the impact of humor in advertising: a reviengrant/cjt780/readings/day 9/weinbergergulas1992.pdf ·...
TRANSCRIPT
The Impact of Humor in Advertising: A ReviewMarc G. Weinberger and Charles S. Gulas
The use of humor has become common practice in advertising; yet our knowledge about its impact has notbeen updated since the last major review almost twenty years ago. In the interim, a great deal of humorresearch has been conducted. The outcome of this research only partially supports earlier conclusions andhighlights the need to apply humor with care. Humor is by no means a guarantee of better ads, but its effectcan be enhanced with careful consideration of the objectives one seeks to achieve as well as the audience,situation, and type of humor.
Marc G. Weinberger (Ph. D.,Arizona State University) is Professorof Marketing, Department ofMarketing, School of Management,University of Massachusetts atAmherst, Amherst, Massachusetts.Charles S. Gulas CM.B.A. Youngs-town State University) is a DoctoralStudent, Department of Marketing,School of Management, University ofMassachusetts at Amherst, Amherst,Massachusetts.
Journal of Advertising,Volume XXI, Number 4December 1992
"People do not buy from clowns."Claude Hopkins 1923
"Good copywriters have always resisted the temptation to entertain."David Ogilvy 1963
"I have reason to believe that... humor can now sell."David Ogilvy 1982
Introduction
Estimates of the use of hiimor in advertising suggest that as much as24.4% of prime time television advertising in the U.S. is intended to behumorous (Weinberger and Spotts 1989). Research conducted by others hasalso indicated similar high (or even higher) levels of usage of humor intelevision ads (Kelly and Solomon 1975; Markiewicz 1972; Speck 1987) andin radio (Weinberger and Campbell 1991). While tbe use of humor is high,
' tbe efficacy of humor as a communications device remains uncertain. Inattempts to delineate its impact, humor has proven to be very elusive. Thislack of knowledge has led advertising copjrwriters and researchers alike toboth praise and decry the effectiveness of humor in advertising as evidencedin tbe opening quotes.
The fact is that humor is a complex topic that has been experimentallystudied by advertisers in several dozen studies over the past twenty-fiveyears. Humor is a multifarious concept that is affected by a wide vEiriety offactors. As a result of the many contingencies imposed by desired goal, typeof humor, medium, placement and audience (see Figure), generalizationsabout tbe effect of bumor are fraught witb pitfalls (Stewfirt-Hunter 1985).Though the broad question of humor's effectiveness in advertising is unan-swerable, we can compile the accounts of humor research in tbe context ofproper constraints to gain insights about its effects. Therefore, the moreappropriate questions to ask are: 1) What communications goals are mostlikely to be achieved tbrough the use of hximor?; 2) What executional ormessage factors are likely to affect the outcome?; 3) For what audience ishumor most appropriate?; and 4) What product factors suggest the use ornon-use of a humorous approach? The purpose of this paper then is tosystematically examine tbe research tbat has been conducted to gain insightinto the effects of humor with regard to these questions.
Journal of Advertising
Humor Research
The widespread use of humor, coupled with theunresolved questions regarding it, has drawn the at-tention of numerous communication researchers. Ina frequently cited review of the early literature in thefield, Stemthal and Craig (1973) drew some tentativeconclusions about the use of humor on a number ofcommunications goals. These conclusions must beviewed as tentative because, although based on athorough review of the extant literature in 1973, thisliterature base was somewhat small and consistedalmost exclusively of non-advertising studies as therewas simply little prior work in advertising to review.
In the years since the Stemthal and Craig work,humor has received extensive further investigationin over 30 studies that have appeared in the marketingliterature, and a great many more studies that haveappeared in the literature streams of education,communication and psychology. This paper synthe-sizes the relevant aspects of this literatvire in order toupdate and expand on the Stemthal and Craig work.Thus, the format to be followed will be to examine theeffect of humor as it applies to various communicationsgoals and then to expand on this work by includingexecution, placement, audience, and product factorsthat have come to light in the past twenty years.
Communications Goals
As alluded to earlier, the nature of the communica-tion goal plays a major role in the appropriateness ofthe use of humor. Stemthal and Craig (1973) listedadvertising goals and the impact of humor on each ofthese goals. Revisited after twenty years of interven-ing research some of these conclusions remain cogent,while others appear to be in need of revision.
Humor and Attention
Studies have shown that 94% of advertising practi-tioners see humor as an effective way to gain atten-tion. Furthermore, 55% of advertising research ex-ecutives believe humor to be superior to non-humorin gaining attention (Madden and Weinberger 1984).While the personal views of advertising executivesshould not be equated with rigorous hypothesis test-ing, these views do refiect a knowledge base built onyears of day to day experience with proprietary re-search results. And in the case of attention, thesepractitioner views appear to be well supported by theavailable empirical evidence. In studies of actual
magazine ads (Madden and Weinberger 1982), televi-sion ads (Stewart and Furse 1986), and radio ads(Weinberger and Campbell 1991) in standard indus-try ad testing situations, humor has been found tohave a positive effect on attention (see Table 1). Simi-larly, this attention effect has also been demonstratedin the laboratory. In a thorough test of attention effectsin the advertising arena. Speck (1987) compared hu-morous ads with non-humorous controls on four at-tention measures: initial attention, sustained atten-tion, projected attention and overall attention. Hefoiind humorous ads to outperform non-humorous adson each of the attention measures.
The attention-attracting ability of humor has alsobeen demonstrated in education research (Powell andAndresen 1985; Zillmann et al. 1980). In a review ofthe education literature, Bryant and Zillmann (1989)conclude that humor has a positive effect on attention;however, they caution that "unqualified direct evi-dence for the effects of using humor in non-mediatedclassroom instruction is still wanting" (p. 59). Thecautionary stance taken by Bryant and Zillmann isappropriate for all the hvtmor-attention studies. Whilethe results seem to indicate a positive impact on at-tention, and in general the past twenty years of re-search largely supports the conclusion drawn byStemthal and Craig (1973) (see Table 1), future re-searchers should be aware that all humor is not cre-ated equal. Related humor, that is, humor directlyconnected to the product or issue being promoted,appears to be more successful than unrelated humor(Duncan 1979; Lull 1940; Madden 1982). In fact,controlling for the relatedness factor makes the find-ings of the experimental studies in advertisingunanimous in their support for a positive effect ofhimior on attention. This indicates that the mereinsertion of "canned" humor into a given ad is unlikelyto have the same impact on attention as the use of amore integrated humor treatment.
Humor and Comprehension
The literature is mixed on the effect that humorhas on comprehension. In a study of 1000 broadcastcommercials, Stewart and Fiorse (1986) found humor-ous content to increase the comprehension of an ad.Other studies have found similar positive results(Duncan, Nelson and Frontczak 1984; Weinbergerand Campbell 1991; Zhang and Zinkhan 1991).However, these studies contrast sharply with the re-sults of other advertising researchers who have founda negative relationship between humor and compre-
December 1992 37
Table 1The Impact of Humor on Attention
Advertising StudiesAuthor(s) & Date Type of Study & Subjects Medium Finding Comment
Stemthal & Craig (1973)
Duncan (1979)
McCollum/Spielman (1982)
Madden (1982)
Madden & Weinberger(1982)
Madden & Weinberger(1984)
literature review
literature review
study of 500 commer-cials from data base,target audiences
lab experiment,326 undergraduates
data-based study of148 liquor ads fromStarch
survey of 140 U.S.ad executives
N/A
N/A
TV
radi
prin
N/A
Duncan & Nelson (1985)
Stewart & Furse (1986)
Speck (1987)
Weinberger & Spotts(1989)
Wu, Crocker & Rogers(1989)
lab experiment, 157 radiomale undergraduates
data-based, study of TV1000 pre-tested ads
lab experiment, TV182 undergraduates
survey of advertising N/Aexecutives, 132 U.S.agencies, 29 U.K.agencies
lab experiment print360 undergraduates
mixed
mixed only related humoreffective
humorous ads out-performed non-humorous ads on "noted,""seen-associated," and"read most" recall measures
humor outperformsnon-humor on 4attention measures
Non-Advertising StudiesAuthor(s) & Date Type of Study & Subjects Medium Finding Comment
Lull (1940)
Gruner(1970)
Ziilmannetal. (1980)
Powell & Andresen (1985)
experiment,1016 undergraduates
communicationexperiment
lab experiment, 70children ages 5-7 years
publicspeech
tapedpublic speech
TV
literature review N/A
interestingness of speech,no advantage over non-humor, topic state medicine
when topic interestingwhen topic uninteresting
studied children's attentionto educational TV
children's attention toclassroom instruction
Bryant & Zillmannn (1989) literature review N/A conclusion drawn fromreview of educationresearch
38 Journal of Advertising
hension (Cantor and Venus 1980; Gelb and Zinkhan1986; Lammers et al. 1983; Sutherland and Middleton1983). This negative view of the effect of humor oncomprehension is shared by the majority of researchexecutives (64%) at U.S. ad agencies. In sum, of theadvertising experiments that attempted to measiu-ethe effects of humor on comprehension, six indicatehumor may enhance comprehension, five produceneutral or mixed findings, and six indicate that humormay harm comprehension (see Table 2). While thesefindings certainly fail to resolve the true effect ofhumor on comprehension, they do call into questionthe existence of a global negative effect hypothesizedby Stemthal and Craig (1973).
