The consumer is not buying your
jokes
A research on the effects of humor in advertisements
Henri Willems
Student number: 1595504
The consumer is not buying your jokes
A research on the effects of humor in advertisements
Henri Willems
Management summary
In this research the effectiveness of using humor in advertisements for high involvement products and its potential difference for utilitarian or hedonic products has been investigated. The most important goal of this research was to create more understanding of the effectiveness of using humor within advertising, while there are quite some contrasting views on its applicability, and further broaden the field of research on humor and advertising by taking a specified look at the role of product nature and involvement on the evaluation of the products shown in the (no) humor ads.
Print ads have specially been designed for this research, whereby the humor ads contain the picture of a product, combined with a pre-tested funny slogan. These same ads have also been designed with no humor, whereby the slogan in the ad has been replaced by a functional description of the product in the ad. Participants in this research were randomly assigned to one of the total eight conditions in this research, which was made up of a combination of humor vs. no humor, hedonic vs. utilitarian product, and high vs. low involvement. After having been assigned to a condition group, they were shown an ad, and asked questions in order to measure their evaluation of the product shown in the ad. After having gathered data, different condition groups were compared and hypotheses tested.
Results coming from this research suggest that for utilitarian products in general, independent of involvement, the use of humor in advertising results in a significantly lower evaluation of the product shown in the ad than when no humor was used. For hedonic products in general, if humor or not was used in the ad, no difference has been found in the evaluation of the products in the ad. Furthermore, other relevant findings were that when humor was used in the ads, product evaluation was significantly lower than when the ad was made without the use of humor. And although the findings with regard to product nature and humor usage could not be significantly specified to product involvement, it is worth mentioning that in this research the participants in the low involvement conditions significantly evaluated products shown in the ads as higher than the participants in the high involvement conditions.
irrespective of the product nature being hedonic or utilitarian, the usage of humor instead of no humor would lead to more positive product evaluations.
Most important implications of this research that are interesting for managers to think about are that results in this research do not argument in favor of the usage of humor in ads for different products. So, although for a lot of firms and managers the use of humor in advertisements is being broadly applied, it should never become a no-brainer while this research has provided counterevidence against the effectiveness of using humor. Managers can use the results from this research, which further specifies the relationship between humor ad effectiveness and product nature, to reflect on the nature of the product that they are planning to launch and conclude on the appropriateness of using humor in the corresponding advertisement campaign. It would be wise for managers to compare findings from this research with previous results on the effectiveness of using humor, lay all findings alongside the specific product that managers are planning to create an advertisement campaign for, and draw a more funded conclusion on the persistence of using humor in the ad or not.
Preface
Performing academically viable research is not easy. Furthermore, it's quite time consuming, even more than I would have expected upfront.
However, I have to say, it's been worth it, I learned a lot, and I'm quite proud of the work I did since starting with this Master's thesis in March 2010.
When starting to think about possible fields of interest to write my thesis in, I quite fast thought about performing a research on the use of humor in the field of advertising. This always had my interest, when seeing a humorous ad, or watching a commercial which used humor in it, I often thought about how effective it actually would be for the firm promoting the product in the ad or commercial to use humor. Wouldn't people seeing the ad or watching the commercial be distracted from the actual product being the subject in the ad and thus 'forget' the product and only remember the joke? With this angle of thought in my head I started further conceptualizing my idea to do a research on the effectiveness of humor in advertisement with regard to the evaluation of consumers of the products in the ads. You are now looking at the eventual result of this original idea which has further evolved into a complete Master thesis research and I can only wish that you will read it with interest.
I would like to take the chance to thank some people that have been supporting and directing me in the last year. First of all I would like to say thank you to dr. Jia Liu who has been my first supervisor in the last eight months, she has been guiding the process I went through from beginning to end and provided me with interesting alternative angles to look at, and constructive feedback which brought this thesis to the level it's currently at. I also thank my second advisor dr. Debra Trampe for her feedback on this thesis.
