26
Extending a contrast resolution model of humor in television advertising: the role of surprise DANA L ALDEN, ASHESH MUKHERJEE, and WAYNE D. HOYER Abstract Few studies have tested models incorporating cognitive äs well äs affective mechanisms that help explain different levels of perceived humorousness in advertising (cf. Alden and Hoyer 1993; Speck 1991). In this study, we extend the current literature by hypothesizing relationships between several message characteristics, perceived humor andattitude towards the ad. In an empirical test ofan extended contrast-resolution model, surprise isfound to be related to substantial amounts of variance in the perception of humor. Surprise is also found to mediale the effects of other antecedent variables such äs type of contrast, strength of contrast andextent ofimagery evoked by the ad. These findings suggest that surprise is a central driving force determining the humorousness ofan adver tisement. The implications ofthe proposed extension of the contrast-resolution model for humor in television advertising are discussed from both theoretic andappliedperspectives, and directions for future research are suggested. Introduction Humor is one ofthe most commonly employed communication strategies in advertising. Researchers estimate that between 11 percent and 24 per- cent of television ads in the United States use humor (Alden, Hoyer and Lee 1993; Weinberger and Spotts 1989; Speck 1991). Similar or higher usage has been reported in other countries (Alden et al. 1993) and in other media such äs radio (Weinberger and Campbell 1991). The apparent popularity of humor may stem, in part, from the managerial belief that it enhances the effectiveness of brand-related Humor 13-2 (2000), 193-217 0933-1719/00/0013-0193 © Walter de Gruyter Brought to you by | University of Arizona Authenticated Download Date | 12/18/14 1:19 AM

Extending a contrast resolution model of humor in television advertising: the role of surprise

  • Upload
    wayne-d

  • View
    215

  • Download
    2

Embed Size (px)

Citation preview

Page 1: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution modelof humor in television advertising:

the role of surprise

DANA L ALDEN, ASHESH MUKHERJEE, andWAYNE D. HOYER

Abstract

Few studies have tested models incorporating cognitive äs well äs affectivemechanisms that help explain different levels of perceived humorousnessin advertising (cf. Alden and Hoyer 1993; Speck 1991). In this study, weextend the current literature by hypothesizing relationships between severalmessage characteristics, perceived humor andattitude towards the ad. In anempirical test ofan extended contrast-resolution model, surprise isfound tobe related to substantial amounts of variance in the perception of humor.Surprise is also found to mediale the effects of other antecedent variablessuch äs type of contrast, strength of contrast andextent ofimagery evokedby the ad. These findings suggest that surprise is a central driving forcedetermining the humorousness ofan adver tisement. The implications oftheproposed extension of the contrast-resolution model for humor in televisionadvertising are discussed from both theoretic andappliedperspectives, anddirections for future research are suggested.

Introduction

Humor is one ofthe most commonly employed communication strategiesin advertising. Researchers estimate that between 11 percent and 24 per-cent of television ads in the United States use humor (Alden, Hoyer andLee 1993; Weinberger and Spotts 1989; Speck 1991). Similar or higherusage has been reported in other countries (Alden et al. 1993) and in othermedia such äs radio (Weinberger and Campbell 1991).

The apparent popularity of humor may stem, in part, from themanagerial belief that it enhances the effectiveness of brand-related

Humor 13-2 (2000), 193-217 0933-1719/00/0013-0193© Walter de GruyterBrought to you by | University of Arizona

AuthenticatedDownload Date | 12/18/14 1:19 AM

Page 2: Extending a contrast resolution model of humor in television advertising: the role of surprise

194 D. L. Alden, A. Mukherjee and W. D. Hoyer

advertising. Indeed, humorous messages that are appropriate for theproduct category and are well integrated with other ad objectives appearto enhance attention, credibility, recall, evaluation and purchase inten-tion (Scott, Klein and Bryant 1990). In addition, humor can have apositive impact on important communication outcomes such äs counter-argumentation (Krishnan and Chakravarti 1990; Scott, Klein and Bryant1990) and comprehension (Stewart and Furse 1986).

Focusing on the effects of humorous versus nonhumorous advertising,scholars have primarily studied this message strategy äs a manipulatedvariable (i.e., humor is present or not). While providing many valuableinsights, this approach does not directly address the fact that advertisingoften varies greatly in terms of perceived humorousness (see Unger 1996).In other words, past research has largely avoided investigating whatconstitutes the "essence" of humor; why many ads attempt humor, butonly some are perceived äs funny. Indeed, only a few studies haveconceptualized and analyzed content-related factors related to percep-tions of humor in advertising (see Alden and Hoyer 1993; Speck 1991).Nevertheless, this represents an important area of study because levelsof perceived humor are likely to have important consequences for otheradvertising outcomes such äs recall, attitude toward the ad and brandattitude (see Chattopadhayay and Basu 1989).

Thus, although research to date has significantly enhanced ourunderstanding of the overall effects of humor in advertising, furtherinvestigation into the antecedents of perceived humor appears to bewarranted. This article presents the results of a study designed to enhancetheory related to the generation of humor in television advertising.Specifically, this study demonstrates the central role of surprise in thegeneration of a humorous response to television advertising. In addition,the study also identifies other key antedents to perceived humor, suchäs type of humorous contrast, strength of humorous contrast, and degreeof imagery evoked by the advertisement.

Background literature

Research on humor in advertising has focused on at least five generaltopics (for a thorough review of the literature, see Weinberger andGulas 1992). First, studies have examined the eifects of humor on avariety of traditional message outcome variables such äs attention,

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 3: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 195

comprehension and persuasion (Osterhouse and Brock 1970; Stewartand Furse 1986; Speck 1987; Scott, Klein, and Bryant 1990; Zhang andZinkhan 1991). Second, researchers have investigated possible moderat-ing effects of non-message factors such äs prior brand attitudes, productcharacteristics (e.g., involvement) and audience characteristics on rela-tionships between humor and traditional advertising outcomes (Gelb andZinkhan 1985; Chattopadhyay and Basu 1989; Weinberger and Campbell1991).

A third area of research has attempted to determine whether thepositive influence of humor on brand attitudes occurs primarily throughcognitive processes such äs enhanced recall (Cho 1994; Zhang andZinkhan 1991) or through affective processes such äs transfer of positiveaffect from the ad to the brand (Zinkhan and Gelb 1990; Aaker, Stayman,and Hagerty 1986). Reaching beyond the United States, a fourth streamhas identified national diiferences and similarities in the use and charac-teristics ofhumor in advertising from multiple countries (Weinberger andSpotts 1989; Alden et al. 1993).

Finally, researchers have tested models that help explain how advertis-ing content can affect perceived levels of humor. A common factorhypothesized to generate humor in these models is "incongruity fromexpectations" and the resolution of such incongruity (Speck 1991; Aldenand Hoyer 1993). This approach is based largely on the work of psycho-logists and linguists who have proposed that incongruity resolution ina playful context is a necessary and sufficient condition for humor (Wickeret al. 1981; Suls 1983; Attardo et al. 1994; Herzog and Larwin 1988;Oppliger and Sherblom 1988).