With a literature as discrepant as this, it is impor-tant to look for factors that may disentangle thesefindings. To this end, it appears that three factorsseem to explain much of the lack of agreement in thestudies. First, there is a lack of a consistent definitionof comprehension among studies. Depending on thespecific measure used, recall may be an indication ofcomprehension or it may merely indicate attention.More importantly, the measures employed may havean impact on the results found. Those studies thatemploy multiple or summated measures of compre-hension (Speck 1987; Weinberger and Campbell 1991)are more likely to find positive or mixed positive effectson comprehension than those studies that employsingle measures (Cantor and Venus 1980; Lammerset al. 1983), indicating that a positive comprehensioneffect may be missed by relatively narrow measures.Further evidence of the importance of measures isfound in the work of Murphy and his colleagues(Murphy, Cunningham and Wilcox 1979). Their studyof context effects demonstrates that different measuresof recall may produce different recall results.
Secondly, humor type may be an important deter-minant in comprehension effects. In one study whichdirectly compared the effects of various humor typeson comprehension. Speck (1987) found significant dif-ferences due to type. His findings indicate that somehumorous ads do better, and some do worse than non-humorous ads on descriptive and message compre-hension and that this differential performance wasattributable to humor type. "Comic wit" was found tounder-perform non-humorous treatments while allother humor tjrpes (i.e., satire, full comedy, senti-mental humor and sentimental comedy) out performedthe non-humor treatment.
Finally, the type of product advertised appears toplay an important role in the impact of humor oncomprehension. This product factor is composed of
two dichotomies, actual vs. fictional products, andhigh involvement vs. low involvement products. Thosestudies employing actual products (Speck 1987;Stewart and Fvirse 1986; Weinberger and Campbell1991; Zhang and Zinkhan 1991) in general indicate apositive effect of humor on comprehension. On theother hand, studies employing fictional products(Cantor and Venus 1980; Gelb and Zinkhan 1986)have found a negative effect of humor on comprehen-sion. However, two studies depart from this generalpattern. The first is an advertising experiment whichused a real industrial product (Lammers et al. 1983).However, since this study used a student sample,none of the subjects was familiar with the product oreven the product category; thus, for all practical pur-poses, this product can be viewed as fictional. There-fore, the negative findings in this study fall in linewith the actual product - fictional product dichotomypresented above. The second study that finds anegative comprehension effect for actual products usedhigh involvement products (luggage and 35mm cam-eras) infrequently purchased by the student sampleemployed (Sutherland and Middleton 1983). Thisfinding points up the other important product di-chotomy, high involvement - low involvement, thatwill be discussed later in the product section.
Given the equivocal findings of the advertising re-search, and the lack of clarity regarding the mea-surement of comprehension, we might do well to turnto non-advertising research to help clarify the issue.In education research, the effect of humor on com-prehension is tj^ically measxired via a written test.While this clearly cannot be claimed to be analogousto the conditions under which advertising is presentedor tested, we believe that these studies do provide arigorous test of the relationship between humor andcomprehension that can provide insight into the im-pact that h\iinor may have on advertising compre-hension. An analysis of the relevant non-advertisingstudies shows eight studies that report a positiveeffect of humor on comprehension and eleven studiesthat indicate a null or mixed effect. None of the non-advertising studies reports a negative effect of humoron comprehension, which again challenges the con-clusion drawn in 1973 by Stemthal and Craig.
Of the education literature, perhaps the strongestsupport for a positive relationship between humorand comprehension appears in work conducted by Ziv(1988). This study indicates that humor can signifi-cantly improve learning. The Ziv experiments com-pared an introductory statistics course that was pre-sented without humor with a course that included
December 1992 39
relevant humor. Both teacher and lecture materialswere held constant. The level of learning was mea-sured at the end of the semester by a standard objec-tive departmental final exam. The average score ofthe humor treatment class on this exam was over tenpercentage points higher than the average score inthe non-humor class. Ziv replicated this experimentwith two psychology classes and found very similarresults. The work conducted by Ziv is supported byother non-advertising researchers (e.g., Chapman andCrompton 1978; Davies and Apter 1980; Gorham andChristophel 1990; Kaplan and Pascoe 1977; Vance1987; Zillmann et al. 1980). This non-advertising lit-erature also supports the hypothesis stated abovethat humor type may moderate the impact of humoron comprehension. Work conducted by Vance (1987)in the education arena parallels that conducted bySpeck (1987) in advertising . Both of these research-ers have found significant effects for humor type. Theeducation literature also points out that relatednessof the humor to the message appears to be very im-portant with regard to comprehension. Studies usingrelated humor were more likely to find that humorenhanced comprehension than those employing un-related humor.
Overall, the inconclusive nature of the results sug-gests that the effect of humor on comprehension is anarea where additional research can be especiallyhelpful, and future researchers should be particularlycognizant of humor type, and relatedness. Advertisingresearchers might also be well advised to use actual,rather than fictional, products in manipulations andemploy several measures of comprehension.
Humor and Persuasion
Sternthal and Craig (1973) concluded that the dis-traction effect of humor may lead to persuasion.However, they note that the persuasive effect of hu-mor is at best no greater than that of serious appeals.These conclusions seem to agree with the opinions ofU.S. ad executives. Madden and Weinberger (1984)found that only 26% of these practitioners agreedwith a statement proclaiming humor to be more per-suasive than non-humor. While U.S. advertising ex-ecutives largely agree with the conclusion of Stemthaland Craig (1973), this opinion is in sharp contrast tothat of their British counterparts, 62% of whom viewedhumor as more persuasive than non-humor and only7% of whom were found to disagree with this assertion(Weinberger and Spotts 1989).
The literature in marketing and communication has
addressed this issue directly, and the evidence for apersuasive effect of humor is mixed at best. Speck(1987) found three out of five humor treatments in-creased two measures of persuasion: intent to use theproduct and change in perceived product quality.Similarly, in an experimental study, Brooker (1981)found a humorous appeal to be more persuasive thana fear appeal. However, neither humor nor fear ap-peals were more persuasive than a straight forwardapproach. An examination of commercials, publishedby McCollunVSpielman (1982), found that 31% of hu-morous commercials exhibited above average scoreson persuasiveness. This figure represents about aver-age performance when compared to other executionaltactics examined by McCollum/Spielman (1982).Stewart and Furse (1986) found no effect of humor onpersuasion. Finally, in their study of radio ads,Weinberger and Campbell (1991) found unrelatedhumor to perform the same or worse on a persuasionmeasure than no humor. Additionally, while relatedhumor was more persuasive than no humor for lowinvolvement-feeling products, it was found to be lesspersuasive on high involvement-thinking products.
Other advertising research also indicates that, muchlike comprehension, other factors may intervene tomoderate the effect of humor on persuasion. For ex-ample, while Lammers and his colleagues (Lammerset al. 1983) found a positive effect for humor on per-suasion, this effect was present only for males. Simi-larly, Chattopadhyay and Basu (1989) found a mod-erated positive persuasive effect for humor. In theirstudy, subjects with a prior positive brand attitudewere more persuaded by humorous treatments whilesubjects with pre-existing negative brand attitudeswere not.
Perhaps the strongest case for a persuasive effect ofhumor is presented in a study by Scott, Klein andBryant (1990), who employed a behavioral measureof persuasion quite different from the measures ofpersuasion used in other studies. They found thatattendance at social events (e.g., town picnics) wasgreater among subjects who received the humoroustreatment of an ad than among those who receivedone of two other types of promotions. The humortreatment was not found to increase attendance incomparison to the other type of promotions at businessevents (e.g., town council meetings). The support fora persuasive effect shown in the Scott, Klein andBryant study must, however, be viewed with cautionin the light of the studies which find no added per-suasive effect of humor (Belch and Belch 1984; Bryantet al. 1981; Duncan and Nelson 1985; Kennedy 1972;
40 Journal of Advertising
Table 2The Impact of Humor on Comprehension
Advertising StudiesAuthor(s) & Date
Stemthal & Craig (1973)
Duncan (1979)
Murphy, Cunningham& Wilcox (1979)
Cantor & Venus (1980)
Madden (1982)
Lammers, Liebowitz,Seymour & Hennessey(1983)
Sutherland & Middleton(1983)
Belch & Belch (1984)
Madden & Weinberger(1984)
Duncan, Nelson,Frontczak(1984)
Stewart & Furse (1986)
Type of Study & Subjects
literature review
literature review
lab experiment,115 undergraduates
lab experiment,117 undergraduates
lab experiment,326 undergraduates
lab experiment,64 undergraduatestarget audiences
lab experiment,107 undergraduates
lab experiment,184 undergraduates
survey of U.S. advertisingexecutives, 68 researchexecutives, 72 creativeexecutives
lab experiment, 157 maleundergraduates
data-based, study of1000 pre-tested ads
Medium
N/A
N/A
TV
radio
radio
radio
print ads
TV
N/A
radio
TV
Finding
-
0
mixed
-
mixed
—
—
0
+
+
Comment
unaided & aided recall ofcommercial and content
fictional products
only related humor effective,one familiar & one unfamiliarproduct
product uses and benefitrecall unfamiliar industrualproducts
recall, high involvement,infrequently purchasedproducts
unaided recall
mixed but generallynegative views
even 'lailed" (unfunny)humor, better than aserious ad
Gelb & Zinkhan (1986)
Nelson (1987)
Zinkhan & Gelb (1987)
Speck (1987)
Wu, Crocker & Rogers(1989)
lab experiment, 120 employed radioadult part-time students
re-examination of Gelb & Zinkhan N/A(1986)
reply to Nelson N/A(1987)
lab experiment, TV182 undergraduates
lab experiment, print360 undergraduates
N/A
N/A
mixed
summed measure of brandand copy recall, fictionalproduct
claims Gelb and Zinkhan'smeasure of recall is not valid
defends use of Cloze proce-dure as recall measure
global effect of humor Ispositive but some humorads weaker than some non-humorous ads
with high involvementproduct humor improvedunaided recall
Continued...