Furthermore, I own special thanks to my girlfriend and forthcoming wife Marijn for all her help and for staying patient in times when I got bored or frustrated at moments when progress seemed to be too little in my view. I also want to express my gratitude to the people that have participated in my research, be it in the pre-test or in the experiment, thank you, especially those who also took the extra time to pass my research questionnaires on to others to further expand the experiment sample.
I would like to finalize this preface by citing Aristotle in order to summarize the overall feeling I personally keep after writing this thesis: "What one must learn to do, one learns by doing"
Groningen, 20 November 2010
Table of contents
1. Introduction
8
1.1 Background and context
8
1.2 Problem analysis
9
1.3 Relevance
12
1.4 Build up thesis
15
2. Theoretical Framework
16
2.1 Introduction
16
2.2 Humor ads and contextual effects
17
2.3 Product involvement and nature
23
3. Methodology
31
3.1 Introduction
31
3.2 Research design
31
3.3 Participants and sample
32
3.4 Variables
32
3.5 Questionnaire and procedure
36
3.7 Plan of analysis
37
3.8 Validity
37
4. Results
39
4.1 Introduction
39
4.2 Manipulation check
39
4.3 Pre-test
41
4.4 Data set
42
4.5 Sample characteristics
43
4.6 Reliability of the scale
47
4.7 Hypotheses tests
48
4.8 Summary of results
53
5. Conclusions & Recommendations
55
5.1 Introduction
55
5.2 Summary
55
5.3 Discussion
56
5.4 General conclusion
61
5.5 Recommendations
62
5.6 Limitations and further research
62
References
64
Appendices
69
1. Questionnaire
69
2. Ads
75
1. Introduction
1.1 Background and context
The usage of humor within advertising is considered a controversial issue for practitioners (advertising managers) as well as for writers, theorists, and researchers (Sternthal and Craig, 1973). Based on research considering the amount of usage of humor in advertising, estimations are that between 30% and 40% of every ad employs some kind of humor (Markiewicz, 1974; Weinberger, Spotts, Campbell and Parsons, 1995; Spotts, Weinberger and Parsons, 1997), hereby proving the intensive application of humor in the world of advertising. However, despite the fact that humor is broadly applied in advertising, its use in advertising remains a controversial issue for managers and researchers alike (Kellaris and Kline, 2007). The main reason for its controversiality is the fact that ads attempting humor may vary dramatically in the level of humor they actually evoke in the target audience (Speck, 1991). This variation in perceived humor is likely to have important consequences for downstream variables of interest to marketers such as message credibility, recall, and attitude towards the ad and brand (Chattopadhyay and Basu, 1990; Shimp, 1997; Weinberger and Gulas, 1992).
Humor is thus being used with quite high frequency within advertising, but is it useful in every setting and for every product and its product nature? A lot of research done on the topic of humor and advertising is based on the same type of products. Most of the times the products used are novel products, which are actually non-existing fictive products, hereby not making a distinction based on the nature of the products. Put otherwise, there are certain differences in the nature of products, which are currently being excluded from most research. In general, products can be divided into two major groups based on their product nature: hedonic and utilitarian products. Hedonic products are "goods that provide more experiential consumption, fun, pleasure and excitement" (Dhar and Wertenbroch, 2000; Hirschman and Holbrook, 1982; Strahilevitz and Myers, 1998). Utilitarian products are "products that are mainly bought for informational reasons, including instrumental and utilitarian reasons (Batra and Ahtola, 1990), thereby being described as primarily instrumental and functional (Hirschmann and Holbrook, 1982; Strahilevitz and Myers, 1998; Dhar and Wertenbroch, 2000). It is interesting to investigate the possible different effects humor in ads will have for utilitarian and hedonic products while, for example, humor may seem at first as being more appropriate for hedonic than utilitarian products based solely on the different characteristics as described above.