For example, Raskin (1985) has advanced a "script based" contrastresolution model of humor. According to this theory, a verbal or writtencommunication is considered a joke when the "text ... is fully com-patible with two different scripts and the two scripts are opposite incertain definite ways" (Raskin 1985). These contrasts can occur, forexample, when scenarios that are expected in everyday life are contrastedwith those that are possible but unusual (e.g., a man walks down a streetand begins to skip like a child) or when scenarios that are possible ineveryday life are contrasted with those that are unreal, fantastic andimpossible (e.g., a man walks down a street, flaps bis arms and beginsto fly). In both cases, deviation from expectations and the cognitiveresolution of these deviations are hypothesized to form the basis forhumor generation.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 4: Extending a contrast resolution model of humor in television advertising: the role of surprise

196 D. L. Alden, A. Mukherjee and W. D. Hoyer

Drawing from the incongruity resolution stream and other theoriesof humor, Speck (1991) proposes a framework for analyzing underlyingprocesses of humor generation in advertising. In addition to incongruityresolution, he suggests that arousal-safety and humorous disparagementmechanisms are involved. Further, he combines these mechanisms withdifferent advertising orientations such äs Information or image orienta-tion to create a taxonomy of humor. A preliminary, exploratory analysisof advertising humor provided initial support for this approach.

Working within an advertising context äs well, Alden and Hoyer (1993)tested the contrast resolution theory of Raskin (1985). They report that65 percent of the dominant contrasts in ads attempting humor were setin either reality-based or unreal, fantasy-based situations. Also, it wasfound that consumers rated reality-based contrasts äs significantly morehumorous than ads using unreal, fantastic contrasts. Offering a post-hocexplanation for this finding, Alden and Hoyer (1993) suggest that reality-based contrasts may on average deviate less from well-formed expecta-tions than those involving the impossible. This in turn may lead to fasterresolution of reality-based contrasts, easier assimilation within cuedSchemas and hence, stronger positive affect (Meyers-Levy and Tybout1989).

While Alden and Hoyer (1993) offer an interesting hypothesis fordifferences in the humor evaluations associated with the two types ofRaskin (1985) contrasts, they did not test for other factors that mayhelp explain their findings. For example, it may be that higher levels ofsurprise generated by reality-based contrasts versus unreal, fantasy-based contrasts is a key factor mediating the eifects of type of contraston perceived humor.

The role of surprise in humor

Within the consumer research literature, the ability of incongruent oratypical advertising content to evoke deeper levels of processing andenhanced positive affect relative to schema-consistent advertising has beenwell established (Meyers-Levy and Tybout 1989; Stayman et al. 1992;Goodstein 1993). Theoretic mechanisms used to account for positiveaffect following exposure to incongruent Stimuli have been primarilycognitive in nature, such äs the transfer of category-level affect to con-gruent or moderately discrepant Stimuli (Fiske and Pavelchak 1986) or

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 5: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 197

the cognitive resolution of the presented incongruity (Mandler 1982).Still to be examined is a possible mediating role played by responsesto incongruity that are more emotional in nature such äs surprise. How-ever, it is not clear that incongruity alone produces a humorous evalua-tion. Rather, we hypothesize that the emotion of surprise plays a keymediational role in the incongruity-humor perception process and thatincongruity and cognitive processes alone (even if accompanied by themild ANS arousal described in Meyers-Levy and Tybout 1989) areinsufficient to account for differences in how funny consumers thinkan ad is.

Surprise is described by Meyer and Niepel (1994, p. 353) in generalterms äs:

... a hypothetical entity that can be observed from three classes of hypotheticalevents: specific physiological changes, behavior patterns, and verbal reports aboutsubjective experience. Surprise is elicited by unexpected events, that is events thatdeviate from a schema.

Izard (1977) identifies surprise äs an emotion when he defines it äs atransitory feeling of uncertainty, set off by any sudden unexpected event.However, the extent to which surprise is a pure emotion has producedcontroversy in the psychological literature. To some extent, whether ornot one considers surprise an emotion depends on the definition ofemotion itself (Meyer and Niepel 1994). Such definitions have typicallybeen more complex than straightforward. For example, Frijda et al. (1992)define emotions äs multivariate phenomena involving cognitions, actionreadiness, feeling and physiological change while Batson et al. (1992) viewthem primarily äs feeling states with tone and intensity that provideInformation and motivation within goal-directed action sequences.

Surprise is similar to pure emotions in that it involves a fairly strongphysiological arousal. However, it is unlike such emotions in that it doesnot imply a specific "tone" or valence (e.g., joy is pleasant and fear isunpleasant while one can be pleasantly or unpleasantly surprised). Thus,Shaver, Wu, and Schwartz (1992) include a "surprise emotion cluster"among six basic-level emotions, based on a sorting task of 135 emotionlabels by large numbers of subjects. However, they qualify their Unding bynoting that there were fewer concepts in the surprise cluster than in otheremotion clusters (e.g., joy, sadness) and that labels within the surprisecluster were less clear in terms of affective valence. Ortony and Turner(1990, p. 317) go a step further when they argue that "it is by no means

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 6: Extending a contrast resolution model of humor in television advertising: the role of surprise

198 D. L. Alden, A. Mukherjee and W. D. Hoyer

clear that surprise is an emotion." According to Ortony and Turner(1990), experiencing emotion necessarily implies an affectively valencedstate. However, surprise can be affectively neutral, negative or positive.Hence, it is not a "true" emotion.

Taking the position that surprise is not inherently valenced in eithera positive or negative direction is the so-called "self-report affect circum-plex model" of emotion described by Larsen and Diener (1992). In thismodel, emotions are arrayed on a two dimensional grid with "pleasant-unpleasant" on one dimension and "high-low activation" on the other.Surprise is located in the neutral quadrant of the pleasant-unpleasantdimension along with such terms äs "aroused," "astonished," and"stimulated." Thus, although affectively unvalenced, surprise is stillconsidered in the circumplex model to be an emotional response entailinga fairly high degree of physiological activation. This moderate to strong"physiological arousal" component is consistent with other researchers'definitions of emotion in general and surprise in particular. Indeed, it wasthis limiting factor that guided the seminal work of James (1884, p. 189)who stated that "the only emotions that I propose expressly to considerhere are those that have a distinct bodily expression."