December 1992 41
Author(s) & Date Type of Study & Subjects Medium Finding Comment
Weinberger & Spotts(1989)
survey of advertising executives,132U.S. agencies, 29 U.K.agencies
Zhang & Zinkhan (1991) lab experiment, 216undergraduates
Weinberger & Campbell data based, pre-(1991) tested ads
N/A
TV
radio
more negative view ofhumor on comprehensionheld by U.S. executives
recall measure, 10questions about informationin ad
study of over 1600 radioads, positive effect notfound with unrelated humor(combined recall index)
Non-Advertising StudiesAuthor(s) & Date Type of Study & Subjects Medium Finding Comment
Gruner(1967)
Gruner(1970)
Kennedy (1972)
Markiewicz (1972)
Markiewicz(1972)
Markiewicz (1974)
Gruner(1976)
communication experiment,128 male undergraduates
communication experiment,144 undergraduates
communication experiment,200 undergraduates
55 7th grade students
86 undergraduates
literature review
literature review
Kaplan & Pascoe (1977) education quasi-experiment, 508 undergraduates
Chapman & Crompton(1978)
Davies & Apter (1980)
Zillmann et al. (1980)
Bryant, Brown, Silberberg& Elliott (1981)
Powell & Andresen(1985)
Vance (1987)
education experiment, childrensubjects ages 5 & 6 yrs
education experiment, 285children ages 8-11
education experiment, 70children ages 5-7
lab experiment, 180undergraduates
literature review
education experiment, 58first grade children
public speech
public speech mixed
slides
slide-tape
TV
textbookillustrations
N/A
audio-tape
mixed
humor enhanced a "dull"speech but did not enhancean "interesting" speech
TV, audio tape,live speech
written essays
written essays
N/A
N/A
TV instruction
0
0
0
0
0
mixed
topic of high personalreievence
immediate comprehensionnot enhanced but recallafter 3 months superior tonon-humor for related points
children's learning fromeducational TV
recall enhanced only foritems related to humor
concludes that humor haspositive effect on compre-hension and retention
effect moderated by typeof humor
Continued.
42 Journal of Advertising
Author(s) & Date Type of Study & Subjects Medium Finding Comment
Weaver, Zillman &Bryant (1988)
Ziv(1988)
Bryant & Zillmann(1989)
Gorham & Christophel(1990)
86 4th and 8th grade students TV
education experiment, 161 classroomundergraduates in first experiment instructionmale and female -132 undergrad-uates in replication all female
literature review N/A
correlational study, 206 classroomundergraduates observing 150 instructionmale and 54 female collegeteachers
semester-long experimentand semester-longreplication
correlates use of humorwith teaching effectiveness
Markiewicz 1972, 1974). Overall, the advertising lit-erature has produced five studies that indicate apositive effect of humor on persuasion, eight studieswhich indicate a neutral or mixed effect, and onewhich produced a negative effect. Among non-adver-tising studies, no positive results have been reported;seven neutral or mixed effects were found and onenegative effect (see Table 3). Some portion of theequivocal nature of these results appears to be at-tributable to an underlying factor that might bebroadly termed intensity of the message. Two studiesthat directly compared levels of intensity (Bryant etal. 1981; Markiewitz 1972) found a significant effectof message intensity on the persuasiveness of a hu-morous message. This intensity factor has two di-mensions: the intensity of the humor and the inten-sity of the surrounding message. Bryant and his col-leagues (1981) examined differences in levels of hvimorand found that the use of low levels of humor providedessentially the same level of persuasion as no humoruse, while extensive use of humor was detrimental topersuasion. The intensity of the surrounding messagewas examined by Markiewitz (1972). Her study re-vealed that the addition of humor to a low intensitysoft sell approach aided the level of persuasion whilethe addition of humor to a hard sell approach actuallyharmed persuasion. This level of intensity factor ap-pears to impact the level of persuasion garnered byhumorous messages. However, since little work hasbeen done in this area, conclusions cannot be drawn.In summary, our synthesis can be interpreted assupport for Stemthal and Craig's (1973) conclusionthat humor may be persuasive but probably no moreso than non-humor.
Humor and Source Credibility
The results of studies examining the effect of hu-mor on source credibility can best be described asmixed. The advertising studies exploring sourcecredibility have produced a smooth distribution ofresults with three advertising studies reporting en-hanced source credibility in humor conditions, fourindicating neutral or mixed effects, and three indi-cating a negative relationship. The non-advertisingstudies parallel these results.
These mixed results appear to be due to a numberof factors such as the nature of the source or nature ofthe humor. Bryant and his colleagues found the effectof humor on credibility to be moderated by gender ofsource. Studying the effectiveness of humor in thecollege classroom, they found that any positive rela-tionship between humor and credibility is slight andonly applicable to male professors (Bryant et al. 1980).Speck's (1987) work indicates that type of humor usedmay also influence humor's impact on credibility.Speck (1987) measured two aspects of source cred-ibility, "knowledgeableness," [sic] and "trustworthi-ness." He found that, while all sources in the experi-ment were viewed as moderately knowledgeable, thesources of non-hvimorous ads were viewed as moreknowledgeable than the humorous sources. However,trustworthiness of a source was demonstrated to beenhanced through the use of one specific humor type."Sentimental humor," a type of humor defined bySpeck as a combination of two humor processes,arousal-safety and incongruity-resolution, in whichthe process of empathy-anxiety-relief occurs, wasfound to outperform other h\imor treatments and non-
December 1992 43
Table 3The Impact of Humor on Persuasion
Advertising StudiesAuthor(s) & Date Type of Study & Subjects Medium
N/A
N/A
TV
Finding
0
mixed
0
+
Comment
miid liumor outperformedmiid fear but did notoutperform straight-forward ad
study of over 500 TVcommercials
Sternthal & Craig (1973) literature review
Duncan (1979) literature review
Brooker(1981)
McCollum/Spielman(1982)
Madden (1982)
Lammers, Leibowitz,Seymour & Hennessey(1983)
Belch & Belch (1984)
Hadden & Weinberger(1984)
Duncan & Nelson (1985)
Stewart & Furse (1986)
Speck (1987)
Chattopadhyay & Basu(1989)
Weinberger & Spotts(1989)
Scott, Kline & Bryant(1990)
lab experiment,240 adults
study of 500 commer-cials, data-basedtarget audiences
lab experiment,326 undergraduates
lab experiment,64 undergraduates
lab experiment, 184undergraduates
survey of U.S. ad-vertising executives, 68 researchexecutives, 72 creative executives
lab experiment, 157male undergraduates
study of 1000 pre-tested ads
lab experiment,182 undergraduates
lab experiment, 80subjects (undergraduates)
survey of advertising executives,132 U.S. agencies,29 U.K.agencies
field experiment,total N-73respondents N=513
radio
radio
TV
N/A
radio
TV
TV
TV
N/A
directmail
mixed
mixed
for malesfor females
mixed opinion
Specie found 3 of 5 hu-mor treatments to increaseboth perceived productquality and intent to useproduct.
when subject had favor-able prior brand attitude
mixed opinion U.K. ex-ecutives view humor asmore persuasive than U.S.executives
humorous ad increasedattendance at socialevents but not atbusiness events
Continued..
44 Journal of Advertising
Non-Advertising StudiesAuthor(s) & Date
Lull (1940)
Kennedy (1972)
Markiewicz (1972)
Markiewicz (1972)
Markiewicz (1972)
Markiewicz (1974)
Gruner(1976)
Bryant, Brown, Siiber-berg&Eiiiott(1981)
Type of Study & Subjects
experiment,1016 undergraduates
communication exper-iment, 200 under-graduates
18 undergraduates
200 undergraduates
86 undergraduates
literature review
literature review
education experi-ment, 180 under-graduates
Medium
publicspeech
TV,audio tape,live speech
TV
directmail
written essay
N/A
N/A
textbookillustrations
Finding
0
0
mixed
mixed
0
0
0
Comment
convincingness andattitude change, no ad-vantage over non-humor,topic state medicine
no persuasion effectfound immediately afterspeech nor 4 weeks later
more persuasive effectfor subjects initiallyopposed
humor Increased persua-siveness of a "soft sell"approach but not of a "hardsell" approach
no humor rated higherin persuasiveness thanmoderate humor, exten-sive humor rated leastpersuasive
humor treatments on measures of trustworthiness.In summary, the overall indication of the research,
both in advertising and non-advertising studies, indi-cates that it is unlikely that source credibility is con-sistently enhanced through the use of humor. Thisresult is consistent with the opinions stated by U.S.and U.K. advertising practitioners (Madden andWeinberger 1984; Weinberger and Spotts 1989). Andthese studies cast doubt on the tentative conclusiondrawn by Stemthal and Craig (1973) that h\imorenhances source credibility (see Table 4).
Humor and Liking
Where source credibility examines cognitive aspectssuch as trust and expertise, source-liking deals withnon-cognitive affect. Stemthal and Craig (1973) con-cluded that humor enhanced the liking of the source.In the years since their work, strong support hasbeen found for this conclusion in both advertising andnon-advertising research (see Table 5). Studies ofteacher effectiveness have demonstrated that teach-ers who incorporate humor into the classroom rate
significantly higher on character scales (Gruner 1967)and are seen as more likable than average (Bryant etal. 1980). Humor has been viewed as a key aspect inteacher effectiveness, and the use of humor by teach-ers was demonstrated to positively influence studentattitudes toward educational programs (Bryant andZillmann 1989). Humor has also been demonstratedto increase the liking of educational materials suchas textbooks (Bryant et al. 1981) and educationaltelevision (Zillmann et al. 1980).
