A Meta-Analysis of Humor in Advertising
by Martin Eisend
The final publication is available at Springer via J. of the Acad. Mark. Sci., DOI 10.1007/s11747-008-
This meta-analysis combines 369 correlations on the effects of humor in advertising and thus
quantifies, updates, and expands previous literature reviews on the effects of humor in
advertising. In line with previous reviews, the meta-analytic correlations demonstrate that humor
in advertising significantly enhances AAD, attention, and positive affect. Contrary to the
assumptions of previous reviews, there is no evidence that humor impacts positive or negative
cognitions, and liking of the advertiser. The meta-analytic findings clarify some ambiguous prior
conclusions: humor significantly reduces source credibility, enhances positive affect, ABR and
purchase intention. The decline from lower order to higher order communication effects is
particularly strong, with the effect size of the impact of humor on AAD being twice as large as the
effect size for ABR. This impact of humor in advertising has been rather stable over the past
decades. A moderator analysis reveals, however, that the findings of academic humor research
are somewhat biased. As for the underlying theory, the positive and linear relationship between
the funniness of the ad and brand attitudes supports an affective mechanism underlying the
impact of humor in advertising.
The use of humor has become common practice in advertising. Approximately one out of five
television ads contain humorous appeals (Beard 2005). TV shows on humorous advertisements
and awards for such advertisements indicate that humor in advertising has even become an
important part of everyday life. The use of humor as an executional tactic is of particular interest
to marketers, since marketing performance depends on effective and successful advertising.
Quite an effort has been made to investigate the impact of humor in advertising and has led to
several literature reviews (Duncan 1979; Madden and Weinberger 1984; Speck 1987; Sternthal
and Craig 1973; Weinberger and Gulas 1992). These reviews draw some clear-cut conclusions
on the impact of humor for some of the outcome variables, but have also come up with a set of
mixed results. Weinberger and Gulas (1992) suggest in their review that broad generalizations
about the persuasive effects of humor may be inappropriate. Chattopadhyay and Basu (1990a)
recommend asking when humor in advertising is effective, rather than if humor is effective. Both
issues, however, the search for generalizable results (“if humor is effective”) and for moderators
that contribute to the variability of results (“when humor is effective”) are specific tasks to be
addressed by the application of a meta-analysis.
The purpose of this study is to provide an integrative meta-analysis of research on humor
effects in advertising. The meta-analysis quantifies, updates, and expands previous literature
reviews, the most recent one having been published 15 years ago. The meta-analytic findings
substantiate (or disprove) the conclusions of those earlier literature reviews. Besides generating
such empirical generalizations, meta-analytic work tries to resolve conflicts in the literature and
to identify gaps in previous research, and thus can help to ensure that the next wave of research
is guided in the most illuminating direction (Farley, Lehmann, and Mann 1998). Hence, the
present meta-analysis further shows whether generalizations on the impact of humor are
appropriate and if not, how one can explain inconsistent findings by applying substantial and
methodological moderator variables. Finally, for the underlying mechanisms of the effect of
humor in advertising, the study investigates the crucial relation between humor-evoked affective
responses and marketing-related measures by showing if and how the funniness of the ad
supports or harms the impact on brand attitudes.
Previous Reviews on the Effects of Humor in Advertising
The widespread use of humor in advertising has led to several literature reviews. Table 1
provides an overview of previous review results. In addition to the reviews by academics, the
table also shows the findings of a study that surveyed senior advertising practitioners to elicit
their views concerning the conclusions of previous academic reviews and the objectives best
achieved by humor.
Insert table 1 about here
Some of the conclusions in the reviews are consistent: Humor in advertising creates attention
and awareness, enhances source liking, attitude towards the ad (AAD), positive cognitions, and
reduces negative cognitions. Some other conclusions are more ambiguous. It is not clear if
humor in advertising enhances comprehension or reduces it. It is not clear how humor impacts
recall or recognition. The conclusions reached for the effect of humor on source credibility are
mixed as well. The effects on brand attitudes (ABR), purchase intention and behavior remain
unclear as well.
Whether humor has an impact on these outcome variables can be studied by calculating a
grand mean effect size that integrates the results of previous studies and by applying significance
tests to the grand mean effect size. The replication nature of such a meta-analysis brings about
the question whether humor effects have changed over the years. There are several reasons that
may lead to changes of the effect sizes. One reason is a substantial change in the influence of
humor in advertising. For this, some authors refer to the transition from the modern to the
postmodern period (which may have started in the 1960s and has been reaching different
domains of society at different paces) in order to describe how consumers have changed and the
way they deal with communication from marketers (Elliot, Eccles, and Hodgson 1993; van Raaij
1993): Consumers have become more experienced and at the same time more skeptical about
influences from marketers. As this may primarily affect the impact of obtrusive marketer
persuasion techniques, it is not clear whether consumers react differently to humorous appeals.
Another reason relates to method changes and improvements; data collection methods and
analytical methods have been improved over the years, leading to stronger effect sizes (Kayande
and Bhargava 1994).
If the set of effect sizes is heterogeneous, the grand mean must be considered an average
rather than a common value, and the variability of the findings can be due to methodological and
conceptual differences of past humor research studies. These differences can be explained by
considering moderating variables. We used the following criteria to select from the potential
variables that moderate the relationship between humor and outcome variables. The moderator
analysis is limited to those subsets of relationships that provide a sufficient number of effect
sizes for the purpose of testing. The moderator analysis is further limited to those moderator
variables where previous literature provides some guidance as to how they may moderate the
effects of humor and/or that help to resolve ambiguous findings.
A sufficient number of effect sizes in this meta-analysis (more than thirty) is available for the
effects of humor on AAD, ABR, and purchase intention, because the primary focus of most
scholarly advertising studies has been on the impact on attitudes as a main predictor of behavior.
Since ABR and purchase intention are highly correlated with each other, moderating variables
should apply to both outcome variables in a similar way (Eisend 2006). Therefore the following
discussion focuses on the impact of humor on AAD and ABR. As for the influence of humor on
attitudes, the effects on AAD versus ABR can differ, and humorous ads that are most liked are not
necessarily most effective in terms of influencing brand attitudes (Woltman Elpers, Mukherjee,
and Hoyer 2004).
Moderating Variables Affecting the Impact of Humor in Advertising on Attitudes
Humor and AAD
Whether humor enhances AAD depends on the characteristics of the humorous stimulus, how
the stimulus is presented, and what kind of recipients are addressed. Previous studies differ in at
least one important instance of each of the factors that can be assessed by means of a meta-
As for demographics, the majority of practitioners believe that humor seems to work best for
younger and well-educated consumers, particularly for males (Madden and Weinberger 1984).
This assumption is in line with Suls “incongruity-resolution theory” of humor comprehension
and appreciation (Suls 1972). Suls proposed a two-stage model of humor: detection and
resolution of incongruity. Humor is based on incongruity wherein a prediction is not confirmed
in the final part of the story. To comprehend humor, it is necessary to revisit the story and to
transform an incongruous situation into a funny, congruous one. Comprehension of humor
requires cognitive abilities, which vary over stages of cognitive development and therefore
depend on age as well as on education. That is, age is negatively related to humor comprehension
whereas education is positively related to humor comprehension (Mak and Carpenter 2007;
McGhee 1986). Comprehension precedes humor appreciation, which in turn influences liking of
the advertising (e.g., Cline and Kellaris 2007; Woltman Elpers, Mukherjee, and Hoyer 2004).