research conducted on the effects of humor in advertising have focused on using more low involvement, often novel products or goods as the target product in an advertising to measure the effects of this advertising on product attitude, brand attitude or other outcome variables, thereby leaving the involvement of consumers for the product excluded. By including the involvement variable in this research, specifically by looking at more high involved products, different slightly contradicting outcomes from research of the past can be tested. Further, the possible differential effects on more hedonic products and more utilitarian products with regard to humor is an interesting angle of incidence to look into, certainly combined with the high involvement variable. A combination of searching for the outcomes of using humor with advertising for more high involved utilitarian and hedonic products is quite uncultivated area at this point and will basically be the main contribution of this research to the field of academics and practice.
1.2 Problem analysis
After having shortly introduced the topic of this thesis, and the most important variables and elements that will be included in the research, hereafter these elements will be looked into more deeply.
1.2.1 Humor
1.2.1.1 Humor in advertising
As already mentioned in the introduction part, humor in advertising is a widely used phenomenon. Humor is something that cannot be defined easily, neither can the effects of humor invoked within people. Or, to state the words of Kellaris and Cline (2007) "a clear understanding of the dynamics of humor in general, let alone humor's effects in advertising, has yet to emerge". So, not only the concept of humor in general is leaving room for further understanding, but also (but maybe that's a logical effect of the former statement) the concept of humor in advertising does.
It will be interesting to test the effectiveness of humor at more high involvement hedonic and utilitarian products, thereby testing and possibly exposing differences between hedonic and utilitarian products.
1.2.1.2 Humor as a contextual cue
Humor in advertising can be seen as a contextual cue. Extensive literature and research has been conducted in the field of importance of media context. Results indicate that, for example, media context can affect ad recall (De Pelsmacker, Geuens, and Anckaert, 2002), ad attitude and cognitions (Coulter, 1998), ad recognition (Moorman, Neijens, and Smit, 2002), level and nature of ad processing (Shapiro, MacInnis, and Park, 2002), brand attitude (Lord, Burnkrant, and Unnava, 2001) and purchase intentions (Yi, 1990). Further, Meyers-Levy and Tybout (1997) have conducted innovatory research on the role of context at encoding and judgment in consumption surroundings. By being aware of the fact that humor in advertising is a contextual cue, this implies thus that humor can have a strong effect on different outcome variables related to a product, such as brand attitude and purchase intentions. It will be interesting to test the influence as humor being a contextual cue in an advertisement for high involvement hedonic and utilitarian products and see if the strong implications of humor will hold in this setting.
1.2.3 Nature of products
It will thus be interesting to test if humor as a contextual cue for hedonic and utilitarian products will result in different outcomes with regard to product evaluation.
1.2.4 Product involvement
To build on the arguments mentioned about the nature of products and why it is interesting to include this aspect in this research, another important variable which will be included is product involvement. Product involvement can be described as an important aspect underlying consumer interest and motivation when searching, using, and buying products or services (Percy and Elliot, 2009) and it can be low or high on the extremes.
Described shortly; consumers being high involved with a product show high motivation with a product and want to spend more time elaborating a communication message that promotes the product and characteristics of the product itself, while consumers being low involved have low motivation with a product and want to spend less time on the elaboration of a communication message that promotes the product and the characteristics of the product itself (Petty and Cacioppo, 1981).
As mentioned in the introduction above, most of the research done in the past on the role of humor in advertising made use of novel, low involvement products to test the effects of humor on the evaluation for these products. The applicability of the conclusions from all research done in this field, having low involvement products as their target products, on more high involvement products will be an interesting angle of research. Also based on the above shortly described differences between low and high involvement products, it seems plausible to have expectations that for high involvement products, humor as a contextual cue might have different effects than for low involvement products. This leads to the expectation that the product involvement is another important variable influencing the effectiveness of humor as a contextual cue in advertising.