In addition to the conceptual literature, it has also been empiricallydemonstrated that surprise is a primary emotional event triggered bystimulus-schema incongruity (Meyer 1986; see also Meyer et al. 1991).Further, perceptions of surprise have been associated with strongincreases in neural activation, changes in brain wave function anduniversally recognized facial response patterns: raised eyebrows, wrinkledforehead, rounded, open eyes and oval shape, opened mouth (Meyer andNiepel 1994, Levenson et al. 1990). Indeed, based on such facial responsepatterns, Rüssel, Suzuki, and Ishida (1993) identify surprise äs one of sixbasic emotions. Thus, at the very least, to the extent that there is a directvisceral response generated by an unexpected deviation from an activeSchema, surprise certainly has an emotional dimension. Further, the emo-tional arousal accompanying surprise typically results in "an inter-ruption of ongoing activities and a spontaneous focusing of attentionon the eliciting Stimulus" in order to "enable and motivate processes thatserve to remove this discrepancy" (Niepel et al. 1994, p. 434; see alsoMeyer et al. 1991).

Hence, in sum, the construct of surprise appears to be largely affectivein character, but also has clear effects on cognitive processes. These dualdimensions of emotional arousal and enhanced cognitive activity may

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 7: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 199

help explain the central role of surprise in the generation of humor. Yet,to date, the role of surprise äs a mediator in the generation of humorin advertising has not been established. To initiate research on the roleof surprise and other related constructs in the generation of a humorousresponse, we now offer and test several specific research hypotheses.

Hypotheses

Our first two hypotheses predict replication of results involving Raskin's(1985) contrasts found by Alden and Hoyer (1993) in television adsattempting humor. Thus, we expect to also find that a majority of adsattempting humor will feature one of two Raskin (1985) contrasts: reality-based or unreal, fantasy-based. In addition, we expect that, on average,advertisements attempting humor that employ reality-based contrasts willbe perceived äs more humorous than those that use unreal, fantasy-basedcontrasts. Both of these hypotheses are now stated formally:

Hl: Among ads that attempt humor, a majority will employ eitherreality-based or unreal, fantasy-based incongruous contrasts.

H2: Among ads that attempt humor, those using reality-based contrastswill be perceived äs more humorous than those using unreal,fantasy-based contrasts.

Turning now to the key construct of surprise, it may be argued that thelevel of surprise generated by an ad and its subsequent impact on humor islikely to depend on a number of factors such äs: (1) situational processinggoal at the time of exposure (e.g., casual enjoyment versus activeInformation seeking); (2) individual differences (e.g., affective responsetendencies, cf., Moore, Harris, and Chen 1995); and (3) contextual messagestrategies such äs degree of unexpectedness from evoked Schemas (Frijdaet al. 1992). Our interest in this study is with the third kind of variable —contextual message factors that impact on the level of surprise and,subsequently, humor. Specifically, we now propose extensions to acontrast-resolution model of humor in television advertising thatincorporates the antecedent message factors of type of contrast, strengthof contrast, and imagery Stimulation, together with the mediating variableof surprise (see Figure 1).

Advertising message strategies are known to vary along a dimension(among others) that runs from resonance or "slice of life" content to

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 8: Extending a contrast resolution model of humor in television advertising: the role of surprise

200 D. L. Alden, A. Mukherjee and W. D. Hoyer

ITvpeofContrastl IStrength of Contrast]

Figure 1. Hypothesized model ofhumor in advertising

fantastic, imaginary and unreal content (Shimp 1996), corresponding toRaskin's (1985) reality-based versus unreal, fantasy based contrasts inhumorous Communications. Highly developed Schemas may be availablein consumers' long-term memories for both 'slice of life'/reality-based äswell äs unreal/fantasy-based contrasts, depending on the amount ofexperience the individual has had with the Situation depicted in thead. Thus, if an individual has had extensive personal experience with theSituation shown in a reality-based contrast, the associated mental Schemaactivated is likely to be extensive and detailed. Similarly, if the individualhas had extensive vicarious experience (e.g., through books, myths) withthe Situation depicted in unreal or fantasy-based contrasts, then too theactivated Schema is likely to be rieh and detailed in nature.

While well-developed Schema of both types (real and unreal) exist inlong term memory, it is our contention that, on average, advertisementsemploying "slice of life" contrasts are more often likely to cue highlydetailed and overlapping Schemas than those employing "unrealcontrasts." This prediction is consistent with past research investigatingthe relative effects of advertising versus direct experience on variables suchäs recall, attitudes and purchase intentions (see Wright and Lynch 1995;Berger and Mitchell 1989). Specifically, it has been noted that directexperience generates richer and more detailed memory traces (i.e.,Schemas) than advertising, unless the advertising is repeated often. Sincemost vicarious experiences that underlie fantasy-based Schemas are not

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 9: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 201

likely to be repeated often, it is likely that unreal, fantasy-based Schemaswill be less detailed than those formed through direct experience. If thisis the case, then the spreading activation model proposed by Anderson(1983) predicts that more rapid and widespread activation of memorynodes in working memory will be associated with more developed and/or overlapping Schemas associated with everyday "slice of life contrasts."As a result, ads that feature reality-based äs opposed to unreal, fantasy-based contrasts should, on average, generate more immediate and morewidespread activation of cued memory nodes in long-term memory. Sinceeach memory node is a distinct source of arousal, a larger number ofactive (aroused) nodes implies more intense physiological and emotionalarousal (for an excellent conceptualization of Anderson's model within anadvertising context, see Grunert 1996). As discussed earlier, this arousalconstitutes the emotion of surprise. Thus we hypothesize:

H3: For ads that attempt humor, those using reality-based contrastswill evoke greater surprise than those using unreal, fantasy-basedcontrasts.

Further, within each type of contrast (i.e., reality-based or fantasy-based),there may be variations in the strength of the contrast (i.e., the extent ofthe discrepancy within a contrast). The strength of a contrast may beviewed äs the schematic distance between the expected and the unexpectedsituations in reality-based contrasts, and the distance between possibleand the impossible situations in fantasy-based contrasts. Strongercontrasts will generate greater Schema incongruity, will be more difficultto resolve and therefore create higher levels of arousal and surprise. Thuswe hypothesize,

H4: The greater the strength of the contrast in an ad, the greater the levelof surprise.

In addition to the type and strength of contrast, we argue that the degreeof imagery evoked by the ad is another antecedent to perceived surprise.In contrast to discursive processing (thinking, problem solving, cognitiveresponding), imagery processing involves more holistic, visual, andsensory representations in working memory (Mclnnis and Price 1987).These two modes of processing are intertwined (Bugelski 1983) but onemode may be predominant. In the case of a visual medium like television(versus print or radio) imagery processing is likely to play a prominentrole. In addition, most television advertisement viewing has been shown

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 10: Extending a contrast resolution model of humor in television advertising: the role of surprise

202 D. L. Alden, A. Mukherjee and W. D. Hoyer

to occur in a low involvement environment (Krugman 1965; Mclnnisand Jaworski 1989) where low elaboration, peripheral route processing(Petty and Cacioppo 1986) is the norm. Further, in such low elabora-tion processing, imagery evocation and elaboration in the mind has beenshown to lead to enhanced incidental learning (Bower 1972; Sheehan1972). Hence humorous ads which evoke imagery are likely to achievegreater comprehension of the details of its humorous contrast throughenhanced incidental learning, an effect sharpened by the dominance ofimagery processing in the television medium. This increased comprehen-sion of the details of the humorous contrast should lead to heightenedperception of the contrast, which should, in turn, enhance surprise. Hencesummarizing the above sequence of effects, we hypothesize that,

H5: The greater the extent of imagery evoked by an ad, the greaterthe level of surprise.