The majority of previous studies of humor in advertising have been based on student samples,
while a smaller group of studies have worked with representative samples or convenience
samples of consumers. Student samples have consisted of younger and better educated
individuals than the consumer samples. Therefore, higher appreciation and stronger effects of
humor on AAD are expected when the sample consists of students.
Previous studies differ according to some methodological factors that in turn are related to the
above-mentioned audience and stimulus characteristics. Some authors have pointed out that
humorous stimuli in studies by academics, which are mostly performed in controlled laboratory
settings, are only mildly amusing, and the effects may therefore differ from the effects of real
world advertisements (Speck 1991; Woltman Elpers, Mukherjee, and Hoyer 2004). Furthermore,
humor is harder to execute in print advertisements due to fewer tools in the executional arsenal
(Gulas and Weinberger 2006). Humor in print media is confined to addressing the whole sensory
spectrum of individuals and is therefore inferior to humor in broadcast media in terms of
funniness. Gulas and Weinberger (2006) further suggest that the impact of humor in print media
may be neutralized through more vigilant, intense, and selective processing. It is therefore not
surprising that TV ads tend to be more humor-dominated than print ads (Speck 1991). Hence,
studies that work with humor in print media should provide weaker effects of humor on AAD than
Humor and ABR
Product categories interact with executional factors (such as humor) to affect advertising
impact, and so the effects of humor on ABR depend on factors related to the particular product. In
their book on humor in advertising, Gulas and Weinberger (2006) notice that only several studies
have explicitly examined the impact of humor on outcome variables for different products, and
that there is little to guide effective usage of different products.
A conceptual scheme to explain the effectiveness of humor when used with different products
can be found in product typologies that have been developed from the rationale of the
elaboration likelihood model (ELM) (Petty and Wegener 1999). Several product typologies have
been developed that attempt to integrate the idea of the ELM, which makes a distinction between
low and high motivation/ability situations when processing an (advertising) message. Each
framework provides a matrix of four fields with an involvement/perceived risk dimension and a
functionality dimension (Rossiter, Percy, and Donovan 1991; Vaughn 1980, 1986; Weinberger,
Campbell, and Brody 1994). In those frameworks, involvement, i.e., whether advertising
information is worth processing or not, is mostly understood as the outcome of perceived risk.
Although the operational definition of involvement and perceived risk varies slightly, the
classification of products in the matrix fields is made rather consistently. The classification of
products in the functionality dimension distinguishes between products of functional value
(“think”, “informational”) and hedonic value (“feel”, “emotional”, “transformational”). Gulas
and Weinberger (2006) summarize those frameworks within a four product color matrix
distinguishing white, red, blue, and yellow goods. Following Rossiter, Percy, and Donovan
(1991) and Gulas and Weinberger (2006), different strategies seem appropriate for each color
and its effect on ABR. Based on this typology, the appropriateness of humor in advertising for
each color can be derived by considering another crucial factor: the relatedness between product
and humor, that is, whether humor is thematically related to the advertiser’s message about the
product or not.
- White goods are high involvement/high risk and functional products that are risky enough to
be worth processing information in a more detailed way. Advertisements for such products
should consider benefit claims to be convincing and the target audience must accept the ad’s
main points but does not have to like the ad, although ad liking does not harm advertising
impact. Humor can serve as an issue relevant argument, particularly when consumers are
engaged in detailed information processing (Zhang and Zinkhan 2006). Hence, issue-relevant
humor provides benefit claims and can help to sell the product, whereas unrelated humor
may help consumers to like the ad but may not further improve the impact on ABR.
- Red goods are high involvement/high risk and hedonic products where the audience
processes information in a more detailed way as well. Advertisements should provide
emotional authenticity, and consumers should like the ad and must identify with the product
portrayed in the ad; information may be provided as well. Humor basically supports effects
on ABR and does not need to be product-related, as humor contributes to one’s liking of the
ad that can transfer to the advertised brand.
- Blue goods are low involvement/low risk and functional products that do not require detailed
information; trial experience is sufficient. As for advertisements, a simple problem-solution
format focusing on the central benefits of the product is most appropriate. It is not necessary
for consumers to like the ad (although ad liking does no harm to advertising effects); both
related and unrelated humor may bear the risk of distracting the consumer from a successful
information transfer of the central benefits of the product. Hence, humor may be effective,
but less so than for other kinds of products in the matrix.
- Yellow goods are low involvement/low risk and hedonic products. Brand attitude strategies
should focus on an emotional appeal that is unique to the brand, and the target audience must
like the ad. As for red goods, humor supports effects on ABR and does not need to be product-
The suggested relationships for the moderator variables are presented in table 2.
Insert table 2 about here
The AAD-ABR Relationship in Humor Studies
Gulas and Weinberger (2006) have suggested that an immediate effect of humor is best
described by mirth, a generic affective response that covers a variety of responses such as
happiness, fun, or pleasure. This affective response impacts higher order outcomes such as
feelings, thoughts, attitudes, or actions. The immediate response to humor varies in intensity
depending on the humorous stimulus. Intensity may impact higher order outcomes in different
A widely accepted explanation for the effect of humor on attitudes is based on the idea of
affective mechanisms, such as evoking a positive affect that is transferred to the brand
(MacKenzie and Lutz 1989). Such affect generalization can be explained by classical
conditioning: positive affective reactions towards the humorous stimulus represent the
unconditioned reaction that is generalized onto the conditioned stimulus: the advertised brand
(Prilluk and Till 2004). Very funny ads seem to be liked best (Woltman Elpers, Mukherjee, and
Hoyer 2004), and a positive impact on ABR should increase with the level of humor intensity in
terms of perceived humor.
Some authors have mentioned that ads that score high on perceived humor and, by this,
impact AAD are not necessarily effective in terms of impacting ABR and purchase behavior
(Weinberger and Campbell 1991; Woltman Elpers, Mukherjee, and Hoyer 2004). One reason for
the missing link between both variables is based on the insight that humor functions as a
distractor from those parts of the message that provide brand benefits. By referring to the widely
accepted insight that humor facilitates attention, Zillmann et al. (1980) have argued that
respondents pay close attention to the humorous part of the message, since humor induces
pleasant reactions and thus functions as incentive to pay attention. But because respondents are
preoccupied with the humor, they are less attentive to other parts of the message. Hence, the
funnier the ad, the higher the incentive value of humor, and the more the ad may distract from
brand-related parts of the message; thus, humor intensity shows no relationship with ABR.
Bryant et al. (1981) provide an explanation for a possible negative persuasive effect of humor
intensity through a source-mediation rationale. The use of humor contributes to decreased
credibility of a source, which reduces persuasion effects. The authors support this effect with the
context of humorous illustrations in college textbooks. However, such textbooks are expected to
provide serious material, whereas in advertising, a source does not necessarily suffer from
decreased credibility, provided the superior use of humor is valued by the recipient. If the source
is witty and has excellent command of her or his material by using humor, the effect may be
positive. A source that is perceived as using humor because of a lack of ability to make her or his
point seriously impairs persuasion. Humor intensity may play a moderating role in advertising,
as very strong humor may be seen as weakness of the marketer to make a serious argument for
the brand. This is also indicated by the Speck (1987) study; the author distinguished between five
types of humor and found out that the ads of the type with the highest level of perceived humor
performed the worst on source trust.