The marketing literature gives similar strong sup-port for enhanced liking through the use of humor,which has been shown to increase both liking of thead (Belch and Belch 1984; Gelb and Pickett 1983;Duncan and Nelson 1985; Speck 1987) and liking ofthe brand (Gelb and Pickett 1983; Gelb and Zinkhan1986; Duncan and Nelson 1985). Overall, ten adver-tising studies and seven non-advertising studies reporta positive effect of humor on liking while only twoadvertising and three non-advertising studies reportneutral or mixed findings. No studies in either groupreport a negative impact of humor on liking. It there-fore seems appropriate to conclude that humor does
December 1992 45
Table 4The Impact of Humor on Source Credibility
Advertising StudiesAuthor(s) & Date
Sternthal & Craig (1973)
Madden (1982)
Sutherland & Middleton(1983)
Belch & Belch (1984)
Madden & Weinberger(1984)
Speck (1987)
Wu, Crocker & Rogers(1989)
Weinberger & Spotts(1989)
Non-Advertising StudiesAuthor(s) & Date
Type of Study & Subjects
iiterature review
iab experiment,326 undergraduate
iab experiment, 107undergraduates
iab experiment, 184undergraduates
survey of U.S. ad-vertising executives, 68research executives and 72creative executives
iab experiment, 182undergraduates
iab experiment, 360undergraduates
survey of advertisingexecutives, 132 U.S. agencies29 U.K agencies
Type of Study & Subjects
Medium
N/A
radio
TV
N/A
TV
N/A
Medium
Finding
+
0
0
+
+
0
mixed
Finding
Comment
humorous commercialsperceived as lesstrustworthy than non-humorous commercials
with luggage ad
with camera ad
for knowledgeableness
for trustworthiness -for certain types of humor
more positive view heidby U.K. executives
Comment
Gruner (1970)
Markiewicz (1972)
Markiewicz (1972)
Bryant, ComiskyCrane & Zillman (1980)
Bryant, Brown, Silberberg& Elliott (1981)
lab experiment, 144undergraduates
36 undergraduates
86 undergraduates
correlation, obser-vation, 49 male and21 female collegeinstructors
education experi-ment, 180under-oraduates
publicspeech
TV
written essays
classroominstruction
text-bookillustrations
mixed
+
0
+ & -
-
humor enhances a "duiispeech but did not en-hance an "interesting"speech
only slight positiveeffect for male in-structors - some nega-tive effects
46
AdvertisingAuthor(s) &
StudiesDate Type
The Impact
of Study & Subjects
ofTable 5
Humor on Liking
Medium
of Source
Finding
Journal of Advertising
Comment
Leavitt (1970)
Stemthal & Craig (1973)
Brooker (1981)
Gelb & Pickett (1983)
Lammers, Leibowitz, Sey-mour & Hennessey (1983)
Belch & Belch (1984)
Duncan & Nelson (1985)
Gelb & Zinkhan (1986)
Speck (1987)
Chattopadhyay & Basu(1989)
Wu, Crocker, & Rogers(1989)
Zhang & Zinkhan (1991)
data-based,target audience
literature review
lab experiment, 240 adults
mall survey, 20% re-sponse rate total of383 respondents fromtarget audiences
lab experiment, 64undergraduates
lab experiment,184 undergraduates
lab experiment, 157male undergraduates
lab experiment, 120employed adults andpart-time students
lab experiment, 182undergraduates
lab experiment, 80undergraduates
lab experiment, 360undergraduates
lab experiment, 216undergraduates
TV
N/A
direct mailad
radio
TV
radio
radio
TV
TV
TV
mixed
factor analysis, humor partof amusing factor
tentative conclusion
weak effect
subjects asked toevaluate either ahumorous or non-humorous ad
for males
four of five humor types in-creased likability signifi-cantly more than non-humor
liking increased by humor-ous ad when subjectpresented with prior positiveinformation
Non-Advertising StudiesAuthor(s) & Date Type of study & subjects Medium Finding Comment
Lull (1940)
Gruner (1967)
Gruner (1970)
communication experi- publicment, 1016 undergraduates speech
communication exper- publiciment, 128 male undergraduates speech
communication exper- publiciment, 144 undergraduates speech
mixed
humorous andnon-humorous speechrated as equally interesting
measured character ofspeaker
humor more effective inenhancing liking of dullspeech than interestingspeech .
continued..
December 1992 47
Author(s) & Date Type of Study & Subjects Medium Finding Comment
Kennedy (1972)
Markiewicz (1972)
Markiewicz (1972)
Gruner(1976)
Ziilmannetal. (1980)
Davies & Apter (1980)
communication exper-iment, 200 under-graduates
55 7th grade students
36 undergraduates
literature review
education experiment70 children aged5-7 years
education experiment285 children aged 8 to 11
Bryant, Comisky, Crane & correlational, ob-Zillman (1980) servatlon of 49 male
and 21 female collegeinstructors
Bryant, Brown, Silberberg education experiment& Elliott (1981) 180 undergraduates
TV, audiotape, livespeech
written essays +
TV +
N/A +
TV
slide-tape +
classroom +instruction
mixed
textbook +illustrations
measured 3 ethosdimensions, dyna-mism, qualification, andsafety humor increaseddynamism immediately afterspeech; 4 weeks later all 3measures superior tonon-humor
speaker image enhancedwith apt humor
fast paced humorenhanced program liking
for male instructors
for female instructors, onlyhostile humor enhancedappeal
indeed have a positive influence on liking.This strong liking response has significant implica-
tions. Recent research indicates that liking may be avery important variable in the effectiveness of an ad(Biel and Bridgwater 1990; Haley and Baldinger 1991).In Haley and Baldinger's (1991) comprehensive studyfor the Advertising Research Foundation, six copytesting methods were employed to study five matchedpairs of commercials with 400-500 respondents percell; thus, a total of nearly 15,000 interviews wereconducted for the study. This research shows thattwo liking measures are the strongest indicators of acommercial's sales success, out-performing all othermeasures. The overall reaction to the commercial, interms of liking, was demonstrated to predict which ofa paired set of commercials would be the sales winner87% of the time, with an index level indicating anassociation three times stronger than random chance.A related dichotomous liking measure had a success-ful prediction rate of 93%, albeit with a lower indexlevel. These recent liking findings provide strongsupport for the importance of this factor in the effec-tiveness of an ad. In concert with the Haley and
Baldinger finding, Biel and Bridgwater (1990) con-cluded that individuals "who liked a commercial 'alot' were twice as likely to be persuaded by it thanpeople who felt neutral toward the advertising" (p.38). Although in the Biel and Bridgwater (1990) workliking was not confined to entertainment value andincluded such factors as personal relevance, a findingby Haley and Baldinger (1991) is directly tied to hu-mor. Their study indicates that a positive response tothe statement, "This advertising is funny or clever,"predicts the success of an ad 53% of the time, whereasagreement with the statement, "This advertising isboring," predicts failure 73% of the time (Haley andBaldinger 1991).
Executional Factors
Humor Type
In previous sections we have made reference tosome executional factors regarding the nature of thehumor that may play a role in determining the effi-cacy of a given humor treatment. These executionalfactors can be subdivided into two groups, the first
48 Journal of Advertising
Figure 1Tactical Considerations for Humor in Advertising
BackgroundFactors
NATURE OF PRODUCT• Involvement• Thought/Emotionality• Familiarity
GOALIntended Impact• Attention• Comprehension• Persuasion• Source
Credibility• Source Liking• Other
MESSAGEHumor Type• Relatedness
to Product• Style of Humor
TARGET AUDIENCE FACTORSDemographics• Age• Gender• EducationPsychographicsPrior Brand AttitudeHumor PreferenceCulture
PLACEMENTMedium Context Repetition• Print • Humorous • Wearout• Broadcast • Non-
Humorous
RESULTING IMPACT ON:• Attention• Comprehension• Persuasion• Source Credibility• Source Liking• Other
December 1992 49
being the relationship between the humor treatmentand the product or message. In other words, is the"joke" in some way dependent on the situation orwould it be equally funny in some other context. Infurther refining the concept of relatedness. Speck(1991) states that there are three types of relatedness:1) intentional - the relationship of humor to messagetype and message processing, 2) semantic - the rela-tionship of humor to product-related themes, and 3)structural - the syntactical function of humor, referringto the integration of the humor and the product claims.While little work regarding the role that relatednessmay play in the effect of humor has been conducted,most of the studies in advertising employed humorthat has some degree of relatedness. Studies thathave directly compared related humor to unrelatedhumor (Kaplan and Pascoe 1977; Madden 1982;Weinberger and Campbell 1991) have generally foundrelated humor to be superior to unrelated humor.
The second executional factor of interest is humortype. Unfortunately, an all-encompassing, generallyaccepted definition of humor does not exist. However,several taxonomies have been proposed tooperationalize the construct of humor. Humor can becategorized on at least two different dimensions,"content" and/or "technique." A commonly used con-tent typology places all humor into one of three clas-sifications: aggressive, sexual, or nonsense (Goldsteinand McGhee 1972). Technique typologies have alsobeen employed; Kelly and Solomon (1975) definedhumorous ads as containing one of the following: 1) apun, 2) an understatement, 3) a joke, 4) somethingludicrous, 5) satire, 6) irony, or 7) humorous intent.