The second stage of the model shown in Figure l focuses on therelationship between surprise and perceived humor. As noted earlier,moderate levels of incongruity have been found by consumer researchersto produce higher levels of positive "liking" than low or high levels.Following such theorists äs Mandler (1982), consumer researchers havehypothesized that cognitive mechanisms involving mild arousal coupledwith enhanced elaboration and successful resolution are responsible forthis effect (cf., Meyers-Levy and Tybout 1989; Stayman, Alden, andSmith 1993). However, we expect that the inverted U relationship betweenincongruity and "liking affect" will not be reproduced in the case ofsurprise and "humor affect" generated by television advertising. Rather,a linear relationship is expected for several reasons.

First, it has been argued that heightened ANS activity or emotionalarousal associated with surprise marshals additional cognitive resourcesrequired to resolve the incongruity created by the humorous attempt(Niepel et al. 1994). As a result, it is likely that the humorous incongruitywill be resolved in most cases, leading to the successful generation ofhumor at all levels of incongruity (see Suls 1983). Second, commercialtelevision advertising is designed to entertain and persuade in a very shortperiod of time. As a result, the extent of presented incongruity äs well ästhe degree of effort required to resolve the incongruity which underliesthe surprise reaction, is likely to be small in most television advertising.As a result, successful resolution is expected in most cases, leading toa monotically positive relationship between surprise and "humor affect."

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 11: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 203

Support for these arguments may be found in the psycho-physiologicalliterature which provides evidence that humor processing consists oftwo distinct elements: a surprise reaction to perceived incongruity andsubsequent positive affective response related to cognitive understand-ing of the relationship between the punch line and the body of the joke(see Brownell et al. 1983). Similarly, a number of researchers have noteda strong positive correlation between surprise and humor in non-advertising contexts (e.g., McGhee and Johnson 1975; Deckers andWinters 1986; Deckers et al. 1987). Finally, psychoanalytic analyses ofhumor used in therapy (Baker 1993) have identified surprise äs an 'affect-releasing' antecedent to humor. Results such äs these lead to our nexthypothesis which states:

H6: For ads that attempt humor, there will be a direct, linearrelationship between surprise and perceived humor.

We have made the case that surprise is a mediator of perceived humor.Thus, variables such äs type of contrast, strength of contrast, and extentof evoked imagery should affect perceived humor through surprise ratherthan directly. To explicitly test whether surprise plays a central mediatingrole in determining levels of perceived humor, we hypothesize:

H7: For ads that attempt humor, surprise will mediate the effects oftype of contrast, strength of contrast, and extent of evoked imageryon perceived humor.

Prior research has shown that attitude towards the ad (Aad) can shapeattitude towards the brand (Abr) directly, äs well äs indirectly through itseffect on brand cognitions (Brown and Stayman 1992; Homer 1990). Weexpect that in a low cognitive elaboration scenario typical of televisionviewing (Krugman 1965), peripheral cues like humor will have anespecially significant effect on Aad, primarily through the former route(Petty and Cacioppo 1986). Consistent with this hypothesis, Unger (1995)finds a direct positive relationship between humor and ad liking in a studywith a limited sample size (10 ads). The present study employs a muchlarger sample to test for a positive effect of humor on attitude toward thead. Thus, our last hypothesis states:

H8: For ads that attempt humor, the higher the level of perceivedhumor, the more favorable will be attitude towards the ad.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 12: Extending a contrast resolution model of humor in television advertising: the role of surprise

204 D. L. Alden, A. Mukherjee and W. D. Hoyer

Methodology

Sampling of Humorous Television Ads

The sampling plan of this study was designed to maximize the variabilityof perceived humor in television ads (and hence the power of ourstatistical tests) while preserving the external validity of the Stimuli. Tothis end, a sample of television ads representing the füll ränge of humorwas collected using a two-step process.

First, randomized cluster samples of ads shown nationally on threenational television networks — CBS, ABC and NBC — were recordedover a period of three randomly selected days of the week in the month ofDecember, 1993. Fach day was split into three time spans of eight hoursand a different time span was randomly assigned to each day for thepurpose of recording the ads. Duplicated ads were eliminated but differentads for the same brand were retained. This resulted in an unduplicatedrepresentative sample of 214 national brand ads. Three coders rated eachad on whether or not humor was intended by the Creators of theadvertisement (yes/no). Coders were trained through a series of pre-evaluation exercises in order to minimize personal biases in the judgmentof humorous intent. 34 ads exhibiting 100 percent intercoder agreementon intended humor were retained for subsequent analysis.

Pretest ratings of these 34 ads indicated that they were negativelyskewed on humor (more were rated äs not very funny). To achieve a morebalanced distribution of humor ratings, a set of high humor ads wasadded to the sample. In particular, 64 award winning humor ads (inEnglish) from the London International Advertising Awards (1992) wereevaluated by the same coders using a 7-point semantic differential scaleof perceived humor. Based on a quartile split, eighteen ads with thehighest humor ratings were added to the sample resulting in a total of52 ads for analysis.

Independent Variable Measures: Expert Ratings of Ads

Three new coders conducted an in-depth coding of the 52 ad sample. Eachof the 52 ads in the sample was shown (more than once if needed) to thecoders who «independently evaluated each ad after extensive training.1

Their main task was to identify whether each ad employed one or more

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 13: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 205

of the incongruity contrasts hypothesized by Raskin (1985): expected/unexpected (i.e., reality-based), possible/impossible (i.e., fantasy-based),or actual/nonactual (i.e., containing both reality and fantasy-basedelements). For those ads found to contain more than one of the contrasts,coders were asked to identify which contrast was featured mostprominently. They were also asked to rate the strength of the mostprominent contrast, and the extent of imagery evoked by the ad usingthree point scales (Low-Medium-High). The intercoder agreement rateexceeded 80 percent across all items on the coding sheet meeting theStandards suggested by Kassarjian (1977). Disagreements were resolvedamong the coders without the involvement of the investigators.

Dependent Variable Measures: Consumer Ratings of Ads

71 undergraduate business students from a large southwestern universityreceived class credit for evaluating each ad. To control for order effects,subjects were randomly divided into three groups (22-25 subjects pergroup) and shown a different randomized sequence of the ads. Using7-point semantic differential scales, subjects rated ads on: humorousness(Very Humorous-Not At All Humorous); surprise (Very Surprising-NotAt All Surprising); and liking (Liked Very Much-Disliked Very Much).Although reliability risks are associated with single item measures,concern with bias due to respondent fatigue from multiple ratings of thelarge sample of ads, limited the number of scales that could be employed.