A curvi-linear relationship between humor intensity and the effects on ABR is due to the idea
that humor plays a cathartic role, evoking arousal that results in pleasure when released (Gulas
and Weinberger 2006). Humor causes such arousal through novelty, complexity, and incongruity
(Berlyne 1972). According to arousal theory, greater arousal results in greater pleasure when
released up to a point (optimal level) and therefore an inverted-U relationship between humor
appreciation/intensity and arousal can be expected. An increasing level of arousal evoked by
humor intensity may reach an optimal affect transfer; beyond this point, too much arousal may
become less pleasant, leading to transfer of negative affect.
Whether such high levels of arousal can be reached in advertising remains an open question.
Indeed, the inverted-U shape is not universally held, and research has provided evidence for a
linear relationship between humor appreciation and arousal (Godkewitsch 1976; Gulas and
Weinberger 2006). A curvi-linear effect has been supported in an education research study by
Bryant et al. (1981). The authors found that the use of low levels of humor has essentially the
same level of persuasion as no humor, while extensive use of humor was detrimental to
persuasion. As for advertising research, only Krishnan and Chakravarti (2003) investigated
moderate and high levels of humor with respect to claim recall. Gulas and Weinberger (2006)
notice that we still do not know what the effects of different humor levels on outcome variables
other than recall are; they point out that manipulation checks of experimental studies in humor
research show that the level of humor is usually not very high. Based on the ambiguous results
for the relationship between arousal and pleasure, they infer that the matter of arousal-evoking
humor effects still remains unresolved.
Effect sizes within a meta-analysis typically provide a wide range of data points. Hence,
perceived humor measures can be used to investigate the relationship between perceived humor
and ABR in a more profound way than has been done in previous studies. The relationship may be
negative (due to reduced credibility), positive (due to an affect transfer), curvi-linear (according
to an optimal-level approach underlying an affect transfer), or there may be no relationship at all
(due to distracting the consumers from brand benefits).
To identify relevant studies for the meta-analysis, a computerized bibliographic keyword
search using Business Source Elite, ABI/Inform (for business publications), PsycINFO and
PSYNDEX (for psychology literature), and the Social Science Citation Index was conducted,
followed by an internet search using Google Scholar. Once a study was identified, references
were examined in a search for further studies. The approach is consistent with recommendations
made by several authors (e.g., Hunter and Schmidt 2004; Rosenthal 1994), and closely follows
the steps taken in earlier meta-analyses published in the marketing literature. The literature
search covered the period from 1960 up to late 2006. Only studies published in English that
investigated the impact of humor in advertising on dependent variables as described above were
considered. These studies had to provide empirical results on the effect of humorous compared to
non-humorous advertising. That is, studies dealing with humor in advertising that did not provide
such comparisons were not considered (e.g., Aaker, Day, and Hagerty 1986; Unger 1995). In the
case where two studies were based on the same data (e.g., an unpublished dissertation and a
journal publication), the study that provided more data detail was used (e.g., the dissertation by
Lee (1997) was included but not the journal publication by Lee and Mason (1999)).
The search resulted in 54 manuscripts. If necessary, authors were contacted in order to receive
further data for those studies with insufficient information to calculate effect sizes. Eventually,
38 manuscripts covering 43 independent studies could be used for the meta-analysis (these
studies are indicated by asterisks in the reference list). With the exception of three studies, they
were all performed in the US. Sixteen manuscripts (of the 54 manuscripts that were found by the
literature review) were excluded due to a lack of statistical information for calculating an effect
size. For ten studies, standard deviations and/or cell sizes were not available (De Pelsmacker and
Geuens 1996; Flaherty, Weinberger, and Gulas 2004; Furnham, Gunter, and Walsh 1998;
Furnham and Mori 2003; Geuens and De Pelsmacker 2002; Madden 1982; Smith 1993; Stanton
and Burke 1998; Zhang 1992; Zhang and Zinkhan 1991); four studies were excluded due to a
(partial) lack of means as well as a (partial) lack of standard deviations (Lammers 1991;
Lammers et al. 1983; Madden and Dillon 1982; Nelson, Duncan, and Frontczak 1985); two
studies provided data from crosstabs of more than four cells that could not be transformed into an
appropriate effect size (Murphy, Cunningham, and Wilcox 1979; Weinberger et al. 1995). The
percentage of exclusions (30 percent) is not uncommon in meta-analyses and corresponds to
figures given in other meta-analyses in the marketing and consumer behavior literature (e.g.,
Brown and Stayman 1992; Szymanski, Troy, and Bharadwaj 1995; Tellis 1988).
The effect size metric selected for the analysis is the correlation coefficient; higher values of
the coefficient indicate a stronger effect of humor on advertising outcome variables compared to
non-humorous messages. Since most papers reported multiple measures of humor effects, the
analysis includes multiple correlations from single studies for particular relationships. Altogether
369 correlations were available for the purpose of the meta-analysis. The ratio of correlations to
the number of studies is not uncommon in meta-analyses when focusing on various dependent
variables (e.g., Churchill et al. 1985; Sultan, Farley, and Lehmann 1990; Szymanski, Bharadwaj,
and Varadarajan 1993).
Integration of Correlations and Moderator Analysis
The meta-analytic procedures were performed using ZumaStat 4.0 and taking a random-
effects perspective (Shadish and Haddock 1994). The integration of the correlations uses weights
for sample size and multiple dependent measures, and considers attenuation corrected
correlations. Sample size weights were applied in order to consider varying sample sizes of the
studies. Two procedures for attenuation correction were applied as suggested by Hunter and
Schmidt (1990, pp. 118-125): (1) effect sizes based on variables that were artificially
dichotomized were corrected, and (2) measurement errors were corrected by considering
reliability coefficients of the dependent and independent variables. A conservative .8 reliability
estimate was applied to objective measures as suggested in the literature (Bommer et al. 1995;
Dalton et al. 2003; Hunter and Schmidt 2004). In order to consider a weight for multiple
measures per study, each sample size was weighted by the ratio 1 to the number of effect sizes
per study measuring the same dependent variable. Using the simple sample size for studies with
multiple dependent measures avoids underestimation of sampling error compared to using the
sum of samples (Hunter and Schmidt 2004).
Mean correlations are significant when the confidence interval does not include zero.
Additionally, mean correlations were tested for significance by z-statistics. In case of
significance, a fail-safe N is calculated which shows how many non-significant results must be
added in order to prove the significance of the integrated effect size as a random error (Rosenthal
The moderator analysis is based on a sample-size weighted regression analysis applying the
predictor variables described above (cf., Hedges 1994). Two coders, a female and a male
graduate student who were not aware of the research questions of the study, were coding the
predictor variables according to the instructions in a coding sheet for each correlation. Based on
information given in the studies, the following variables were coded: fictitious versus real
advertising, broadcast versus print media, student versus non-student samples, white versus red
versus blue versus yellow goods, and product-related versus product-unrelated humor. The
coding scheme for product types was derived from the definition and the broad range of product
examples provided in previous studies (e.g., Weinberger and Campbell 1991). Given the rather
manifest nature of most of the moderator variables, Cohen’s Kappa indicated “excellent” results
(higher than .9) for the dummy codings. The few differences were resolved through discussion.