1.2.5 Product evaluation
1.2.6 Research question
In summary, the main topic underlying this research is the effectiveness of humorous advertisements on product evaluation. The main variables which are expected to be of influence on the effectiveness of humorous advertisements are product involvement (high vs. low) and product nature (hedonic vs. utilitarian).
Putting the above mentioned elements together, the following research question will be formulated:
"What is the effect of humorous advertisements on high involvement products and to what extent does this effect differ for hedonic and utilitarian goods?"
1.2.7 Conceptual model
The research question described above can be captured in an conceptual model, which can be seen below in Figure 1. The hypotheses and proposed relations within the conceptual model will be further explained and discussed in the theoretical framework (chapter 2).
Humor Advertisement Product Evaluation Product Involvement - high - low Product Nature - hedonic - utilitarian
Figure 1: Conceptual framework
1.3 Relevance
Hereafter, the academic relevance of this research will be described, followed by the most important reasons for its practical relevance.
There are several arguments for the academic relevance of this research. For starters, it seems that more research on the relation of humor in advertising with for example brand attitude, product attitude, product evaluation, purchase intention and more is based on the usage of low involvement, often novel and fictive products. In this research the subjects will be high involvement products, both hedonic and utilitarian of nature. Further, investigation of the effects of humor on an outcome variable for different product types (white goods, red goods, blue goods, yellow goods: Weinberger et al, 1994) has been done only rarely (Gulas and Weinberger, 2006), while a combination of high involvement and hedonic or utilitarian and the investigation of possible differences between these two product types related to humor is quite new in the field of academics and therefore seems contributory to the literature.
Another field of research, related to humor and advertising, concerns the effects of using emotions and emotional advertising (e.g. Johar and Sirgy, 1991; Batra and Stephens, 1994). Humor and emotions seem quite closely related to each other (Alden, Mukherjee and Hoyer, 2000), so it seems an extra underpinning for the hypotheses of this research to test if the effects that occurred in emotional advertising research, also hold in humorous advertising research.
Another, more recently published, research is that of Strick et al (2009) proving the positive influence of mere association of humor with products, thereby implicitly changing the attitudes of respondents towards these products, and enhancing product liking, evaluation and product choice. Products used in the research however were low involvement goods, it seems academically relevant and interesting to broaden this research by testing if these outcomes also hold for more high involvement hedonic and utilitarian goods.
Furthermore, the highly valued findings from the research of Meyers-Levy and Tybout (1997) about the role of context at encoding and judgment in consumption surroundings will be applied in this research to test the results and see if, in a high involvement product setting with humor as a contextual cue, the outcomes that can be expected based on the research of Meyers-Levy and Tybout will hold true. Especially because of the fact that the subject in this research are high involvement goods, it will be interesting to see if respondents will indeed show sufficient motivation to process and partial out the inappropriate information of the context (for utilitarian high involvement goods), leading to overcorrecting and a less positive judgment. And if the expected effects of using humor as a context for hedonic high involvement goods will indeed lead to a assimilation effect and thus more positive judgment. An alternative is that the findings of Raghunathan and Irwin (2001) will persist and the opposite effect will happen.
involvement means using the central route leading to the expectation that using humor only in advertisement for these products would not be sufficient in persuading consumers while they need more cognitive elaboration) will hold true. Different contradicting evidence on this aspect from other research of e.g. Eisend (2009) and Spotts et al (1997) proofs at least an interesting basis for discussion and academic relevance of this thesis.
Lastly, this thesis will contribute to existing literature on advertising research and the effects of using humor.
1.3.2 Practical relevance
This thesis will contribute to practitioners in the field of advertising and marketing because of the applicability of outcomes. As described above in the academic relevance part, interesting outcomes for more high involvement products, both hedonic and utilitarian, on the field of effectiveness of associating humor with it, will be the result.