Results

Hl: Use of Reality-based and Unreal, Fantasy-based Contrasts

Alden et al. (1993) reported that 69 percent of the ads that attemptedhumor in their sample used contrast resolution mechanisms hypothesizedby Raskin (1985). We also found Raskin contrasts to be present in a largemajority (92 percent) of ads. Furthermore, the proportions of contrasttypes, within ads using contrasts, were found to be similar to thosereported by Alden et al. (1993): 48 percent reality-based contrasts and48 percent unreal, fantasy-based contrasts versus 44 percent and 50 percentrespectively in their study. Two ads that did not employ either type of

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 14: Extending a contrast resolution model of humor in television advertising: the role of surprise

206 D. L. Alden, A. Mukherjee and W. D. Hoyer

contrast were eliminated from subsequent analysis, leaving 50 ads forhypothesis testing.

H2: Contrast Type and Perceived Humor

As predicted by H2, ads with reality-based contrasts were perceived ässignificantly more humorous (X=4.64) than ads with unreal, fantasy-based contrasts (X=3.89, t[44] = 2.35, p < 0.01). This result replicatesAlden and Hoyer (1993) and provides the starting point for investigatingthe role of other variables, such äs surprise, in mediating this effect.

H3-H5: Type of Contrast, Strength of Contrast, Imageryand Surprise

To test the effects of type of contrast, strength of contrast, and extent ofimagery evoked on surprise, a regression analysis was performed. Beforefitting the regression equation, we conducted the KS test to check theregression assumption of normal distribution of the dependent variableand found strong support for this assumption (z = 0.90, p < 0.39).Multicollinearity was also not a threat to stability of the results, äs allthe variance Inflation factors were less than 2.0, again less than the cutoffof 10 recommended by Neter, Wasserman and Kutner (1985, p. 392).The results of this analysis are shown in Table l.

H3 predicts that ads using reality-based contrasts would have agreater mean surprise rating than ads using EU contrasts. As shown inTable l, the effect of type of contrast on surprise was only marginallysignificant (p < 0.06), although in the predicted direction (Mreaiity=4.15,^fantasy = 3.87). Thus H3 reccived marginal support.

Hypothesis H4 proposes that the strength of the contrast in an ad has apositive linear relationship with surprise. Consistent with this hypothesis,

Table 1. Regression

Dependent variable

Surprise

Humor

analysis: Perceptions of humor in advertising

Independent variable

Type of contrastStrength of contrastImagerySurprise

Beta coefficient

0.190.630.240.94

p value

p < 0.06p < 0.001p < 0.02p < 0.001

Adj.R2

0.510.88

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 15: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 207

a strong positive relationship between strength of contrast and perceivedsurprise was indicated by a significant positive beta coefficient (/?=0.63,p < 0.001). Thus it appears that the strength of contrast is more importantthan type of contrast in generating surprise.

Finally, H5 predicts that the degree of imagery evoked by the ad has apositive effect on surprise. Consistent with H5, we found that the degreeof imagery evoked has a significant positive effect on surprise (j8=0.24,p < 0.02). We also found that the set of statistically significant antecedentsto surprise (strength of contrast and degree of imagery evoked) explainedmore than 50 percent of its variance. Hence these variables constitute apowerful set of antecedents to surprise.

H6: Surprise and Humor

Hypotheses H6 proposes a direct linear relationship between surprise andperceived humor. A regression of surprise on humor was used to test thishypothesis (see Table 1). As shown in Table l, the regression equation givesan overall adjusted R square of 0.88 [F(2,43)= 198.9, p < 0.001], and asignificant positive beta coefficient for surprise (/?=0.94, p < 0.001). Thus,a strong linear relationship, which explains more than 80 percent of thevariance in humor, is in evidence, supporting hypothesis H6. This findingassumes added significance when compared to the approximately 11 percentofhumor variance explained by the type of contrast alone. Thus, surprise,although a conceptually distinct construct from humor (see Brownell et al.1983; Rüssel, Suzuki, and Ishida 1993; Niepel et al. 1994), appears to be apowerful antecedent to perceived humor in television advertising.

H7: Mediating Role of Surprise

In order to confirm the implicit mediating role of surprise in the proposedmodel ofhumor (Figure 1), we performed a modified 2 test, similar toHastak and Olson (1989). Squares of partial correlations were used äsSurrogates for effect coefficients generally used in ANOVA analysis. Themediating role of surprise in the model was confirmed since its inclusionled to a 99 percent, 98 percent, and 72 percent drop in the squared partialcorrelations of strength of contrast, degree of imagery evoked, and typeof contrast respectively, with perceived humor.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 16: Extending a contrast resolution model of humor in television advertising: the role of surprise

208 D. L. Alden, A. Mukherjee and W. D. Hoyer

H8: Humor and Aa<^

Finally, hypothesis H8 predicts that perceived humor has a positive effecton attitude towards the ad (Aad). To test this notion, attitude towards thead was correlated with the humor ratings. Results indicated a strongrelationship between humor and Aad [r = 0.95, p < 0.001]. Hence H8 isstrongly supported, and perceived humor does have a significant positiveeffect on evaluation of the ad. Figure 2 summarizes the results of theforegoing analysis and shows the significant model of humor äs supportedby the data.

Path Analysis

Finally, using the significant standardized parameter estimates obtainedin the preceding analysis, a path analysis was performed to assess thedirect, indirect, and total effects of each of the antecedent variableson perceived humor (see Duncan 1966 for a description of this pro-cedure). Such analysis is based on specifying all possible effects that arecontained a correlation and then estimating the amount of correla-tion attributable to each effect. Hence path analysis uses the individualstandardized regression coefficients äs the basis for empirical estimation

fTvoeofContrastr [Strength of Contrast]

| Perceived Humor)

Attitude towards the Ad

* This path is marginally significant @ p<.06. All others are significant @ p<.05.

Figure 2. Significant model of humor in advertising

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 17: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 209

Table 2. Effects ofantecedent variables on humor

Dependentvariable

Humor

AntecedentVariable

Type ofcontrastImagerySurpriseStrengthof contrast

Directeffect

0.20-0.94

-

Indirecteffect

0.17*0.21-

0.56

Totaleffect

0.37*0.210.94

0.56

* Marginally SignificantNote: Effects were computed using only statistically significant structural parameters.

and decomposition of the strength each implied causal relationshipdepicted in the model (Figure 1). The results of this analysis are reportedin Table 2.

The most striking result of the preceding analysis is the powerful role ofsurprise äs a mediator in the process ofhumor evocation. In addition, wealso identify the strength of contrast and the degree of imagery evoked ässignificant antecedents to surprise.