Alden, Hoyer and Lee (1993) have introduced an-other approach to the understanding of humor intothe advertising literature. This work focuses on theunderlying process that creates humor. Their litera-ture review, as well as their empirical data, suggeststhat much of what is seen as humorous is some formof incongruent contrast. This work further suggeststhat the prevalence of incongruity holds cross-cul-turally. In each of the four countries examined byAlden and his colleagues, the majority of humoroustelevision ads contained one or more incongruentcontrasts. This data indicates that 69% of humorousTV ads in the U.S. employ incongruity (Alden, Hoyerand Lee 1993).
A broader based method of categorizing humor isproposed by Speck (1991). He states that humor iscomposed of distinct basic processes: arousal-safety,incongruity-resolution, and hvtmorous disparagement.These processes may act alone or in combination to
form five humor types: HTl) comic wit (incongruity-resolution), HT2) sentimental humor (arousal-safety),HT3) satire (incongruity-resolution and humorousdisparagement, HT4) sentimental comedy (arousal-safety and incongruity-resolution, and HT5) full com-edy (arousal-safety, incongruity-resolution, and hu-morous disparagement).
Disappointingly, little work has directly comparedhumor types. However, one study that attempts to dothis (Speck 1987) indicates that significant differ-ences exist in hvimor effect between types. For ex-ample, in measuring the effect of humor on overallattention (an average of the five attention measureshe employs). Speck (1987) found effects ranging fromstrongly positive for full comedy to an essentially nulleffect for sentimental humor. This result is intuitivelyappealing in that one might expect full comedy withall three humor processes operating to draw the at-tention of the viewer. It should also be noted that allof the humor types outperformed non-humor on at-tention. Intuitively, appealing results are also foundin Speck's analysis of source liking which indicatesthat sentimental humor is associated with liking whilethe more aggressive satire is not (see Table 6).
The findings reported by Speck (1987) provide agood beginning as an exploration of humor type. Theyindicate that there is no one type of humor that has auniversally positive or negative impact. However, wemust caution against drawing any sweeping conclu-sions from this work. Speck's (1987) study analyzedonly five ads from each humor type. Lacking cor-roborating studies conceming hvimor types, we musttherefore regard these results as tentative. It is fartoo early to draw any general conclusions regardingthe appropriateness or inappropriateness of any giventype of humor in achieving a certain communicationsgoal.
Placement
The type of medium, the context in which an adappears, and the degree of repetition for humorousads are all topics that have been explored over thepast twenty years. In their survey. Madden andWeinberger (1984) found U.S. ad executives generallybelieved that radio and TV were the media best suitedto the use of hvimor, while print media were consid-ered not well suited to using humor. These surveyresults are corroborated to some degree in researchthat shows that in the U.S., 30.6% of radio ads areintended to be humorous (Weinberger and Campbell1991), while 24.4% of TV ads and just 9.9% of print
50 Journal of Advertising
Tabie 6Differential Effects of Humor Type on Communications Goals*
Overall"attention
Messagecomprehension
Descriptivecomprehension
Perceivedsource trust
Perceivedsource knowledge
Source liking
Comic Wit
.06
-.15
-.18
-.1
-.09
-.01
SentimentalHumor
-.02
-
-
.00
-.21
.31
Satire
.07
.23
.21
-.26
-.05
-.23
SentimentalComedy
.06
-.01
-.13
.27
.23
.23
Fuli Comedy
.27
-.11
.13
-.10
-.21
.15
No HumorControl
-.45
-.11
-.04
.11
.16
-.19
* adapted from Speck (1987).** a summed scale of five attention measures.
ads have humorous intent (Weinberger and Spotts1992). What is clear from this data is that the use ofhumor is consistent with the views of the ad executivessurveyed. What is unclear from any of the research iswhether humor in the broadcast media is more effec-tive than humor used in print.
At a more micro level, within media, the context inwhich a humorous ad appears may also affect theimpact of the ad. Studies have indicated that an in-teraction effect may exist between program environ-ment and commercials (Goldberg and Gom 1987;Kamins, Marks, and Skinner 1991; Mathur andChattopadhyay 1991). However, work conducted di-rectly in the humor area in general does not support astrong context interaction (Cantor and Venus 1980;Madden 1982; Markiewicz 1972). Some evidence of acontext effect is found in the work of Murphy and hiscolleagues (Murphy, Cunningham and Wilcox 1979).Their study finds a program interaction, but it affectsonly some measures, they find no context effect forunaided product recall. Therefore, while the generaladvertising literature indicates context interactions,the humor literature does not support this view.However, with so little evidence for or against con-text effects, no conclusive statement about these effectscan be made (see Table 7).
Finally, also at the micro level is the issue of howoften to run a humorous ad. Some evidence indicates
that hiimorous ads may wear out faster than non-humorous treatments over repeated exposvires (Gelband Zinkhan 1985). This finding has intuitive appealsince the surprise element often present in humor islikely to decay after the first exposure. Other re-searchers (Belch and Belch 1984) have, however, foundevidence that humorous ads decay at the same rateas non-humorous counterparts. More recently,Zinkhan and Gelb (1990) conclude, "not all humorouscommercials ( or comedy acts) 'wear out' with repeti-tion; some seem to get better, as anticipation of whatwill be presented evokes an anticipatory humorousresponse" (p.44O). They also posit that the social set-ting in which the humor is received may affect thehumor response. Humor is perceived as funnier whenreceived as a member of a group. This finding is alsosupported by Zhang and Zinkhan (1991), and thisinteraction with group members may help postponethe wear-out of humorous ads.
Audience Factors
The majority of practitioners believe that humor-ous ads are best suited to a target audience composedof better educated younger males (Madden andWeinberger 1984). The advertising literature gener-ally supports this belief. Several studies have indi-cated an interaction between gender and humor ef-
December 1992 51
Author(s) & Date
Media
Madden & Weinberger(1984)
Weinberger & Campbell(1991)
Weinberger & Spotts(1992)
Context
Murphy, Cunningham &Wilcox(1979)
Canton & Venus (1980)
Madden (1982)
Kamins, Marks, & Skinner(1991)
Repetition
Belch and Belch (1984)
Gelb & Zinkhan (1985)
Type of Study & Subjects
survey of 140 U.S.ad executives
data based, pre-tested ads
content analysis,U.S. and U.K.
lab experiment, 115undergraduates
lab experiment, 117undergraduates
lab experiment, 326undergraduates
lab experiment, 124undergraduates
lab experiment, 184undergraduates
iab experiment, 120adults
Table 7Placement of Ad
Medium
N/A
radio
TV, maga-zines
TV
radio
radio
TV
TV
radio
Finding
+
N/A
N/A
humorous adsrecalled in non-humor context,unaided productrecall same inall contexts
no significantcontext effects
no effect ofserious orhumorouscontext
"happy" adsbetter inhappy con-texts
no signifi-cant differ-ence inwearout be-tween humor-ous and non-humorous ads
rating ofhumorousnessof ad de-clined overmultiple ex-posures
Comment
broadcast best and printleast suited for humor
humor use highin radio
humor use highin TV, low in magazines
happy ads notnecessarily humorous,results of study may beaffected by health care pro-duct chosen formanipulation
Zilnkhan & Gelb (1990) conceptual N/A wearout delayed byanticipation of liked humorand viewing in group setting
52 Journal of Advertising
fectiveness (Gorham and Christophel 1990; Lammerset al. 1983; Madden and Weinberger 1982; Stewart-Hunter 1985; Whipple and Courtney 1980,1981), andrace and humor effectiveness (Madden andWeinberger 1982) (see Table 8).
The effect of gender may be partially explained byapparent differences in humor appreciation. In a re-view of the literature, Whipple and Covirtney (1981)conclude that men appear to enjoy aggressive andsexual humor more than women do, and women ap-pear to have a greater appreciation for nonsensicalhumor. However, they caution that the results arenot conclusive and that these preferences may bechanging as society changes. Further, the perspectiveof the creator of the humor may be an importantmoderator, as will be discussed.
Contrary to the preponderance of the findings inmarketing, research in education generally has notfound significant gender effects on humor response.Both in an extensive educational experiment discussedearlier (Ziv 1988), and in other experiments (Daviesand Apter 1980; Weaver, Zillmann and Bryant 1988;Zillmann et al. 1980), the positive effect of humor onlearning was not found to differ by gender.
The dichotomy of results of gender effects on humorraises some interesting issues. Humor is very closelytied into the culture, experiences, and points of ref-erence that are shared between the humor originatorand the humor receiver. For example, research hassuggested that the gender response to sexual humoris reversed when the creator of the humor is female(Gallivan 1991), and the characteristics of the butt ofthe joke may influence which audiences find the jokefunny (Gruner 1991). If this is indeed the case, thenmuch of the variation based on gender, and perhapsrace and age as well, may be explained by divergentperspectives of the creator of the humorous manipu-lation and the receiver of that manipulation. Thus,the "shared point of view" between the creator of ahumorous ad and the target of the ad is a potentiallyimportant intervening variable in humor effectiveness.This issue has been largely overlooked by researchers.