Practitioners can use these outcomes to base their decisions with regard to the usage of humor in advertising for certain hedonic or utilitarian products when consumers are highly involved on. Is it, for example, wise to use humor in advertisement when you are the brand manager of a high involvement hedonic product such as Burton Snowboards? Or, to take an example of a high involvement utilitarian product, should humor be used in advertisement for Miele washing-machines? The findings from this research will be applicable to brand managers, marketing managers and other managers that have to make important decisions about the contents of its advertisements.
at Unilever to decide for which of these two products humor in an advertisement can contribute positively to the evaluation of this product, or maybe even for none of the two?
In summary, the practical relevance of this research lies mainly in the applicability of the outcomes for firms selling high involvement products on the marketplace, while they can really underpin the choices they make with regard to the (non-) usage of humor in advertisement for certain products.
1.4 Build up thesis
2. Theoretical Framework
2.1 Introduction
Hereafter, the main reasoning underlying the problem statement mentioned in chapter 1, as well as the basics underlying the conceptual model, will be discussed. Furthermore, the hypotheses resulting from this discussion will be described.
First of all, there seems to be a poverty in literature concerning the research on the role of humor divided over all the product categories, or, to state Spotts et al (1997): "Despite the knowledge that product category matters, most prior humor and message research has either ignored product variation or was unable to cope with such analysis because small data sets did not allow subdivision of the sample". Based on the gathered information about how advertisers are using humor in magazine advertising (extracted from collection of magazines evaluated by Starch during a five-year period), Spotts, Weinberger and Parsons (1997) made an overview of humor used in advertising, broken down to four product categories in the Product Color Matrix (Weinberger et al, 1994).
The product categories in this Product Color Matrix are made up from two variables, 'Risk' and 'Consumer Objective'. Risk can be translated also as 'Involvement' which can be high or low. Consumer Objective can be divided into Functional tools and Expressive toys, while in this research Functional tools will be translated as utilitarian goods and Expressive toys will be translated as hedonic goods.
These variables together form a 2x2 matrix, where four product types are the result:
white goods: high involvement and utilitarian products
red goods: high involvement and hedonic products
blue goods: low involvement and utilitarian products
yellow goods: low involvement and hedonic products
with their utilitarian counterpart on the same level of involvement. But on the other hand, overall, the only striking difference in usage of humor is with white goods (high involvement and utilitarian products) with a low 23% score. The other three categories all score above 60% and differences in usage of humor between red, yellow and blue goods are not shockingly high.
Eisend (2009) however, performed a meta-analysis on the role of humor in advertising, thereby also addressing the differences in the impact of humor on the different product types. Following his meta-analysis, humor seemed significantly appropriate for red and yellow goods (high and low involvement hedonic products), while it is appropriate for white goods (high involvement and utilitarian products) only when humor is related to the product. For blue goods (utilitarian and low involvement products), humor is concluded not to have a significant further advantage, independent of its relatedness with the product.
Summarizing, the research of Spotts et al (1997) investigating the usage in practice of humor in advertising, reveals a higher usage of humor for more hedonic products (red and yellow goods) when compared to their counterparts on the same level of involvement. However, difference in usage of humor between low involvement products is negligible, with a usage rate of 71% for yellow goods (hedonic) and 64% for blue goods (utilitarian). The only real convincing fact is the low usage (23%) of humor for high involvement and utilitarian products (white goods). Looking however at the research performed by Eisend (2009) on the effects of humor in advertising, humor is significantly contributing and appropriate for hedonic products, but less or not for utilitarian products.
Hence, humor for more high involvement products seems most appropriate at first sight for more hedonic products, however proof for the role of humor within advertising for more utilitarian products is a little more turbid.
In the following parts of this theoretical review, the main ingredients and subjects of this research will be described on results from past research and implications for this research. In the end, the literature review will finish off by stating the hypotheses being subject within this research.
2.2 Humor ads and contextual effects
Hereafter the first two important elements of this research, humor and contextual effects, are being featured.