Humor level was assessed by calculating the effect sizes of the available manipulation check
measures of perceived humor. The publication year of the study was used as an indicator of the
year of the humor effect. Furthermore, the number of items to measure the dependent variable
was used as an indicator of measurement refinement.
Table 3 shows the results of the meta-analytic correlations when applying a random-effects
model using the Q-method for variance estimation. The results differentiate between ad- and
brand related recall and recognition effects. Humor significantly enhances AAD, ABR, positive
affective reactions, attention, and purchase intention. Humor reduces negative affective reactions
and credibility. The results show considerable variance, with the strongest effects of humor on
attention (r = .416) and AAD (r = .374). Both effects can be considered as large according to the
classification by Cohen (1977). The high fail-safe N lends broad credence to the significance of
the results based on larger samples of correlations, as a remarkable number of additional studies
with non-significant results would be necessary in order to prove the significance of the
integrated effect size to be a random error.
Insert table 3 about here
In order to investigate whether humor effects have changed over the year, correlations were
calculated between the publication year of the study and the effect size for AAD, ABR, and
purchase intention. Only those outcome variables were used, because they were based on a set of
studies that were published in at least ten different years. The bivariate correlations are not
significant (AAD: r = -.131, p = .227; ABR: r = .112, p = .445, purchase intention: r = .180, p =
.232). To control for possible confounds, partial correlations were calculated controlling for
perceived humor (as measure of humor intensity) as well as for the number of measurement
items as an indicator for measurement refinements. The partial correlations controlling for
perceived humor are not significant (AAD: r = -.237, p = .058; ABR: r = -.082, p = .641, purchase
intention: r = -.149, p = .439). When controlling for both perceived humor and the number of
measurement items, the partial correlations for purchase intention could not be calculated,
because it is almost exclusively measured with single-item scales. The correlations for AAD and
ABR remain insignificant (AAD: r = -.098, p = .449; ABR: r = -.081, p = .647).
The suggested moderator variables are applied to the subsets of correlations between humor
and AAD and humor and ABR, both of which reveal high heterogeneity that can be reduced by
moderator variables. For each regression model, a number of correlations had to be dropped due
to missing values of the predictor variables, a common problem in meta-analysis (Stock 1994):
five correlations were dropped for the moderator model dealing with the correlations between
humor and AAD, and eight correlations were dropped for the moderator model dealing with the
correlations between humor and ABR.
Table 4 shows the results of the moderator variables applied to the correlations between
humor and AAD. The effects are stronger for the student sample, for real ads, and for broadcast
media. Hence, the predictions of the method factors that explain the variance of the effects of
humor in advertising are all met. The results are in line with the effects of perceived humor.
Perceived humor is stronger for real ads than for fictitious ads (t = 4.442, p < .001), and
perceived humor is stronger for advertising in broadcast media than in print media (t = 3.585, p =
.001). Because only three correlations of studies with non-student samples provided a
manipulation check, no test was performed. As mentioned before, most academic studies are
done in laboratory settings using print ads to evaluate humor’s effects. When testing the joint
influence of students (bias upward) and print ads (bias downward) by adding an interaction
variable, the interaction effect was not significant (b = .020, se = .086, p = .819), showing that
both effects seem to cancel out and do not operate independently.
Insert table 4 about here
Table 5 shows the results for the moderator variables of product categories and humor
relatedness applied to the correlations between humor and ABR. In line with the assumptions, the
impact of humor is enhanced for hedonic motives in both cases of related humor (yellow goods)
and unrelated humor (red goods). In the case of white goods (high involvement/risk and
functional), humor leads to stronger effects when it is related to the product. As for blue goods,
humor does not provide any further advantage, either in the case of related or unrelated humor.
Insert table 5 about here
Five of the studies in the meta-analysis provide a correlation coefficient for the relationship
between AAD and ABR, leading to a mean weighted correlation between both variables of .557
(p < .001). The studies in the meta-analysis include 35 correlations for the relationship between
humor and brand attitudes that provide a manipulation check measure of perceived humor. The
results of a bivariate weighted regression model show that perceived humor enhances ABR (b =
.603, se = .098, p < .001, R2 = .526, F = 37.685, p < .001). Adding a quadratic term to the
regression model only slightly changes the results, and the influence of the quadratic term is non-
significant (Table 6). The results support the positive and linear relationship between humor
intensity and the impact of humor on ABR (Figure 1).
Insert table 6 about here
Insert figure 1 about here
The results of the meta-analysis quantify and update previous literature reviews by providing
a set of empirical generalizations on the effects of humor in advertising. In line with the
assumptions of previous reviews, the meta-analytic correlations demonstrate that humor in
advertising significantly enhances AAD, attention, and positive affect. Contrary to earlier review
assumptions, there is no evidence that humor impacts positive or negative cognitions and liking
of the advertiser. Some of the conclusions of existing reviews were partly ambiguous. This meta-
analysis provides empirical results that can help to resolve ambiguous conclusions: humor
significantly reduces source credibility, enhances positive affect, ABR, and purchase intention.
Application of a random-effects model to effect size integration shows that the impact on ad-
related and brand-related recall and recognition is non-significant. The effect on comprehension
is also non-significant as is the effect on purchase behavior. For those results that are based on
large samples of studies and a large number of subjects, those findings should be rather robust.
However, most of the integrative findings are based on a quite heterogeneous set of effects (AAD,
ABR, attention, purchase behavior, recall and recognition) and the grand mean of the correlation
coefficient has to be considered as an average rather than a common correlation value. In
particular, when a small number of effects is heterogeneous as it is the case for purchase
behavior, the results have to be interpreted with caution.
Basically, however, the results support the notion that humor is more effective in generating
lower order than higher order communication effects; the effect size of the impact of humor on
AAD is even twice as large as the effect size for ABR. Such a decline from lower order to higher
order effects is more pronounced compared to the effects of other content elements in advertising
(e.g., Eisend 2006; Grewal et al. 1997). Practitioners should be aware that other executional
tactics may lose less of their impact on their way from attention and ad liking to brand preference
and brand choice. Still, in order to come to an overall conclusion for the effects on sales, the
analyst should bear in mind that s/he has to consider the total effect (both direct and indirect
Table 7 summarizes the integration results of the meta-analysis and compares them to the
assumptions of previous review. Most of the findings replicate what has been identified in the
last major review by Weinberger and Gulas (1992). In addition, the main results have not
changed over the years. Hence, the meta-analytic results may not only provide more certainty but
also suggest that the impact of humor in advertising has been rather stable over the past decades.
Insert table 7 about here
The moderator models show that humor effects on AAD depend on the characteristics of the
audience and the humorous stimulus; the effects of humor on ABR partially depend on the
product type as well as the relatedness of humor and product. The results of the moderator
analysis for humor effects on AAD in particular show that the academic stream of research on
humor in advertising may produce findings that are biased by the fact that those studies are
performed in laboratory settings. This bias stems from at least two factors. First, the use of
students enhances the effects of humor in advertising on AAD, because students represent young
and well educated consumers with cognitive abilities that help them to comprehend and
appreciate humor more easily. Second, print advertisements and fictitious ads reduce the effects
of humor, as they are less humorous than real advertisements and ads in broadcast media.