In addition to gender, race, and age, other audiencefactors may impact the effectiveness of humor andare worthy of consideration. Of particular interest ishow the crossing of national boundaries affects hvimorappreciation and effectiveness. Humor is a universalhuman process exhibited by people of all cultures andthroughout all of recorded history (Alden, Hoyer andLee 1993). However, the research that has examinedhumor in advertising cross-culturally indicates dif-ferential use of humor among countries, both in hu-
mor types employed and in absolute levels of humorused (e.g. Alden, Hoyer and Lee 1993; Weinbergerand Spotts 1989). Furthermore, empirical evidenceindicates that people of different cultural backgroundsrespond to humor differently. In an experiment thatcompared Israeli Jews of Eastern and Western de-scent, Weller and his colleagues found significant dif-ferences in the appreciation for absurd jokes betweenthe two groups (Weller, Amitsour, and Pazzi 1976).They posit that these differences are due to "habits ofthought and mental attitudes rooted in culturalbackgrounds" (p. 163). These findings imply that evenwhen language differences are removed, jokes maynot be easily "translatable" between cultures.
Other audience factors may also affect hvmior ap-preciation. For example, conservatism has been shownto be a predictor of response to humor; subjects ratedhigh on measures of conservatism have been demon-strated to judge incongruity-resolution humor to befunnier than their liberal counterparts (Hehl and Ruch1990; Ruch and Hehl 1986). Another audience factorof note includes audience and product interactionssuch as prior brand attitude. Chattopadhyay and Basu(1989) indicate that humor has greater positive effect,with regard to persuasion, for those audience memberswith a prior positive brand attitude. These and otheraudience factors should be kept in mind in the designof humorous ads and future humor research.
Product Factors
Another potentially important situational factor isthe nature of the product. Advertising executives sur-veyed believed that low involvement products such asconsumer non-durables are best suited for humorousad treatments (Madden and Weinberger 1984). Theusage of humor bears evidence of this belief. In ananalysis of over 1600 radio ads, Weinberger andCampbell (1991) found significant variation in humorapplication across different cells of the FCB productgrid. In the lowest use cell, high involvement-feelingproducts (fashion clothes, perfume, etc.), only 10.0%of all commercials were humorous in nature. This iscontrasted with the low involvement-feeling products(snack foods, beer, wine, etc.) cell where 39.6% of theads were humorous. More importantly, the impact ofhumor also appears to differ across the cells of theFCB matrix. Weinberger and Campbell (1991) foundthat related humor, when used with high involvementfeeling products (cell 2), resulted in significantly higherrecall scores than either unrelated humor or no hu-mor. Conversely, no humor appeared to be the most
December 1992 53
Table 8Audience Factors
Advertising StudiesAuthor(s) & Date Type of Study & Subjects Medium Finding Comment
Shama & Coughlin(1979)
Whipple & Courtney(1980)
(1981)
lab experiment 403undergraduates
lab experiment, 284graduates & under-graduates
literature review
radio & no gender effect,TV race effect found
print gender effect
gender effect
humor more ef-fective for white subjects
male graduatestudents ratead with femaleliterature review
butt of joke significantlyhigher than females. Hostileand aggressive humor moreeffective for males.
Madden (1982)
Madden & Weinberger(1982)
Lammers, Leibowitz,Seymour & Hennessey
Sutherland &Middleton (1983)
Madden & Weinberger(1984)
Weinberger & Spotts(1989)
Alden, Hoyer & Lee(1993)
Zhang & Zinkhan(1991)
Non-Advertising Studies
Smith etal. (1971)
Weller, Amitsour &Pazzi(1976)
lab experiment, 326undergraduates
data-based study of148 print ads.Starch
lab experiment, 64undergraduates
lab experiment, 107undergraduates
survey of U.S. ad-vertising execu-tives, 68 researchexecutives and 72 cre-ative executives
survey of advertis-ing executives, 132U.S. agencies, 29U.K. agencies and182 undergraduates
content analysis,cross-cultural
lab experiment, 216undergraduates
lab experiment, 215undergraduates
70 adults
radio no gender effect
print race & gender dif-ference
radio gender differenc:es
print no gender effectfound
N/A age, gender & educa-tion difference
N/A less age gender &education differenc-es in the opinion ofU.K. executives
TV high use of incon-gruity across countries
TV humor more effectivewhen presented togroups of subjects
written no gender differ-test ences found
oral laughter response tojokes absurd humor lower
for Israeli subjectsof Eastern origin
humor most ef-fective forwhite males
humor increasesmale liking, decreasedfemale liking
thought as moreeffective foryounger, well-educated males
continued...
54 Journal of Advertising
Author{s) & Date
Chapman & Crompton(1978)
Davles& Apter (1980)
Zillman & Bryant(1980)
Zillman etal. (1980)
Weaver, Zillman &Bryant (1988)
Ziv (1988)
Gorham & Christophel(1990)
Type of Study & Subjects
education experi-ment, children sub-jects ages 5 & 6yearstarget audiences
education experi-ment, 285 childrensubjects ages 8-11132 U.S. agencies29 U.K agencies
lab experiment, 90undergraduates
education experi-ment, 70 childrenages 5-7 years
education experi-ment, 86 children,10 & 14 years
education experi-ment, 161 undergrad-uates In first ex-periment male & fe-male, 132 undergrad-uates in replicationall female
correlational study,206 undergraduatesobserving 150 maleand 54 female col-lege instructors
Medium
slides
slide-tape
N/A
TV
TV
class-roominstruction
class-roominstruction
Finding Comment
humor moreeffective formales
no gender effect
no gender effectfound for "mirth"
no gender effect
no gender effect
no gender effect a semester longexperiment anda semester longreplication
male students morepositively affectedby humor
effective executional tactic for high involvementthinking products (cell 1). These results are largely inconcert with those of other researchers (Bauerly 1990;Scott, Klein and Bryant 1990; Stewart and Furse1986).
Furthermore, studies have shown humorous ads tobe more successful for existing products than for newproducts (McCollunVSpielman 1982; Stewart andFurse 1986). This factor raises an important meth-odological issue. While the use of fictional products inexperimental studies eliminates pre-existing productattitudes that may confound the results of experi-ments, a fictional product is a "new product" thatmay diminish the effectiveness of the humor treat-ment. This may mean that weak effects of advertis-ing humor found in some lab studies may arise fromtests using unfamiliar products.
Other Questions For Another Time
While we have provided what we believe to be thecurrent state of knowledge regarding humor in ad-vertising, it is clear that this knowledge is incompleteand additional research must be conducted before wecan clarify the remaining gray areas. Specifically, webelieve that the following areas require additionalresearch:
• Relatedness of the humor to the product ap-pears to be a strong predictor of the success ofa given ad. However, it has been studied insurprisingly few papers. Future work shouldconsider this factor. In particular. Speck's(1991) typology of relatedness could be usedas a basis to examine the issue.
December 1992 55
Table 9Nature of Product or Topic
Advertising StudiesAuthor(s) & Date
McCollum/Spielman (1982)
Madden & Weinberger(1984)
Weinberger & Spotts(1989)
Bauerly (1990)
Scott, Klein & Bryant(1990)
Weinberger & Campbell(1991)
Type o( Study & Subjects
data basedstudy of 500commercials, targetaudiences
survey of U.S. ad-vertising execu-tives, 68 researchexecutives, 72 crea-tive executives
content analysisstudy, 450 U.S. and247 U.S. commercials
mall intercept sur-vey, 226 respondents
field experiment,513 respondents
data-basedstudy of over 1600pre-tested ads
Medium
TV
N/A
TV
N/A
directmail
radio
Finding Comment
humor better forestablished prod-ucts and is bettersuited to certainproduct categories
consumer non-dur-ables best suitedto humor treatment
humor most commonlyused for low in-volvement products
soft drinks andsnack foods bestsuited to humor
humorous ad in-creased attendanceat social events,but not to businessevents
most common use forlow involvementproducts, most ef-fective for highinvolvement/feelingand low involve-ment/feeling situations
The context in which a humorous ad is placedhas been largely overlooked in the research.Mood established by adjacent programs oreven by ads within a pod might be examined.
Audience factors appear to be vitally importantfactors in the success or failure of a humorousad.. Researchers should pay particular atten-tion to the nature of the audience with regardto age, gender, education, culture and priorbrand attitude and should also investigatehow'these factors relate to the object of thehumor (i.e., the butt of the joke).
Humor types may affect humor response. Dif-ferences in humor types have largely beenignored by advertising researchers, but earlywork in this area by Speck (1987) indicatesthat different communications goals might
best be addressed using different types of hu-mor. However, since only one study has sys-tematically investigated humor types, this isan area in need of much future research.
• Researchers should consider message intensity,particularly its effect on persuasion. Somework indicates the intensity of the messageincorporating humor may have a significanteffect on the persuasiveness of the message(Bryant et al. 1981; Markiewitz 1972).
While calling for research in a number of areas wedo not wish to leave the impression that researchersto date have been remiss. We recognize the difficul-ties in conducting research in the hvmior area, and,therefore, we are not disappointed with the extantwork. However, we believe that much additional in-sight can be gained through further research.
56 Journal of Advertising
Discussion and Conclusion
Discussion
The preceding discussion makes it clear that muchwork remains to be done to fully understand the im-pact of humor in advertising. However, our synthesisof the current literature leads us to several conclu-sions. Before stating these conclusions, it is neces-sary to add a cautionary note. Humor research isplagued by many complexities, as mentionedthroughout the text. In schematic format, many ofthese factors are noted in Figure 1. Each of thesevariables can intervene in the relationship between agiven humorous ad and its outcome, thereby creatingdozens of contingent relationships with possibly di-vergent effects. The factors shov^m in Figure 1 thusshow that broad generalizations about the effect ofhumor in advertising are inappropriate.