Different academics, scientists and philosophers from the past have tried to capture and describe humor. But what is humor really? Artistotle and Plato in their time have tried to describe why people are becoming amused and what humor is. Furthermore, there are different writers that have tried to describe humor in theories such as Eliot and Baudelaire (Ziv, 1988). There have also been newer and more original theories developed on the topic of humor throughout the years, such as the humor taxonomy by Speck (1987) and others (Attardo, 1994; Raskin, 1985). However, as Kellaris and Cline (2007) also stated in their article on humor and ad memorability: "a clear understanding of the dynamics of humor in general, let alone humor's effects in advertising, has yet to emerge".
To begin with Plato, he described humor as something that is funny and 'ridicule' because of the weak people that are not able to retaliate when they are being ridiculed. Aristotle later discussed the important fundaments underlying humor, and he suggested that for humor to persist, one should be able to observe "ugliness without becoming disgusted". Both descriptions are based on the explanation that it is funny when people are misfortunate, probably because seeing other people as misfortunate lifts the person's own superiority. These theories are also known as Superiority Theory.
There are several other theories on humor produced later on, such as the General Theory of Verbal Humor (Raskin and Attardo, 1991) on the expression of incongruities between semantic scripts in verbal humor, or the Ontic-Epistemic Theory of Humor (Marteinson, 2006) mainly focusing on the role of laughter as a reaction to a cognitive impasse, or more recently the Benign Violation Theory (McGraw and Warren, 2010) stating that everything that conflicts with the established thoughts of people about how the world should be, without becoming dangerous, can be humorous.
However, there are three main theories or processes of humor which have been based on the works of, among others, Monro (1951), Keith-Spiegel (1972), McGhee (1979), and Morreall (1983), namely Arousal-Safety, Incongruity-Resolution, and Humorous Disparagement. All these three theories will be explained hereafter.
2.2.1.1 Arousal-Safety
To give an example on this, consider a movie where the main character is presumably killed in action, which leads to triggering of arousal and potential affective uncertainty expressed for example in fear ("what will happen now that the hero from the movie is killed?"). In a next shot, the main character appears not to be deadly shot, but only wounded. This offers the viewers an safety judgment, which will probably even make them smile ("thank God, he's still alive!"). Arousal-Safety requires four main elements (Speck, 1967):
1. arousal related to the discomforting presence, behavior, intention, or fate of another person 2. affective uncertainty
3. a play signal 4. a safety judgment
2.2.1.2 Incongruity-Resolution
Incongruity can occur if two or more elements in a stimulus (for example an ad) cannot be assimilated by usage of a single processing schema, or when the stimulus as a whole is not congruent with one's expectations (Speck, 1967). Different researchers suggest that humor can only occur when (at least partially) a resolution for an incongruent situation will take place (Shultz, 1972; Rothbart and Pien, 1977). Shultz (1976) even states that 'resolution of incongruity is necessary to distinguish humor from nonsense'. With Incongruity-Resolution, a situation is created which seems incongruent to the recipient, after having acknowledged this, the humor will be the result of the resolution of this incongruence.
Think for example of an advertisement in which you see a monk standing beside a football field swearing toward the referee that he's the worst referee ever seen. This leads to an incongruent situation, monks beside a football pitch is one incongruent aspect, but him swearing and offending a referee really leads to a situation of incongruity. Next shot you see the monk leaving the football field, going to his car and throwing off his monk suit, turning him into a 'normal' everyday man. Then the text 'Van Halen, for all your costumes: no matter the purpose' appears on the screen, providing the resolution.
Incongruity-Resolution requires four main elements (Speck, 1967): 1. an initial play manipulation
2.2.1.3 Humorous Disparagement
Disparagement is a concept in humor which has been described earlier in the works of Aristotle (ao. Monro, 1951; Keith-Spiegel, 1972; Morreall, 1983) and is mainly focused on the social role of humor, to keep control and censure (Speck, 1967). Disparagement humor is based always on a relationship between three persons: 1. joke-teller, 2. joke-hearer, 3. victim. Hereby a joke-teller attacks a victim, when he's present or not, where the hearer has to form a reaction on this attack. If the joke-hearer laughs, he agrees with the joke.