Discussion and future research

An initial contribution of the present study is to replicate the Undingsof previous researchers. Thus, äs reported in Alden and Hoyer (1993),we found that Raskin's (1985) contrast structures (reality-based versusunreal, fantasy-based) are widely used in television advertising thatattempts humor. Second, the linkage between these contrast typesand perception of humor in television advertising was reaffirmed withreality-based contrasts judged by viewers to be more humorous thanunreal, fantasy-based contrasts. The robustness of this result suggests thatRaskin's contrast model has an important nomological role to play inthe future development of theory regarding humor in advertising. Finally,we also observed that perceived humor had a strong positive effect onattitude towards the ad. This is of practical significance äs it strengthenscurrent beliefs about the ability of humor to generate positive effectson brand attitude via attitude towards the ad, especially under lowelaboration conditions (see Brown and Stayman 1992).

The results of this study also break new theoretical ground bydescribing the effects of some additional antecedents and mediators of

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 18: Extending a contrast resolution model of humor in television advertising: the role of surprise

210 D. L. Alden, A. Mukherjee and W. D. Hoyer

humor, such äs strength of contrast, degree of imagery evoked andsurprise. Notably, surprise was shown to be the most powerful antecedentof a humorous response, explaining over 80 percent of the observedvariance in humor ratings. It was also shown to mediate the effects ofother antecedent variables such äs type of contrast, strength of contrast,and imagery on humor, explaining more than 70 percent of the effect ofthese variables on humor. This would suggest that surprise is the drivingforce in determining the humorousness of an advertising, and thatsurprise would be an extremely useful variable in predicting when ahumorous campaign will actually be successful.

As discussed earlier, this central role of surprise äs an antecedentto humor is consistent with Suls (1972) conceptualization of humor äsa labeled psychological release resulting from a two stage process:perception of incongruity, followed by its resolution. Thus in mostsuccessful humorous executions — and this study focused on adsprejudged to be humorous — surprise should be a strong determinant ofoverall perceived humor. This central role of surprise is consistent withearlier research (e.g., McGhee and Johnson 1975; Brownell et al. 1986;Deckers and Winters 1986) that found a positive correlation betweensurprise and humor in various contexts.

Another important finding was that key role of strength of thedominant contrast in determining the level of surprise generated by ahumorous ad. Strength of contrast in an ad is an executional element,which can be manipulated by contrasting different elements of the ad likeSituation, visuals, music, or color scheme. The results of this studyindicate that using sharply distinct versions of these elements in the samead may be an effective way to create surprise, which in turn is the keyantecedent of humor.

Along with strength of contrast, the type of contrast in Operation wasexpected to influence perceived surprise (i.e., reality-based contrasts wereexpected to be more surprising than fantasy-based contrasts). However,this effect was only marginally significant. This outcome may be aconsequence of ads being clustered in the moderate incongruity region(i.e., both reality-based and fantasy-based contrasts having moderate orlower levels of incongruity). This clustering effect would be furtheraccentuated by a possible downward adjustment of the consumerreference point for judging incongruity due to the experimental contextof watching a series of humorous ads. That is, the subjects in this studywere exposed to a large number of humorous ads; after a while they would

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 19: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 211

have seen a variety of incongruous ads and might also become primedto expect incongruity. As a consequence, their Standards for judgingthe extent of incongruity may go up, leading to depressed perceivedincongruity which would be reflected in the surprise ratings. Note alsothat almost 50 percent of the ads in this study contained more than onecontrast, and it is possible that this could also account for the obtainedlack of difference between reality and fantasy-based contrasts on surprise.The ambiguity of the effect of type of contrast on surprise furtherhighlights the importance of strength of contrast in the generation ofsurprise and humor.

It was also found that the degree of imagery evoked has a significantpositive effect on surprise. Assuming that the contrast resolutionmechanism underlies humor — and an overwhelming majority ofhumorous ads were seen to use this mechanism — the results supportthe idea that imagery-oriented executions may increase comprehensionof the contrast and thereby heighten surprise. For advertisers, theimplication is that contrast resolution executions that utilize imageryevoking elements such äs visuals of a sea cruise or nostalgic music, willachieve higher surprise response.

The study also identifies several other avenues for future research. Itwas mentioned earlier that the perception of incongruity (surprise), and itsresolution are the key underlying antecedents to the labeled psychologicalrelease called humor. While the resolution process has been proposed towork in conjunction with surprise to evoke humor, the resolution processwas not specifically examined in this study. Future research shouldexamine the effect of different kinds of resolution and the timing ofthe resolution, on perceived humor. In addition to the perception ofincongruity and its resolution, it has also been argued that a "playful"context is necessary for successful evocation of humor. This contextualfactor distinguishes humorous executions from other kinds of appeals likefear and sex, and constitutes another prerequisite for successful humor.In this study since ads were preselected to be humorous, most of them canbe expected to include resolution cues äs well äs a "playful" setting.Hence both the factors discussed above have been, in a sense, favorablycontrolled in this study. In a future study, the precise impact of thesefactors may be isolated and measured.

In conclusion, we have proposed and tested a model of humor intelevision advertising and shown that it possesses high explanatory power.As mentioned earlier, the reason for the strength of this model is that it

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 20: Extending a contrast resolution model of humor in television advertising: the role of surprise

212 D. L. Alden, A. Mukherjee and W. D. Hoyer

represents what ads can do, äs opposed to what ads actually do — aconsequence of our sampling of a wide variety of humorous ads. Despitethis conscious effort to explore the füll ränge of humor in advertising, wesuspect that commercial objectives of advertising probably limit Variationin the use of discrepancy äs an antecedent to humor. Future studies withnon-commercial and artificially created advertising would validate andextent the conclusions of this study. These results also give direction toadvertisers in developing optimal messages to achieve high humorousness.Juxtapositioning sharply opposite versions of ad elements like Situation,visuals, music, or color scheme in an imagery evoking execution to inducesurprise, and using surprise ratings of ads äs a key variable to pretesthumorous ads for effectiveness.

University of Hawaii and University of Texas

Notes

Correspondence address: [email protected]. Before starting the coding task, all coders were intensively trained on the details of

the coding task and the dimensions of the constructs being measured. This was done by(a) presenting definitions and examples of the three kinds of contrasts in terms ofthe conceptualization of Raskin (1985), (b) screening a example set of three ads for eachtype of contrast. These best examples had been identified in an earlier study (Alden andHoyer 1993) äs being the most typical of the type of contrast in question on thebasis of maximum inter-coder agreement, and (c) conducting a warm-up coding taskwhere coders were independently asked to identify the type of contrast in another setof fifteen ads. The contrasts operating in these fifteen ads had been consensually iden-tified by a different set of coders in an earlier study (Alden and Hoyer 1993). Hence coderresponses in the present warm-up task could be compared against this benchmark. Coderaccuracy in identifying the contrast in this warm-up task ranged between 86 percentand 93 percent. The other constructs being measured (like strength of contrast,degree of imagery Stimulation, etc.) were also explained to the coders, both verbally andon the coding sheet. This method of training the coders is consistent with ourtheoretical development, and also parallele the approach taken by Alden and Hoyer(1993), facilitating replication and extension of their work.