Although both biases seem to cancel each other out, effects from studies performed in laboratory
settings should be interpreted with caution. A validation of these studies by performing
additional studies with non-student samples, with ads in broadcast media, as well as with real ads
is advisable for future humor research.
The results of the moderator model also call for more questions. Comics, funny stories, and
other humorous print sources evoke mirth and laughter and are not necessarily inferior in terms
of funniness compared to film adaptations. Why is this not the case in print advertisements? An
explanation may be provided by the “media-fit” of different humorous styles: do humorous
styles used in advertising fit better with TV than with print media? Are there alternatives for
print advertising which have not yet been considered (e.g., comic adaptations)? Another question
refers to the fictitious ads that are typically used in studies by marketing academics. It would be
interesting to see in what way the humor perception and appreciation of academics differ from
that of students or the rest of the population. Such results could help to control the bias caused by
humorous advertisements that were created by scientists compared to real ads.
As for the impact on ABR, the propositions of the existing product typologies are largely
supported. Humor is appropriate for red and yellow goods, and it is appropriate for white goods
when humor is related to the product and therefore provides a brand claim. The impact of humor
is weaker for functional and low involvement goods (blue goods) in both cases of related and
unrelated humor. Practitioners should keep in mind, though, that humor basically enhances ABR,
and it is simply the strength of the effects that varies – depending on the combination of product
category and humor-product relatedness. So far, relatedness of humor has been suggested by
other authors as a potential moderator for humor effects (e.g., Speck 1991; Weinberger and
Campbell 1991). The meta-analysis shows how the effects of humor-product relatedness depend
on product categories; the meta-analysis thereby tests a classification that goes beyond previous
approaches of humor-relatedness effects (e.g., Cline and Kellaris 2007).
The results of the relationship between perceived humor and the correlations between humor
and ABR show a linear and positive relation between both variables. The result is in line with the
assumption of an affect transfer mechanism, and shows that humor does not necessarily distract
the attention of consumers from processing the brand-related parts of the message. The affective
mechanism seems to be the superior mechanism underlying the effect of humor in advertising.
Cognitive effects are apparently weaker, which supports the notion that the impact of humor is
less likely to be based on cognitive mechanisms such as enhanced information processing,
distraction, or persuasion effects due to source credibility. Such tentative conclusions are in
accordance with the integrative results of the meta-analysis. They show that humor enhances
attention but does not impact cognitive responses, indicating that humor neither distracts from
processing the message nor stimulates message processing. Though the overall effect of humor
on credibility is negative, the results of the relationship between perceived humor and the
correlations between humor and ABR are not in line with the credibility-mediation effect on brand
attitudes. Humor is a stimulus that apparently improves brand attitudes foremost by affective
processes and not by cognitive processes.
The study has some limitations that represent common problems of meta-analytic techniques.
One limitation results from restricting the sample to studies published in English. While such
language bias is commonly accepted in meta-analyses for practical reasons (one would need
many coders from different countries in order to review all non-English language marketing
journals) and for substantial reasons (English is the preferred language of science, and high
quality research is always published in Anglo-American journals), the culture-dependent effects
of humor in the present meta-analysis may be largely ignored as the majority of the studies
included in the meta-analysis were performed in the US. But use and appreciation of humor
shows large differences from culture to culture (Alden, Hoyer, and Lee 1993; Toncar 2001),
making the same kind of humor lead to different effects for audiences from different countries.
Further research is needed in order to sort out which kind of humor is most appropriate for each
culture in terms of effectiveness. A useful approach could be to test the congruency between
humor taxonomies (Speck 1991) and cultural dimensions (Hofstede 2001; House et al. 2004).
Moderator models are always restricted by the studies that are available for the analysis. The
regression model presented here could be refined by considering interaction effects between
audience, humor, and products. Inclusion of additional predictors always leads to the problem of
increasing missing values of the predictor variables (Stock 1994). As further studies become
available, additional moderators can be applied to the regression models and meaningful
moderator models could be applied to other dependent variables that are affected by humor in
advertising (attention, affect, source credibility). Nevertheless, this meta-analysis revealed some
insightful findings, provides some empirical generalizations and resolves some conflicting
findings in the literature. By identifying gaps in previous research, the meta-analysis will help to
ensure that the next wave of research is guided in a promising direction. While empirical work
on the impact of advertising is widespread, including detailed studies of executional elements,
only several meta-analyses have been performed in order to review quantitatively such
advertising effects (Lehmann and Reibstein 2006). As research integration is an essential step of
knowledge accumulation and refinement in science, linking past research with future scientific
endeavors and providing empirical generalizations that are useful for practical marketing
decisions, more meta-analytic work on the effects of executional elements in advertising would
help both researchers and practitioners.
(*asterisks denote manuscripts that were used for the meta-analysis)
Aaker, David, George Day, and Michael R. Hagerty. 1986. "Warmth in Advertising:
Measurement, Impact and Sequence Effects." Journal of Consumer Research 12 (March): 365-
Alden, Dana L., Wayne D. Hoyer, and Chol Lee. 1993. "Identifying Global and Culture Specific
Dimensions of Humor in Advertising: A Multinational Analysis." Journal of Marketing 57
*Alden, Dana L., Ashesh Mukherjee, and Wayne D. Hoyer. 2000a. "The Effects of Incongruity,
Surprise and Positive Moderators on Perceived Humor in Television Advertising." Journal of
Advertising 24 (2): 1-15.
*----. 2000b. "Extending a Contrast Resolution Model of Humor in Television Advertising: The
Role of Surprise." Humor: International Journal of Humor Research 13 (2): 193-217.
Beard, Fred K. 2005. "One Hundred Years of Humor in American Advertising." Journal of
Macromarketing 25 (1): 54-65.
*Belch, George E. and Michael E. Belch. 1984. "An Investigation of the Effects of Repetition on
Cognitive and Affective Reactions to Humorous and Serious Television Commercials." In
Advances in Consumer Research. Ed. Thomas C. Kinnear. Provo, UT: Association for Consumer
*Berg, Eron M. and Louis G. Lippman. 2001. "Does Humor in Radio Advertising Affect
Recognition of Novel Product Brand Names?" Journal of General Psychology 128 (2): 194-205.
Berlyne, Daniel E. 1972. "Humor and Its Kin." In Psychology of Humor. Eds. Jeffery H.
Goldstein and Paul E. McGhee. New York: Academic Press, 43-60.
Bommer, William H., Jonathan L. Johnson, Gregory A. Rich, Philip M. Podsakoff, and Scott B.
MacKenzie. 1995. "On teh Interchangeability of Objective and Subjective Measures of
Employee Performance: A Meta-Analysis." Personnel Psychology 48 (3): 587-605.
*Brooker, George. 1981. "A Comparison of the Persuasive Effects of Mild Humor and Mild Fear
Appeals." Journal of Advertising 10 (4): 29-40.
Brown, Steven P. and Douglas M. Stayman. 1992. "Antecedents and Consequences of Attitude
Toward the Ad: A Meta-Analysis." Journal of Consumer Research 19 (June): 34-51.
Bryant, Jennings, Dan Brown, Alan R. Silberberg, and Scott M. Elliott. 1981. "Effects of
Humorous Illustrations in College Textbooks." Human Communication Research 8 (1): 43-57.
*Cantor, Joanne and Pat Venus. 1980. "The Effect of Humor on Recall of a Radio
Advertisement." Journal of Broadcasting 24 (1): 13-22.
