. Krugmans 1972 point that the first exposure of afleeting broadcast ad allows only a "What is it?" response) should be sufficient.An overall brand attitude measure is used. For attitude measurement, a pre-post design is mostefficient. A pre measure is not sensitizing because consumers consciously experience attitudeshifts with informational ads. For diagnostic measurement of brand benefit beliefs, Likert-type(agree-disagree) scales are recommended, because the consumer must agree that the brandprovides the benefit; note, however, the 0-1 measures recommended if the model is lowinvolvement-informational. Finally, since informational ads are supposed to work immediately,it is appropriate to include a purchase intention measure.For ads based on the transformational brand attitude strategy, a high quality version of the test adis required. Transformational ads depend on production values for contributing to the positivelyderived brand attitude, so a transformational ad should be tested in a version as close to thefinished ad as possible. Whereas one (unlimited or natural) exposure is sufficient fortransformational print ads, multiple exposures are necessary for transformational broadcast ads toallow them the opportunity to "build" their brand attitude effect. Three to four exposures of thetest commercials represent a more valid simulation of real world conditions than just one or twoexposures.An overall brand attitude measure is usually appropriate for high involvement-transformationalmodels; but it is unrealistic to expect low involvement-transformational advertising to producean immediate attitude because of the low involvement, virtually subconscious way in whichthese ads operate, and thus an overall brand attitude measure is omitted in this case. For the samereason that transformational advertising produces "soft" attitude shifts, a non-sensitizingexperimental-control (two group) design is recommended rather than a pre-post design. Brand
  24. 24. benefit beliefs are also more appropriately measured on the "softer" semantic differential type ofrating scales, anchored by "image adjectives" than on Likert-type agree-disagree scales. Finally,transformational attitude shifts are not validly captured by an immediate purchase intentionmeasure because the translation of attitude into planned action is not an immediate, consciousprocess as in informational advertising. The kid glove response registration represented bysemantic differential-type "image" measures is the most valid way of testing the typically fragileeffects of transformational advertising.Composition of a valid ad test therefore depends crucially on careful prior identification of asuitable advertising communication model. The eight basic models outlined in this paper shouldprove adequate for most advertising applications. If a more specifically tailored model is desired,the eight basic models together with the strategy checklist should be of considerable assistance inindicating the questions to be answered in designing the inputs for a specialized model.
  26. 26. SECTION B
  27. 27. SECTION C
  28. 28. SECTION D
  29. 29. REFERENCESBauer, R.A. (1967), "Source Effect and Persuasibility: A New Look," in Risk Taking andInformation Handling in Consumer Behavior, D.F. Cox, ed., Cambridge, MA: HarvardUniversity Press, 559-578.Belch, G.E. (1982), "The Effects of Television Commercial Repetition on Cognitive Responsesand Message Acceptance," Journal of Consumer Research, 9 (June), 56-65.Bettman, J.R. (1979), "Memory Factors in Consumer Choice: A Review," Journal of Marketing,43 (Spring), 37-53.Cialdini, R.B., R.E. Petty, and J.T. Cacioppo (1981), "Attitude and Attitude Change," AnnualReview of Psychology, 32, 357-404.Colley, R. (1961), Defining Advertising Goals for Measured Advertising Results, New York:Association of National Advertisers.Ehrenberg, A.S.C. (1974), "Repetitive Advertising and the Consumer," Journal of AdvertisingResearch, 14 (April), 25-34.Fennell, G. (1975), "Motivation Research Revisited," Journal of Advertising Research, 15(June), 23-27.Fennell, G. (1978), "Consumers Perceptions of the Product-Use Situation," Journal of Marketing,42 (April), 38-47.Finn, D.W. (1982), "Try It, Youll Like It: A Case Against the Low-Involvement Hierarchy,"Texas Christian University, M.J. Neeley School of Business, working paper no. 6.Fishbein, M. and I. Ajzen (1975), Belief, Attitude, Intention, and Behavior: An Introduction toTheory and Research, Reading, MA; Addison-Wesley.Harris, R.J., T.M. Dubitsky, R.L. Perch, C.S. Ellerman, and M.W. Larsen (1980), "RememberingImplied Advertising Claims as Facts: Extensions to the Real World," Kansas State University,Department of Psychology, working paper no. 80-2.Houston, M.J. and M.L. Rothschild (1977), "A Paradigm for Research on ConsumerInvolvement," University of Wisconsin-Madison, Graduate School of Business, working paperno. 11-77-46.Hovland, C.I., I.L. Janis, and H.H. Kelly (1953), Communication and Persuasion, New Haven,CT: Yale University Press.Howard, J.A. and J.N. Sheth (1969), The TheorY of Buyer Behavior, New York, Wiley.