References

Aaker, David A., Douglas M. Stayman, and Michael R. Hagerty1986 "Warmth in Advertising: Measurement, Impact and Sequence Effects,"

Journal of Consumer Research 12(March), 365-381.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 21: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 213

Alden, Dana L. and Wayne D. Hoyer1993 "An Examination of Cognitive Factors Related to Humorousness in

Television Advertising," Journal of Advertising 22(2), 29-37.Alden, Dana L., Wayne D. Hoyer, and Chol Lee

1993 "Identifying Global and Culture-Specific Dimensions of Humor inAdvertising: A Multinational Analysis," Journal of Marketing 57(April),64-75.

Anderson, John. R1983 "A Spreading Activation Theory of Memory," Journal of Verbal Learning

and Verbal Behavior 22, 261-295.Attardo, Salvatore, Donalee H. Attardo, Paul Bates, and Marnie J. Petray

1994 "The Linear Organization of Jokes: Analysis of Two Thousand Texts,"Journal of Humor 7(1), 27-54.

Baker, Ronald1993 "Some Reflections on Humor in Psychoanalysis," International Journal of

Psycho-Analysis 74(5), 951-960.Batson, C. Daniel, Laura L. Shaw, and Kathryn C. Oleson

1992 "Differentiating Affect, Mood, and Emotion," in Emotion, MargaretS. Clark (ed.), Newbury Park, CA: Sage Publications, Inc., 294-322.

Berger, Ida E. and Andrew A. Mitchell1989 "The Effect of Advertising on Attitüde Accessibility, Attitüde Confidence,

and the Attitude-Behavior Relationship," Journal of Consumer Research16(3), 269-279.

Bower, Gordon1972 "Mental Imagery and Associative Learning," in Cognition in Learning and

Memory, ed. Lee Gregg, New York: John Wiley.Brown, Steven P. and Douglas M. Stayman

1992 "Antecedents and Consequences of Attitüde Towards Ad: A Meta-Analysis," Journal of Consumer Research 19(June), 34-51.

Brownell, Hiram H., Dee Michel, John Powelson, and Howard Gardner1983 "Surprise But Not Coherence: Sensitivity to Verbal Humor in Right

Hemisphere Patients," Brain and Language 18(1), 20-27.Bugelski, B. Richard

1983 "Imagery and the Thought Processes," in Imagery: Current Theory,Research and Application, ed. Aness Sheik, New York: John Wiley, 72-95.

Chattopadhyay, Amitava and Kunal Basu1989 "Prior Brand Evaluation äs a Moderator of the Effects of Humor in

Advertising," Journal of Marketing Research 26(4), 466-476.Cho, Hyongoh

1994 "Antecedents of Perceived Humorousness and their Relationships toHumorous Devices in Advertising," Working Paper, Department ofAdvertising, The University of Texas at Austin.

Deckers, Lambert, and Jo A. Winters1986 "Surprise and Humor in Response to Discrepantly Short and/or Heavy

Stimuli in a Psychological Task," Journal of General Psychology 113(1),57-63.

Deckers, Lambert, Kuhlhorst and L. Freeland1987 "The Affects of Spontaneous and/or Voluntary Facial Reactions on

Surprise and Funniness Ratings," Motivation and Emotion 11, 403-412.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 22: Extending a contrast resolution model of humor in television advertising: the role of surprise

214 D. L. Alden, A. Mukherjee and W. D. Hoyer

Duncan, Otis Dudley1966 "Path Analysis: Sociological Examples," American Journal of Sociology,

72(July), 6-35.Fiske, Susan T. and Mark A. Pavelchak

1986 "Category-based versus Piecemeal-based Affective Responses: Develop-ments in Schema-triggered Affect," in Affect and Cognition: The 17th AnnualCarnegie Symposium on Cognition, Margaret S. Clark and Susan T. Fiske(eds.), Hillsboro, NJ: Erlbaum, 55-78.

Frijda, Nico H., Andrew Ortony, Joep Sonnemans, and Gerald L. Clore1992 "The Complexity of Intensity: Issues Concerning the Structure of Emotion

Intensity," in Emotion, Margaret S. Clark (Ed.), Newbury Park, CA: Sage,25-59.

Gelb, Betsy D. and George M. Zinkhan1985 "The Effect of Repetition on Humor in a Radio Advertising Study," Journal

of Advertising 14(4), 13-20.Goodstein, Ronald C.

1993 "Category-based Applications and Extensions in Advertising: MotivatingMore Extensive Ad Processing," Journal of Consumer Research 20(1),87-99.

Grunert, Klaus G.1996 "Automatic and Strategie Processes in Advertising Eifects," Journal of

Marketing 60(October), 88-101.Hastak, Manoj and Jerry C. Olson

1989 "Assessing the Role of Brand-Related Cognitive Responses äs Mediatorsof Communication Effects on Cognitive Structure," Journal of ConsumerResearch 15(March), 444^456.

Herzog, Thomas R. and David A. Larwin1988 "The Appreciation of Humor in Captioned Cartoons," Journal of

Psychology 122(6), 597-607.Homer, Pamela M.

1990 "The Mediating Role of Attitüde Towards Ad: Some Additional Evidence,"Journal of Marketing Research 27(February), 78-86.

Izard, Carroll E.1977 Human Emotions, NY: Plenum Press.

James, William1884 "What is an Emotion?," Mind 9, 188-205.

Kassarjian, H. H1977 "Content Analysis in Consumer Research," Journal of Consumer Research

4, 8-18.Krishnan, H. Shankar and Dipankar Chakravarti

1990 "Humor in Advertising: Testing Effects on Brand Name and Message ClaimMemory," Proceedings of the Summer Educators' Conference, WilliamBearden and A. Parasuraman, eds. Chicago: American MarketingAssociation, 10-16.

Krugman, Herbert E.1965 "The Impact of Television Advertising: Learning Without Involvement,"

Public Opinion Quarterly 29(Fall), 349-356.Larsen, Randy J. and Edward Diener

1992 "Promises and Problems with the Circumplex Model of Emotion," inEmotion, Margaret S. Clark (ed.), Newbury Park, CA: Sage, 25-59.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 23: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 215

Levenson, R. W., P. Ekman, and W. V. Friesen1990 "Voluntary Facial Action Generates Emotion-Specific Nervous System

Activity," Psychophysiology 27, 363-384.Mandler, George

1982 "The Structure of Value: Accounting for Taste," in Affect and Cognition:The 17th Annual Carnagie Symposium, Margaret S. Clark and Susan T.Fiske eds., Hillsdale, NJ: Lawrence Erlbaum Associates, 3-36.

McGhee, Paul E. and Susan F. Johnson1975 "The Role of Fantasy and Reality Cues in Children's Appreciation of

Incongruity Humor," Merril-Palmer Quarterly 21(1), 19-30.Mclnnis, Deborah J. and Bernard J. Jaworski

1989 "Information Processing from Advertisements: Toward an IntegrativeFramework," Journal of Marketing 53(October), 1-23.