*Chattopadhyay, Amitava and Kunal Basu. 1990a. "Humor in Advertising: The Moderating Role
of Prior Brand Evaluation." Journal of Marketing Research 17 (November): 466-476.
----. 1990b. "Humor in Advertising: The Moderating Role of Prior Brand Evaluations." Journal
of Marketing Research 17 (November): 466-476.
*Chung, Hwiman and X. Zhao. 2003a. "Effects of Humor Ad: Moderating Role of Product
Familiarity." In Proceedings of the American Academy of Advertising Conference. Denver, CO:
American Academy of Advertising, 11.
*Chung, Hwiman and Xinshu Zhao. 2003b. "Humor Effect on Memory and Attitude:
Moderating Role of Product Involvement." International Journal of Advertising 22 (1): 117-145.
Churchill, Gilbert A., Jr., Neil M. Ford, Steven W. Walker, and Orville C. Walker, Jr. 1985. "The
Determinants of Salesperson Performance: A Meta-Analysis." Journal of Marketing Research 22
*Cline, Thomas W. 1997. "The Role of Expectancy and Relevancy in Humorous Ad Executions:
An Individual Difference Perspective." Dissertation. University of Cincinnati.
*Cline, Thomas W., Moses B. Altsech, and James J. Kellaris. 2003. "When Does Humor
Enhance or Inhibit Ad Responses?" Journal of Advertising 32 (3): 31-45.
Cline, Thomas W. and James J. Kellaris. 2007. "The Influence of Humor Strength and Humor-
Message Relatedness on Ad Memorability." Journal of Advertising 36 (1): 55-67.
*----. 1999. "The Joint Impact of Humor and Argument Strength in a Print Advertising Context:
A Case for Weaker Arguments." Psychology & Marketing 16 (1): 69-86.
Cohen, Jacob. 1977. Statistical Power Analysis for the Behavioral Sciences. New York:
*Conway, Michael and Laurette Dubé. 2002. "Humor in Persuasion on Threatening Topics:
Effectiveness Is a Function of Audience Sex Role Orientation." Personality and Social
Psychology Bulletin 28 (7): 863-873.
Dalton, Dan R., Catherine M. Daily, S. Trevis Certo, and Rungpen Roengpitya. 2003. "Meta-
Analyses of Financial Performance and Equity: Fusion or Confusion?" Academy of Management
Journal 46 (1): 13-26.
*De Pelsmacker, Patrick and Maggie Geuens. 1998. "The Advertising Effectiveness of Different
Levels of Intensity of Humor and Warmth and the Moderating Role of Affect Intensity." In
Proceedings of the Academy of Marketing Science. Norfolk, VA: Academy of Marketing
----. 1996. "The Communication Effects of Warmth, Eroticism and Humour in Alcohol
Advertisements." Journal of Marketing Communications 2: 247-262.
Duncan, Calvin P. 1979. "Humor in Advertising: A Behavioral Perspective." Journal of the
Academy of Marketing Science 7 (4): 285-306.
*Duncan, Calvin P. and James E. Nelson. 1985. "Effects of Humor in a Radio Advertising
Experiment." Journal of Advertising 14 (2): 33-40, 64.
*Duncan, Calvin P., James E. Nelson, and Nancy T. Frontczak. 1984. "The Effects of Humor on
Advertising Comprehension." In Advances in Consumer Research. Ed. Thomas C. Kinnear.
Provo, UT: Association for Consumer Research, 432-437.
Eisend, Martin. 2006. "Two-Sided Advertising: A Meta-Analysis." International Journal of
Research in Marketing 23 (2): 187-198.
Elliot, Richard, Susan Eccles, and Michelle Hodgson. 1993. "Re-Coding Gender
Representations: Women, Cleaning Products, and Advertising's 'New Man'." International
Journal of Research in Marketing 10 (3): 311-324.
Farley, John U., Donald R. Lehmann, and Lane H. Mann. 1998. "Designing the Next Study for
Maximum Impact." Journal of Marketing Research 35: 496-501.
Flaherty, Karen, Marc G. Weinberger, and Charles S. Gulas. 2004. "The Impact of Perceived
Humor, Product Type and Humor Style in Radio Advertising." Journal of Current Issues and
Research in Advertising 26 (1): 25-36.
Furnham, Adrian, Barrie Gunter, and Deidre Walsh. 1998. "Effects of Programme Context on
Memory of Humorous Television Commercials." Applied Cognitive Psychology 12 (6): 555-567.
Furnham, Adrian and Tadataka Mori. 2003. "The Effect of Programme Context on Memory For
Humorous Television Advertisements in Japan." Psychologia 46 (1): 53-66.
*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.
*Gelb, Betsy D. and George M. Zinkhan. 1986. "Humor and Advertising Effectiveness After
Repeated Exposures to a Radio Commercial." Journal of Advertising 15 (2): 15-20,34.
Geuens, Maggie and Patrick De Pelsmacker. 2002. "The Role of Humor in the Persuasion of
Individuals Varying in Need for Cognition." In Advances in Consumer Research. Eds. Susan M.
Broniarczyk and Kent Nakamoto. Valdosta, GA: Association for Consumer Research, 50-56.
Godkewitsch, Michael. 1976. "Physiological and Verbal Indices of Arousal in Rated Humour."
In Humour and Laughter: Theory, Research and Application. Eds. Antony J. Chapman and
Hugh C. Foot. London: John C. Wiley & Sons, 117-138.
Grewal, Dhruv, Sukumar Kavanoor, Edward F. Fern, Carolyn Costley, and James Barnes. 1997.
"Comparative Versus Noncomparative Advertising: A Meta-Analysis." Journal of Marketing 61
Gulas, Charles S. and Marc G. Weinberger. 2006. Humor in Advertising. A Comprehensive
Analysis. Armonk, NY: M. E. Sharpe.
Hedges, Larry V. 1994. "Fixed Effect Models." In The Handbook of Research Synthesis. Eds.
Harris Cooper and Larry V. Hedges. New York: Russell Sage Foundation, 285-299.
Hedges, Larry V. and Ingram Olkin. 1985. Statistical Methods for Meta-Analysis. Orlando, FL:
Hofstede, Geert H. 2001. Culture's Consequences: Comparing Values, Behaviors, Institutions,
and Organizations Across Nations. Thousand Oaks, CA: Sage.
House, Robert J., Paul J. Hanges, Mansour Javidan, Peter W. Dorfman, and Vipin Gupta. 2004.
Culture, Leadership, and Organizations. The GLOBE Study of 62 Societies. Thousand Oaks, CA:
Hunter, John E. and Frank L. Schmidt. 1990. Methods of Meta-Analysis. Correcting Error and
Bias in Research Findings. Newbury Park, CA: Sage.
----. 2004. Methods of Meta-Analysis. Correcting Error and Bias in Research Findings.
Thousand Oaks: Sage.
Kayande, Ujwal and Mukesh Bhargava. 1994. "An Examination of Temporal Patterns in Meta-
Analysis." Marketing Letters 3 (2): 141-151.
*Krishnan, H. Shanker and Dipankar Chakravarti. 2003. "A Process Analysis of the Effects of
Humorous Advertising Executions on Brand Claims Memory." Journal of Consumer Psychology
Lammers, H. Bruce. 1991. "Moderating Influence of Self-Monitoring and Gender on Responses
to Humorous Advertising." The Journal of Social Psychology 131 (1): 57-69.