  30. 30. Kelman, H.C. (1958), "Compliance, Identification, and Internalization: Three Processes ofOpinion Change," Journal of Conflict Resolution, 2 (Summer), 51-60.Krugman, H.E. (1965), "The Impact of Television Advertising: Learning Without Involvement,"Public Opinion Quarterly, 29 (Fall), 349-356.Krugman, H.E.(1972),Why Three Exposures May Be Enough," Journal of AdvertisingResearch, 12 (December), 11-14.Lavidge, R.J. and G.A. Steiner (1961), "A Model for Predictive Measurements of AdvertisingEffectiveness," Journal of Marketing, 25 (October), 59-62.Little, J.D.C. (1979), "Decision Support Systems for Marketing Managers," Journal ofMarketing, 43 (summer), 9-26.Lutz, R.J. and ;.B. MacKenzie (1981), "Construction of a Diagnostic Cognitive Response Modelfor Use in Commercial Pretesting," University of California, Los Angeles, Graduate School ofManagement, working paper no. 110.Lutz, R.J. and P.J. Reilly, (1974), "An Exploration of the Effects of Perceived Social andPerformance Risk on Consumer Information Acquisition," in Advances in Consumer Research,Vol. 1, S. Ward and P. Wright, eds., Ann Arbor, MI: Association for Consumer Research, 393-405.Maloney, J.C. (1962), "Curiosity Versus Disbelief in Advertising," Journal of AdvertisingResearch, 2 (June), 2-8.Mandler, G. (1979), "Emotion," in The First Century of Experimental Psychology, E. Hearst, ed.,Hillsdale, NJ: Lawrence Erlbaum Associates. 275-321.McGuire, W.J. (1969), "The Nature of Attitudes and Attitude Change," in The Handbook ofSocial Psychology, Vol. 3, G. Lindzey and E. Aronson, eds., Reading, MA: Addison-WesleY.136-314.McGuire, W.J. (1976), "Some Internal Factors Influencing Consumer Choice," Journal ofConsumer Research, 2 (March). 302-319.Mitchell, A.A. and J.C. Olson (1981), "Are Product Attribute Beliefs the Only Mediator ofAdvertising Effects on Brand Attitude?," Journal of Marketing Research, 18 (August), 318-332.Naples, M.J. (1979), Effective Frequency: The Relationship Between Frequency and AdvertisingEffectiveness, New York: Association of National Advertisers. Inc.Nelson, P.E. (1970),"Information and Consumer Behavior," Journal of Political Economy, 78(March/April), 311-329.
  31. 31. Percy, L. and J.R. Rossiter (1980), Advertising StrategY: A Communication Theory Approach,New York: Praeger.Perloff, R.M. and T.J. Brock, (1980), "And Thinking Makes It So: Cognitive Responses toPersuasion," in Persuasion: New Directions in Theory and Research, M.E. Roloff and G.R.Miller, eds., Beverly Hills, CA: Sage, 67-99.Peter, J.P. and Tarpey, L.X. (1975), "A Comparative Analysis of Three Consumer DecisionStrategies," Journal of Consumer Research, 2 (June), 29-37.Petty, R.E. and Cacioppo, J.T. (1979), "Issue Involvement Can Increase or Decrease Persuasionby Enhancing Message Relevant Cognitions," Journal of Personality and Social Psychology, 37(June), 1915-1926.Petty, R.E. and Cacioppo, J.T. (1983), "Central and Peripheral Routes to Persuasion: Applicationto Advertising," in L. Percy and A.G. Woodside, eds., Advertising and Consumer Psychology,Lexington, MA:Lexington, 3-23.Petty, R.E., Ostrom, T.M.., and Brock, T.C., eds. (1981), Cognitive Responses in Persuasion,Hillsdale, NJ: Lawrence Erlbaum Associates.Preston, I.L. (1982), "The Association Model of the Advertising Communication Process,"Journal of Advertising, 11 (no. 2), 3-15.Ray, M.L. (1982), Advertising and Communication Management, Englewood Cliffs, NJ:Prentice-Hall.Ray, M.L. and P.R. Webb (1974), "Three Learning Theory Traditions and Their Application inMarketing," in Combined Proceedings: American Marketing Association Conference, R.Curham, ed., Chicago, IL: American Marketing Association, 100-103.Robertson, T.S. (1971), Innovative Behavior and Communication, New York: Holt, Rinehart andWinston.Rogers, E.X. (1962), Diffusion of Innovations, New York: Free Press.Rossiter, J.R. and L, Percy (1980), "Attitude Change Through Visual Imagery in Advertising,"Journal of Advertising, 9 (Winter), 10-16.Rossiter, J.R. and L. Percy (1981), "Visual Communication in Advertising," ColumbiaUniversity, Graduate School of Business, working paper no. 231-A.Rossiter, 3.R. and L. Percy (1983), Draft chapters for text, Advertising and PromotionManagement, New York: McGraw-Rill, available from authors.
  32. 32. Schlinger, M.J. and L. Green (1980), "Art-work Storyboards Versus Finished Commercials,"Journal of Advertising Research, 20 (December), 19-23.Shimp, T.A. and J.T. Yokum (1982), "Advertising Inputs and Psychophysical Judgments inending-Machine Retailing," Journal of Retailing, 58 (Spring), 95-113.Smith, R.E. and W.R. Swinyard (1982), "Information Response Models: An IntegratedApproach," Journal of Marketing, 46 (Winter), 81-93.Thorson, E. and M.L. Rothschild (1983), "Using a Text Comprehension Analysis to CompareRecognition and Recall of TV Commercials," in Advertising and Consumer Psychology, L.Percy and A.E. Woodside, eds., Lexington, MA: Lexington Books.Vaughn, R. (1981), "Row Advertising Works: A Planning Model," Journal of AdvertisingResearch, 20 (October), 27-33.Webster, F.E. and Y. Wind (1972), Organizational Busing Behavior, Englewood Cliffs, NJ:Prentice-Hall.Wells, W.D. (1981), "How Advertising Works," Mimeo, Chicago, IL: Needham, Harper &Steers Advertising, Inc.Wright, P.L. (1980), "Message-Evoked Thoughts: Persuasion Research Using ThoughtVerbalizations," Journal of Consumer Research, 7 (September), 151-175.Wyer, R.S. (1974) Cognitive Organization and Change: An Information Processing Approach,Potomac, MD: Lawrence Erlbaum Associates.Zajonc, R. (1980), "Feeling and Thinking: Preferences Need No Inferences," AmericanPsychologist, 35 (February), 151-175.