Mclnnis, Deborah J. and Linda L. Price1987 "The Role of Imagery in Information Processing: Review and Extensions,"

Journal of Consumer Research 13(March), 473-491.Meyer, Wulf-Uwe

1986 "The Role of Surprise in the Attribution Process," PsychologischeRundschau 39(3), 136-147.

Meyer, Wulf-Uwe and Michael Niepel1994 "Surprise," Encyclopedia of Human Behavior, Volume 4, V. S. Ramachadran

(ed.), San Diego: Academic Press, 353-358.Meyer, Wulf-Uwe, Michael Niepel, Udo Rudolph, and Achim Schutzwohl

1991 "An Experimental Analysis of Surprise," Cognition and Emotion 5(4),295-311.

Meyers-Levy, Joan and Alice Tybout1989 "Schema Congruity äs a Basis for Product Evaluation," Journal of

Consumer Research 16(June), 39-54.Moore, David J., William Harris, and Hong C. Chen

1995 "Affect Intensity: An Individual Difference Response to AdvertisingAppeals," Journal of Consumer Research, 22(2), 154-164.

Neter, John, William Wasserman, and Michael H. Kutner1985 Applied Linear Statistical Models, 2nd. ed., Homewood, IL: Richard

D. Irwin Inc.Niepel, Michael, Udo Rudolph, Achim Schützwohl, and Wulf-Uwe Meyer

1994 "Temporal Characteristics of the Surprise Reaction Induced by Schema·discrepant Visual and Auditory Events," Cognition and Emotion 8(5), 433-452.

Oppliger, Patrice A. and John Sherblom1988 "Late Night With David Letterman: A Humorous Balance," Communi-

cation Research Reports 5(2), 193-196.Ortony, Andrew and Terence J. Turner

1990 "What's Basic About Basic Emotions?," Psychological Review 97(3),315-331.

Osterhouse, Robert and Timothy Brock1970 "Distraction Increases Yielding to Propaganda by Inhibiting Counter-

arguing," Journal of Personality and Social Psychology, 15(August), 344-358.Petty, Richard E. and John T. Cacioppo

1986 "The Elaboration Likelihood Model of Persuasion," in Advances inExperimental Social Psychology, Vol. 19, Leonard Berkowitz, ed. NewYork: Academic Press Inc., 123-205.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 24: Extending a contrast resolution model of humor in television advertising: the role of surprise

216 D. L. Alden, A. Mukherjee and W. D. Hoyer

Raskin, Victor1985 Semantic Mechanisms of Humor, Boston: D. Reidel.

Rüssel, James A., Naoto Suzuki, and Noriko Ishida1993 "Canadian, Greek, and Japanese Freely Produced Emotion Labels for

Facial Expressions," Motivation and Emotion 17(4), 337-351.Scott, Cliff, David M. Klein, and Jennings Bryant

1990 "Consumer Response to Humor in Advertising: A Series of Field StudiesUsing Behavioral Observation," Journal of Consumer Research 16(March),498-501.

Shaver, Phillip R., Shelley Wu, and Judith C. Schwartz1992 "Cross-Cultural Similarities and Differences in Emotion and Its

Representation: A Prototype Approach," in Emotion, Margaret S. Clark(ed.), Newbury Park, CA: Sage, 175-212.

Sheehan, Peter1972 "A Functional Analysis of the Role of Visual Imagery in Unexpected

Recall," in The Function and Nature of Imagery, ed. Peter Sheehan,New York: Academic Press, 149-174.

Shimp, Terence A.1996 Advertising, Promotion and Supplemental Aspects of Integrated Marketing

Communications, Fort Worth, TX: The Dryden Press.Speck, PaulS.

1991 "The Humorous Message Taxonomy: A Framework for the Study ofHumorous Ads," in Current Issues & Research in Advertising 13,James H. Leigh and Claude R. Martin Jr. eds., Ann Arbor, MI:Michigan Business School.

Speck, Paul Surgi1987 "On Humor and Humor in Advertising," Unpublished doctoral dissertation,

Texas Tech University.Stayman, Douglas, Dana L. Alden, and Karen H. Smith

1992 "Some Effects of Schematic Processing on Consumer Expectations andDisconfirmation Judgements," Journal of Consumer Research 19(2),240-255.

Stewart, David M. and David H. Furse1986 Effective Television Advertising, Lexington, MA: D.C. Heath & Co.,

Chicago.Suls, Jerry

1972 "Two-stage Model for the Appreciation of Jokes and Cartoons: Infor-mation Processing Analysis," in J. H. Goldstein, and P. E. McGhee eds.,The Psychology of Humor, 81-100, San Diego, CA: Academic Press.

1983 "Cognitive Processes in Humor Appreciation," in Handbook of HumorResearch, Jeffery Goldstein ed., New York: Springer-Verlag Inc., 39-57.

Unger, Lynette S.1995 "A Cross-Cultural Study of the Affect-Based Model of Humor in

Advertising," Journal of Advertising Research January/February, 66-69.1996 "The Potential for Using Humor in Global Advertising," Humor 9(2),

133-168.Weinberger, Marc G. and Charles S. Gulas

1992 "The Impact of Humor in Advertising: A Review", Journal of Advertising21(4), 35-39.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 25: Extending a contrast resolution model of humor in television advertising: the role of surprise

Extending a contrast resolution model ofhumor 217

Weinberger, Marc G. and Harlan E. Spotts1989 "Humor in U.S. Versus U.K. TV Advertising," Journal of Advertising 18(2),

39-44.Weinberger, Marc G. and Leland Campbell

1991 "The Use and Impact of Humor in Radio Advertising," Journal ofAdvertising Research 31(December/January), 44-52.

Wicker, F. W., I. M. Thorelli, W. L. Barron III and M. R. Ponder1981 "Relationships Among Affective and Cognitive Factors in Humor," Journal

of Research in Personality 15, 359-370.Wright, Alice A. and John G. Lynch

1995 "Communication Effects of Advertising and Direct Experience WhenBoth Search and Experience Attributes are Present," Journal of ConsumerResearch 21(4), 708-718.

Zhang, Yong and George M. Zinkhan1991 "Humor in Television Advertising," Advances in Consumer Research, 18,

Rebecca H. Holman and Michael R. Solomon eds. Provo, UT: Associationfor Consumer Research, 813-818.

Zinkhan, George M. and Betsy D. Gelb1990 "Humor Repetition and Advertising Effectiveness," in Advances in

Consumer Research, 17, Marvin E. Goldberg, Gerald Gorn, and RichardW. Pollay, eds. Provo, UT: Association for Consumer Research, 438-441.

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM

Page 26: Extending a contrast resolution model of humor in television advertising: the role of surprise

Brought to you by | University of ArizonaAuthenticated

Download Date | 12/18/14 1:19 AM