Lammers, H. Bruce, Laura Liebowitz, George Edward Seymour, and Judith E. Hennessey. 1983.
"Humor and Cognitive Response to Advertising Stimuli: A Trace Consolidation Approach."
Journal of Business Research 11 (2): 173-185.
*Lee, Moon J. and Mary Ann Ferguson. 2002. "Effects of Anti-Tobacco Advertisements Based
on Risk-Taking Tendencies: Realistic Fear vs. Vulgar Humor." Journalism & Mass
Communication Quarterly 79 (4): 945-963.
*Lee, Yih Hwai. 1997. "The Immediate and Delayed Effects of Advertising: The Role of
Information Incongruency." Dissertation. University of North Carolina at Chapel Hill.
Lee, Yih Hwai and Charlotte Mason. 1999. "Responses to Information Incongruency in
Advertising: The Role of Expectancy, Relevancy, and Humor." Journal of Consumer Research
26 (September): 156-169.
Lehmann, Donald R. and David J. Reibstein. 2006. Marketing Metrics and Financial
Performance. Cambridge, MA: Marketing Science Institute.
MacKenzie, Scott B. and Richard J. Lutz. 1989. "An Empirical Examination of the Structural
Antecedents of Attitude Toward the Ad in an Advertising Pretesting Context." Journal of
Marketing 53 (April): 48-56.
Madden, Thomas J. 1982. "Humor in Advertising: Applications of a Hierarchy of Effects
Paradigm." Dissertation. University of Massachusetts.
Madden, Thomas J. and William R. Dillon. 1982. "Causal Analysis and Latent Class Models: An
Application to a Communication Hierarchy of Effects Model." Journal of Marketing Research
19 (November): 472-490.
*Madden, Thomas J. and Marc G. Weinberger. 1982. "The Effects of Humor on Attention in
Magazine Advertising." Journal of Advertising 11 (3): 8-14.
----. 1984. "Humor in Advertising: A Practitioner View." Journal of Advertising Research 24
Mak, Wingyun and Brian Carpenter. 2007. "Humor Comprehesion in Older Adults." Journal of
the International Neuropsychological Society 13 (4): 606-614.
McGhee, Paul E. 1986. "Humor Across the Life Span: Sources of Developmental Change and
Individual Differences." In Humor and Aging. Eds. L. Nahemow, K. A. McCluskey-Fawcett, and
Paul E. McGhee. New York: Academic Press, 27-51.
*Michaels, Steven L. 1997. "Cognitive and Affective Responses to Humorous Advertisements."
Dissertation. Wayne State University.
*Mukherjee, Ashesh and L. Dube. 2001. "The Use of Humor in Threat-Related Advertising: An
Experimental Processing Perspective." In European Advances in Consumer Research. Eds.
Andrea Gröppel-Klein and Franz-Rudolf Esch. Valdosta, GA: Association for Consumer
*Mullinax, Sharon Leann. 1984. "The Use of Humor to Prevent Wearout in Advertising."
Dissertation. Lamar University.
Murphy, John H., Isabella C. M. Cunningham, and Gary Wilcox. 1979. "The Impact of Program
Environment on Recall of Humorous Television Commercials." Journal of Advertising Research
8 (2): 17-21.
Nelson, James E., Calvin P. Duncan, and Nancy T. Frontczak. 1985. "The Distraction
Hypothesis and Radio Advertising." Journal of Marketing 49 (Winter): 60-71.
*Perry, Stephen D., Stefan A. Jenzowsky, Cynthia M. King, Huiuk Yi, Joe Bob Hester, and
Jeanne Gartenschlaeger. 1997. "Using Humorous Programs as a Vehicle for Humorous
Commercials." Journal of Communication 47 (1): 21-39.
Petty, Richard E. and Duane T. Wegener. 1999. "The Elaboration Likelihood Model: Current
Status and Controversies." In Dual Process Theories in Social Psychology. Eds. Shelly Chaiken
and Yaacov Trope. New York: Guildford Press, 41-72.
Prilluk, Randi and Brian D. Till. 2004. "The Role of Contingency Awareness, Involvement, and
Need for Cognition in Attitude Formation." Journal of the Academy of Marketing Science 32 (3):
Rosenthal, Marylu C. 1994. "The Fugitive Literature." In The Handbook of Research Synthesis.
Eds. Harris Cooper and Larry V. Hedges. New York: Russell Sage Foundation, 85-94.
Rosenthal, Robert. 1979. "The 'File Drawer Problem' and Tolerance for Null Results."
Psychological Bulletin 86 (3): 638-641.
Rossiter, John R., Larry Percy, and Robert J. Donovan. 1991. "A Better Advertising Planning
Grid." Journal of Advertising Research 31 (October/November): 11-21.
*Scott, Cliff, David M. Klein, and Jennings Bryant. 1990. "Consumer Response to Humor in
Advertising: A Series of Field Studies Using Behavioral Observation." Journal of Consumer
Research 16 (March): 498-501.
Shadish, William R. and C. Keith Haddock. 1994. "Combining Estimates of Effect Sizes." In The
Handbook of Research Synthesis. Eds. Harris Cooper and Larry V. Hedges. New York: Russell
Sage Foundation, 261-281.
*Shifman, Richard Bennett. 1994. "Take my Brand . . .Please: Attitudinal Effects of Functional
Relationships Among Type of Humorous Appeal, Context, and Seriousness of Salient Product
Attributes in Print Advertisements." Dissertation. Temple University.
*Skinner, Deborah, R. Mackoy, and G. Osland. 2000. "Does Need for Cognition Moderate the
Effectiveness of Ironic Humor in Advertising? Or What Does it Take to Get the Message?" In
Proceedings of the AMA Summer Marketing Educator´s Conference. Chicago, IL: American
Marketing Association, 139-140.
Smith, Stephen M. 1993. "Does Humor in Advertising Enhance Systematic Processing?" In
Advances in Consumer Research. Eds. L. McAlister and Michael L. Rothschild. Provo, UT:
Association for Consumer Research, 155-158.
Speck, Paul Surgi. 1991. "The Humorous Message Taxonomy: A Framework for the Study of
Humorous Ads." Current Issues and Research in Advertising 14 (1): 1-44.
*----. 1987. "On Humor and Humor in Advertising." Dissertation. Texas Tech University.
*Spotts, Harlan E., Martin G. Weinberger, and Amy L. Parasons. 1997. "Assessing the Use and
Impact of Humor on Advertising Effectiveness: A Contingency Approach." Journal of
Advertising 26 (3): 17-32.
Stanton, John L. and Jeffrey Burke. 1998. "Comparing the Effectiveness of Executional
Elements in TV Advertising: 15- versus 30-second Commercials." Journal of Advertising
Research 38 (6): 7-14.
Sternthal, Brian and Samuel Craig. 1973. "Humor in Advertising." Journal of Marketing 37
*Stewart, David W. and David H. Furse. 1986. Effective Television Advertising: A Study of 1000
Commercials. Massachusetts: D.C. Heath and Company.