It is also important to note that the complex rela-tionships described above are further complicated bymethodological problems that make them difficult tostudy. In many ways these complexities parallel thosefound in attempting to study other message charac-teristics such as music or warmth. In the study of anyof these message factors, the development of a controlad that is equivalent to the test ad in every way, savefor the tactic being tested, is a difficult task. Thefactor being studied must not be present in the controlad yet this ad must seem plausible as a stand-alonead, and must be equivalent in length to the test adwhile not adding anything that might confound re-sults. In the case of humor this becomes particularlyproblematic since we have argued that humor per-forms best when it is related to the product beingpromoted. To remove related humor removes an es-sential part of the message. The problem of findingadequate controls is a vexing problem indeed, and weapplaud the valiant efforts made in this regard butinclude this particular methodological problem in ourconclusions because not only is it a problem all humorresearchers face, but it is also a problem interpretersof this research must face. Each study included inthis review is affected to one degree or another bythis non-equivalence problem. While we do not con-sider this a fatal flaw, it is important to bear it inmind when reviewing humor literature. It is our belief,however, that this problem, as well as other problemsthat may seriously damage a given study, are some-what ameliorated when studies are combined.
Also problematic in the area of humor research isthe idiosyncratic response by members of the audi-
ence to the humor treatment and perhaps to the typeof humor employed. While other message character-istics like warmth also elicit different results in dif-ferent individuals, these results are likely to be mat-ters of degree rather than type. Humor, on the otherhand, is truly idiosyncratic. Some forms of humor,such as satire, sexual humor, and other forms of ag-gressive humor, may generate strong positive feelingsin some audience members while eliciting strongnegative feelings in others. In this respect, humor isperhaps a riskier tactic than many other approaches.Further complicating the study of humor is that humormay create warmth (Aaker, Stayman and Hagerty1986) and other emotional responses, in addition tothe laughter or smile that is usually associated with ahumor response. This generation of a warmth re-sponse to humor confounds the effects, making it un-clear if other humor treatments would have generatedsimilar effects or if alternatively other warmth treat-ments may have generated similar results. We reit-erate that perhaps the best way to resolve thesecomplexities in the study of humor is to examinehumor effects across numerous studies so that thestrengths of one study can balance the weaknesses ofanother. Therefore, conclusions drav^ni for a synthesissuch as this are somewhat insulated from the eccen-tricities of individual studies.
Conclusions
• Humor attracts attention. The vast majority ofstudies conducted in both advertising andeducation bear this out.
• Humor does not harm comprehension. Whilesome studies indicate that a harmful effectmay occur, it is more likely for humor to haveno effect. In fact, some evidence exists that itmay even aid comprehension. This more op-timistic view of humor is strongly supportedin the educational research and in the viewsof British advertising executives.
• Humor does not appear to offer an advantageover non-humor at increasing persuasion.Though some examples of increased persua-sion do exist, they tend to be qualified bygender, prior attitude and the nature ofproduct or the event promoted. Despite strongsupport by British agency executives for hu-mor and persuasion, the cvirrent conclusionfrom the overall literature concurs with theview that hximor does not offer significant
December 1992 57
advantage over non-humor when persuasionis the goal.
• Humor does not enhance source credibility. Itappears that in general humor has no differ-ential effect and a number of studies actuallysuggest that there are cases when himiorharms source credibility.
• Humor enhances liking. In fact, the link be-tween humor and liking is stronger than forany of the other factors. In light of an in-creased emphasis in advertising on affect, thisfinding should not be underestimated.
• Related humor is superior to unrelated humor.In studies that have directly compared thetwo forms, a differential advantage has beenshown for related humor. However, to date,there has not been sufficient research con-ducted to determine if specific forms of relat-edness have a differential advantage.
• Audience factors affect humor response. Thoughvariations do exist with regard to audiencepreference and audience reaction to humor,this is a shifting sand that needs close scru-tiny. What is funny to a certain gender, ethnic,or age group needs to be assessed in relationto the group's perspective and who is the buttof the humor. For example, in the past, re-searchers have concluded men enjoy sexualhumor and women do not. This finding wouldperhaps be reversed when examining sexualhumor written for a women's audience withmales as the butt of the joke. This suggeststhat perhaps more than other forms of adver-tising, humor needs pretesting.
• The nature of the product affects the appropri-ateness of a hvimor treatment. Though hu-mor is used with many types of products, itsuse is more successful with existing ratherthan new products. Humor also appears to bemore appropriate for low involvement prod-ucts and feeling-oriented products.
In conclusion, evidence from twenty years of re-search conducted since Stemthal and Craig's land-mark review has caused us to reassess the role ofhumor. Humor is not, and never has been, a magicwand that assures more successful advertising, how-ever success is defined. In spite of the wave of in-creasing numbers of humorous ads that may lead oneinto overstating the case for humor in advertising, itis important to understand that humor can be ap-propriate and effective in some situations and not inothers. This review attempted to remove some of the
uncertainty about the use of humor in advertising byidentifying the contingencies that define humor's ef-fectiveness. As is often the case, a great deal has beenlesimed but far more needs to be studied in futureresearch.
ReferencesAaker, David A., Douglas M. Stayman and Michael R. Hagerty
(1986), "Warmth in Advertising: Measurement, Impact, andSequence Effects," Journal of Consumer Research, 12 (March)365-381.
Alden, Dana L., Wayne D. Hoyer, Choi Lee (1993), "IdentifyingGlobal and Culture-Specific Dimensions of Humor in Adver-tising: A Multinational Analysis," Journal of Marketing(forthcoming).
Bauerly, Ronald J. (1990), "Humor in Advertising: Does the Prod-uct Class Matter," Proceedings of The Atlantic Marketing As-sociation, 9-13.
Belch, George E. and Michael A. Belch (1984), "An Investigation ofthe Effects of Repetition on Cognitive and Affective Reactionsto Humorous and Serious Television Commercitds," Advancesin Consumer Research, Vol. 11, Thomas C. Kinnear, ed., Provo,UT: Association for Consumer Research, 4-10.
Biel, Alexander and Carol A. Bridgwater (1990), "Attributes ofLikable Television Commercials," Journal of Advertising Re-search, 30 (JunQ/July) 38-44.
Brooker, George (1981), "A Comparison of the Persuasive Effects ofMild Humor and Mild Fear Appeals," Journal of Advertising,10 (4), 29-40.
Bryant, Jennings, Dan Brown, Alan R. Silverberg and Scott M.Elliott (1981), "Effects of Humorous Illustrations in CollegeTextbooks," Human Communication Research, 8 (1) 43-57.
, Paul W. Comisky and Dolf Zillmann (1979), "Teachers'Humor in The College Classroom," Communication Educa-tion, 28 (May), 110-118.
, Paul W. Comisky, Jon S. Crane and Dolf Zillmann(1980), "Relationship Between College Teachers' Use of Humorin the Classroom and Students' Evtduation of Their Teachers,"Journal of Educational Psychology, 72 (4), 511-519.
and Dolf Zillmann (1989), "Using Humor to PromoteLearning in the Classroom," Journal of Children in Contem-porary Society, 20 (1-2), 49-78.
Cantor, Joanne and Pat Venus (1980), "The Effect of Humor onRecall of a Radio Advertisement," Journal of Broadcasting, 24(1), 13-22.
Chapman, A.J. and P. Crompton (1978), "Humorous Presentationsof Material and Presentations of Humorous Material: A Reviewof the Humor and Memory Literature and Two ExperimentalStudies," in Practical Aspects of Memory, M.M. Gruneberg, ed..Academic Press, 84-92.
Chattopadhyay, Amitava and Kunal Basu (1989), "Prior BrandEvaluation as a Moderator of the Effects of Humor in Adver-tising," Journa/ of Marketing Research, 26 (4), 466-476.
Davies, Ann P. and Michael J. Apter (1980), "Humour and itsEffect on Learning in Children," in Children's Humour, PaulMcGhee and A. Chapman, eds, John Wiley & Sons, 237-253.
Ehincan, Calvin P. (1979), "Humor in Advertising: A BehavioralPerspective," Journal of the Academy of Marketing Science,7(4), 285-306.
, James E. Nelson, and Nancy T. Frontczak (1984), "TheEffects of Humor on Advertising Comprehension," In Advancesin Consumer Research, Vol. 11, Thomas C. Kinnear, editor,Provo, UT: Association for Consumer Research, 432-437.
58 Journal of Advertising
and James E. Nelson (1985), "Effects of Humor in aRadio Advertising Experiment," «7ouma/ of Advertising, 14 (2),33-40, 64.
Gallivan, Joanne (1991), "What is Funny to Whom, and Why?"paper presented at Ninth Intemational Conference on Hum.ourand Laughter, Brock University, St. Catharines, OntarioCanada.
Gelb, Betsy D. and Charles M. Pickett (1983), "Attitude-Toward-The-Ad: Links to Humor and To Advertising Effectiveness,"Journal of Advertising, 12 (2), 34-42.
and George M. Zinkhan (1985), "The Effect of Repetitionon Humor in a Radio Advertising Study," Journal of Adver-tising, 14(4), 13-20, 68.
and (1986), "Humor and Advertising Effec-tiveness after Repeated Exposures to a Radio Commercial,"Journal of Advertising, 15(2), 15-20, 34.
Goldberg, Marvin E. and Gerald J. Gom (1987), "Happy and SadTV Programs: How They Affect Reactions to Commercials,"Journal of Consumer Research, 14 (December), 387-403.
Goldstein, Jeffery and Paul McGhee, eds. (1972), The Psychology ofHumor, New York: Academic Press.
Gorham, Joan and Diane M. Christophel (1990), "The Relationshipof Teachers' Use of Humor in the Classroom to Immediacyand Student Learning," Communication Education, 39(January), 46-62.