Stock, William A. 1994. "Systematic Coding for Research Synthesis." In The Handbook of
Research Synthesis. Eds. Harris Cooper and Larry V. Hedges. New York: Russell Sage
Suls, Jerry M. 1972. "A Two-stage Model for the Appreciation of Jokes and Cartoons." In The
Psychology of Humour. Theoretical Perspectives and Empirical Issues. Eds. P. E. Goldstein and
J. H: McGhee. New York: Academic Press, 81-100.
Sultan, Fareena, John U. Farley, and Donald R. Lehmann. 1990. "A Meta-Analysis of
Applications of Diffusion Models." Journal of Marketing Research 27 (February): 70-77.
*Sutherland, John C. and Lisa A. Middleton. 1983. "The Effect of Humor on Advertising
Credibility and Recall." In Proceedings of the 1983 Conference of the American Academy of
Advertising. Ed. Donald W. Jugenheimer. Lawrence, KS: American Academy of Advertising,
*Sutherland, John C. and Sudha Sethu. 1987. "The Effect of Humor on Television Advertising
Credibility and Recall." In Proceedings of the 1997 Conference of the American Academy of
Advertising. Ed. Florence G. Feasley. Columbia, SC: American Academy of Advertising, R3-R8.
Szymanski, David M., Sundar G. Bharadwaj, and Rajan P. Varadarajan. 1993. "An Analysis of
the Market Share-Profitability Relationship." Journal of Marketing 57 (July): 1-18.
Szymanski, David M., Lisa C. Troy, and Sundar G. Bharadwaj. 1995. "Order of Entry and
Business Performance: An Empirical Synthesis and Reexamination." Journal of Marketing 59:
Tellis, Gerard J. 1988. "The Price Elasticity of Selective Demand: A Meta-Analysis of
Econometric Models of Sales." Journal of Marketing Research 25 (November): 331-341.
Toncar, Mark F. 2001. "The Use of Humor in Television Advertising: Revisiting the US-UK
Comparison." International Journal of Advertising 20 (4): 521-539.
Unger, Lynette S. 1995. "Observations: A Cross-cultural Study on the Affect-Based Model of
Humor in Advertising." Journal of Advertising Research 35 (1): 66-71.
van Raaij, W. Fred. 1993. "Postmodern Consumption." Journal of Economic Psychology 14:
Vaughn, Richard. 1980. "How Advertising Works: A Planning Model." Journal of Advertising
Research 20 (October/November): 27-33.
----. 1986. "How Advertising Works: A Planning Model Revisited." Journal of Advertising
Research 26 (February/March): 57-66.
*Weinberger, Marc G. and Leland Campbell. 1991. "The Use and Impact of Humor in Radio
Advertising." Journal of Advertising Research 30 (6): 44-52.
Weinberger, Marc G., Leland Campbell, and Beth Brody. 1994. Effective Radio Advertising.
New York: Lexington Books.
Weinberger, Marc G. and Charles S. Gulas. 1992. "The Impact of Humor in Advertising: A
Review." Journal of Advertising 21 (4): 35-59.
Weinberger, Marc G., Harlan Spotts, Leland Campbell, and Amy L. Parsons. 1995. "The Use
and Effect of Humor in Different Advertising Media." Journal of Advertising Research 35 (3):
Woltman Elpers, Josephine L. C. M., Ashesh Mukherjee, and Wayne D. Hoyer. 2004. "Humor in
Television Advertising: A Moment-to-Moment Analysis." Journal of Consumer Research 31
*Wu, Bob T. W., Kenneth E. Crocker, and Martha Rogers. 1989. "Humor and Comparatives in
Ads For High- and Low-Involvement Products." Journalism Quarterly 66: 653-661,780.
Zhang, Yong. 1992. "Audience Involvement and Persuasion in Humorous Advertising."
Dissertation. University of Houston.
*----. 1996a. "The Effect of Humor in Advertising: An Individual-Difference Perspective."
Psychology & Marketing 13 (6): 531-546.
*----. 1996b. "Responses to Humorous Advertising: The Moderating Effect of Need for
Cognition." Journal of Advertising 25 (1): 15-32.
Zhang, Yong and George M. Zinkhan. 1991. "Humor in Television Advertising: The Effects of
Repetition and Social Setting." In Advances in Consumer Research. Eds. Rebecca H. Holman
and Michael R. Solomon. Provo, UT: Association for Consumer Research, 813-818.
----. 2006. "Responses to Humorous Ads. Does Audience Involvement Matter?" Journal of
Advertising 35 (4): 113-127.
Zillmann, Dolf, Brien R. Williams, Jennings Bryant, Kathleen R. Boynton, and Michelle A.
Wolf. 1980. "Acquisition of Information from Educational Television as a Function of
Differently Paced Humorous Inserts." Journal of Educational Psychology 72 (2): 170-180.
Assumed effects of the impact of humor in advertising in previous literature reviews
Reviews by academics
Attitude towards the advertiser
Cognitive responses, positive
Cognitive responses, negative
Note: “+” indicates that humor enhances the outcome variable; “-“ indicates that humor decreases the
outcome variable, “O” indicates that humor has no additional effect over serious messages, “?” refers
to mixed findings, empty cells indicate that the review does not provide any conclusions related to the
particular outcome variable.
Assumed moderating effects on the impact of humor on ABR
Type of product
White goods (high involvement/risk, functional)
Blue goods (low involvement/risk, functional)
Red goods (high involvement/risk, hedonic)
Yellow goods (low involvement/risk, hedonic)
Note: “+” indicates that moderator enhances the effect of humor on ABR; “-“ indicates that the moderator
decreases the effect of humor on ABR, “O” indicates that the moderator has no effect.
Meta-analytic correlations (random-effects model)
Attitude towards the advertiser
Cognitive responses, positive
Cognitive responses, negative
a “k” refers to the number of effect sizes.
b All effect sizes were sample size weighted and corrected for measurement error and artificial
dichotomization of continuous variables; weights for considering multiple effect sizes were applied.
c The fail-safe N is computed for α = .05 and provided for significant mean correlations.
*p < .05; **p < .01; ***p < .001.
Impact of moderator variables on the correlations between humor and AAD (n = 82)
Ficitious vs. real ads
Print vs. broadcast media
Note: No intercept was included as only dummy variables were applied to the regression analysis.
Explained variance = .482
Impact of moderator variables on the correlations between humor and ABR (n = 41)
Type of product
White goods (high involvement/risk, functional)
Blue goods (low involvement/risk, functional)
Red goods (high involvement/risk, hedonic)
Yellow goods (low involvement/risk, hedonic)
Note: No intercept was included as only dummy variables were applied to the regression analysis.
Explained variance =.630
Impact of perceived humor on the correlations between humor and ABR (n = 35)
Perceived humor (quadratic term)
R2 = .526; F(2,33) = 18.332, p < .001
Assumed effects of the impact of humor in advertising in previous literature reviews and findings of the meta-analysis
Reviews by academics
Sternthal and Craig
Attitude towards the advertiser
Cognitive responses, positive
Cognitive responses, negative
Note: “+” indicates that humor enhances the outcome variable; “-“ indicates that humor decreases the outcome variable, “O” indicates that humor
has no additional effect over serious messages, “?” refers to mixed findings, empty cells indicate that the review does not provide any
conclusions related to the particular outcome variable.
Relationship between humor intensity and the correlation between humor and ABR
correlation between humor and attitude towards the brand
humor intensity (perceived humor)