Gruner, Charles R. (1967), "Effect of Humor on Speaker Ethos andAudience Information Gain," Journal of Communication, 17(3), 228-233.
(1970), "The Effect of Humor in Dull and InterestingInformative Speeches," Central States Speech Journal, 21 (3),160-166.
(1976), "Wit and Humor in Mass Communication," inHumor and Laughter: Theory, Research and Applications,Antony J. Chapman and Hugh C. Foot, eds.. New York: JohnWiley & Sons, 287-311.
(1991), "On the Impossibility of Having a Taxonomy ofHumor," paper presented at Ninth International Conference onHumour and Laughter, Brock University, St. Catharines,Ontario, Canada.
Haley, Russell I. and Allan L. Baldinger (1991), "The ARF CopyResearch Validity Project," Journal of Advertising Research,31 (April/May), 11-31.
Hehl, Franz-Josef and Willibald Ruch (1990), "Conservatism as aPredictor of Responses to Humor-III. The Prediction of Appre-ciation of Incongruity-Resolution Based Humour By ContentSaturated Attitude Scales in Five Seimples," Personality andIndividual Differences, 11 (5), 439-445.
Hopkins, Claude (1923), (1927), Scientific Advertising and My Lifein Advertising, reprinted by Advertising Publications Inc.,Chicago, 1966.
Kamins, Michael A., Lawrence J. Marks, and Deborah Skinner(1991), "Television Commercial Evaluation In The Context ofProgram Induced Mood: Gongruency Versus Consistency Ef-fects," Journal of Advertising, 20 (2), 1-14.
Kaplan, Robert M. and Gregory C. Pascoe (1977), "Humorous Lec-tures and Humorous Examples: Some Effects upon Compre-hension and Retention," Journal of Educational Psychology,69 (1), 61-65.
Kelly, J. Patrick and Paul J. Solomon (1975), "Humor in TelevisionAdvertising," t/ourraaZ of Advertising, 4(3), 31-35.
Kennedy, Allan James (1972), "An Experimental Study of the Ef-fect of Humorous Message Content Upon Ethos and Persua-siveness," Unpublished doctoral dissertation. University ofMichigan.
Lammers, H. Bruce, Laura Laebowitz, George Edward Seymour,and Judith E. Hennessey (1983), "Humor and Cognitive Re-sponses to Advertising Stimuli: A Trace Consolidation Ap-proach," Journal of Business Research, 11 (2), 173-185.
Leavitt, Glark (1970), "A Multidimensional Set of Rating ScalesFor Television Commercials," Journal of Applied Psychology,34 (5), 427-429.
Lull, P.E. (1940), "The Effectiveness of Humor in PersuasiveSpeech," Speech Monographs, 1, 26-40
Madden, Thomas J. (1982), "Humor in Advertising: Applications ofa Hierarchy of Effects Paradigm," Unpublished doctoral dis-sertation. University of Massachusetts - Amherst.
and Marc G. Weinberger (1982), "The Effects of Humoron Attention in Magazine Advertising," Journal of Advertis-ing, 11(3), 8-14.
and (1984), "Humor in Advertising: A Prac-titioner View," Journal of Advertising Research, 24 (4), 23-29.
Mathur, Mahima and Amitava Ghattopadhyay (1991), "The Im-pact of Mood Generated By Television Programs on Responsesto Advertising," Psychology & Marketing, 8(1), 59-77.
Markiewicz, Dorothy, (1972), "The Effects of Humor on Persuasion,"Unpublished doctoral dissertation, Ohio State University.
(1974), "Effects of Humor on Persuasion," Sociometry,37(3), 407-422.
McCoUunVSpielman and Go. Inc. (1982), "Focus on Funny," Topline,3 (3), 1-6.
Murphy, John H., Isabella G.M. Gunningham and Gary Wilcox(1979), "The Impact of Program Environment on Recall ofHumorous Television Commercials," Journal of AdvertisingResearch, 8 (2), n-21.
Nelson, James E. (1987), "Comment on 'Humor and AdvertisingEffectiveness after Repeated Exposvires to a Radio Commer-cial,'" Journal of Advertising, 16 (1), 63-68.
Ogilvy, David (1963), Confessions of an Advertising Man, London:Longman.
Powell, J.P. and L.W. Andresen (1985), "Humour and Teaching inHigher Education," Studies in Higher Education, 10 (1), 79-90.
Ruch, Willibald and Franz-Josef Hehl (1986), "Conservatism As APredictor of Responses to Humour-II. The Location of Sense ofHximour in a Comprehensive Attitude Space," Personality andIndividual Differences, 1 (6), 861-874.
Scott, Gliff, David M. Klein and Jennings Bryant (1990), "ConsumerResponse to Humor in Advertising: A Series of Field StudiesUsing Behavioral Observation," Journal of Consumer Re-search, 16 (March), 498-501.
Shama, Avraham and Maureen Goughlin (1979), "An Experimen-tal Study of the Effectiveness of Humor in Advertising," inEducators' Conference Proceedings, Neil Beckwith, ed., Ghi-cago: American Marketing Association, 249-252.
Smith, Ronald E., James G. Ascough, Ronald F. Ettinger and DonA. Nelson (1971), "Humor, Anxiety, and Task Performance,"Journal of Personality and Social Psychology, 19 (2), 243-246.
Speck, Paul Surgi, (1987) "On Humor and Humor in Advertising,"Unpublished doctoral dissertation, Texas Tech University.
(1991), "The Humorous Message Taxonomy: AFramework for the Study of Humorous Ads," in Current Is-sues & Research and Advertising, James H. Leigh and GlaudeR. Martin, Jr., eds.. The Division of Research Michigan Busi-ness School, The University of Michigan, 1-44.
Stemthal, Brian and Samuel Graig (1973), "Humor in Advertis-ing," Journal of Marketing, 37(4), 12-18.
Stewart, David M. and David H. Furse (1986), Effective TelevisionAdvertising, Lexington, MA: D.C. Heath and Company, Chi-cago.
December 1992 59
Stewart-Hunter, David (1985), "Humour in Television Advertis-ing: The Search for the Golden Rule," ADMAP, May, 268-279.
Sutherland, John C. and Lisa A. Middleton (1983), "The Effect ofHumor on Advertising Credibility and Recall of the AdvertisingMessage," In Proceedings of the 1983 Convention of TheAmerican Academy of Advertising, D.W. Jugenheimer, editor,Lawrence, KS: William Allen White School of Journalism andMass Communications, University of Kansas, 17-21.
Vance, Charles M. (1987), "A Comparative Study On The Use OfHumor In The Design Of Instruction," Instructional Science,16, 79-100.
Weaver, James, Dolf Zillmann and Jennings Bryant (1988), "Effectsof Humorous Distortions on Children's Letiming From Edu-cational Television: Further Evidence," Communication Edu-cation, 37 (July), 181-187.
Weinberger, Marc G. and Leland Campbell (1991), "The Use andImpact of Humor in Radio Advertising," Journal of Advertis-ing Research, 31 (December/January), 44-52.
and Harlan E. Spotts (1989), "Humor in U.S. VersusU.K. TV Advertising," Journal of Advertising, 18 (2), 39-44.
and (1992), "Differences in British and
and . (1980), 'Male and Female Differences
American Television and Magazine Advertising: Myth or Re-ality," (forthcoming) in Proceedings of the Association forConsumer Research: European Summer Conference, Gary J.Bamossy and W. Fred van Raaij, editors.
Weller, Leonard, Ella Amitsour, and Ruth Pazzi (1976), "Reactionsto Absurd Humor by Jews of Eastern and Western Descent,"Journal of Social Psychology, 98 (April), 159-163.
Whipple, Thomas W. and Alice E. Courtney (1981), "How Men andWomen Judge Hxmior, Advertising Guidelines for Action andResearch," In Current Issues and Research in Advertising, J.H.Leigh and C.R. Martin, Jr., editors, Ann Arbor, MI: Divisionof Research, Graduate School of Business Administration,The University of Michigan, 43-56.
in Response to Nonsensical Humor in Advertising," in Pro-ceedings of the American Academy of Advertising, J.E. Haefner,ed., 71-74.
Wu, Bob T.W., Kenneth E. Crocker and Martha Rogers (1989),"Humor and Comparatives in Ads For High and Low Involve-ment Products," Journalism Quarterly, 66 (Autumn), 653-661,780.
Zhang, Yong and George M. Zinkhan (1991), "Humor in TelevisionAdvertising," in Advances in Consumer Research, Vol. 18,Rebecca H. Holman and Michael R. Solomon, eds., Provo, UT:Association for Consumer Research, 813-818.
Zillmann, Dolf find Jennings Bryant (1980), "Misattribution Theoryof Tendentious Humor," Journal of Experimental and SocialPsychology, 16, 146-160.
, Brien R. Williams, Jennings Bryant, Kathleen R.Boynton and Michelle A. Wolf (1980), "Acquisition of Informa-tion From Educational Television as a Function of DifferentlyPaced Humorous Inserts," Journal of Educational Psychol-ogy, 72 (2), 170-180.
Zinkhan, George M. and Betsy D. Gelb (1987), "Humor and Adver-tising Effectiveness Reexamined," Journal of Advertising, 16(1), 66-68.
and (1990), "Repetition, Social Settings,Perceived Humor and Wearout," in Advances in ConsumerResearch, Vol. 17, Marvin E. Goldberg, Gerald Gom and Ri-chard W. Pollay, eds., Provo, UT: Association for ConsumerResearch, 438-441.
Ziv, Avner (1988), "Teaching and Learning with Humor: Experimentand Replication," Journal of Experimental Education, 57 (1),5-15.