Furthermore, disparagement humor can come in different ways (Speck, 1967), namely: 1. Satire: if the joke-teller's intent is mainly didactic
2. Put-down humor: if joke-teller's intent is only to embarrass the victim in front of others 3. Sarcasm: if joke-teller uses put-down humor but doesn't care about reactions of joke-hearers 4. Self-deprecating humor: joke-teller makes fun of himself
To give an example of humorous disparagement, think of an advertisement where a new BMW car is stopping over at a traffic light next to a smaller, unknown other car. Then a pretty woman passes by that waves at the quite handsome man in the smaller car, which makes him think that she might be interested to step in his car. The next moment she walks past him, goes to the BMW car and steps into that car, together leaving when the traffic lights hit green. The man in the smaller car is left behind disappointed, when the text scrolls through the screen: "A man needs more than only a pretty face; BMW can give you that". This situation clearly shows an example of someone being victimized (the man in the smaller car) by the story-teller (the advertiser) which is appreciated by the story-hearer (recipients of the advertisement) in a funny way.
Humorous Disparagement requires the following four main elements (Speck, 1967): 1. a play manipulation
2. arousal associated with the disparagement 3. uncertainty regarding one's response
4. elements of incongruity-resolution facilitating the necessary 'misattribution'
In summary, humor and the above mentioned main humor processes tend to have much the same elements which build up these processes, such as play manipulation, arousal and tension. Put shortly, humor can function as a mechanism that makes people feel safe again, for redefining categories, or for enforcing social boundaries (Speck, 1967).
2.2.2 Humor ads and context
to influence consumers reactions and feelings towards an advertisement. Contextual cueing stems from the field of psychology and is referring to the different ways the human brain can gather information based on visual elements and its environments (Chun and Jiang, 1998). A cue from a context, and the context in total, will be assessed by consumers in order to determine its appropriateness.
Meyers-Levy and Tybout (1997) have explored the role of context at encoding and judgment in consumption surroundings. Past literature on the subject of context has suggested that consumers most of the times assimilate their responses toward the descriptive implications associated with a contextual cue (Allen and Janiszewski, 1989; Herr, 1989; Mathur and Chattopadhyay, 1991; Meyers-Levy and Sternthal, 1993; Shumann and Thorson, 1990; Shimp, Stuart, and Engle, 1991). However, other research has provided proof for a contrast effect among consumers (Herr, 1989; Kamins, Marks, and Skinner, 1991; Lynch, Chakravarti, and Mitra, 1991; Meyers-Levy and Sternthal, 1993) where they take distance from the contextual influences.
Martin (1990) has done research on the topic of assimilation and contrast effects, thereby finding that two factors are important in determining if assimilation or contrast effects will persist, namely 1. appropriateness of the context with regard to the target object and 2. the cognitive resources devoted by consumers to form a response on the context.
If the contextual influence is assessed by consumers and perceived to be an appropriate basis for forming a reaction, then assimilation will persist. However, when the contextual influence is perceived to be an inappropriate basis for forming a reaction, then contrast effect will persist. Important to mention here is that a negative effect resulting from an inappropriate context can only occur when recipients have enough motivation to process the message. If they have sufficient motivation, they can partial out the inappropriate information, overcorrect for it, eventually leading to a less positive judgment.
effect and judgment towards the advertisement (Meyers-Levy and Tybout, 1997). Since the topic of research here concerns high involvement goods, it seems logical to expect that consumers buying these high involvement goods have sufficient motivation to process the product, message and context and partial out possible inappropriate information.
Research of Raghunathan and Irwin (2001) building on the findings of the past, for example the works of Meyers-Levy and Tybout (1997), on contextual influences on target objects, has provided deviating results on the influence of context on a target object. They found that when context and target match with each other (comparable with assimilation effects mentioned by Meyers-Levy and Tybout (1997)), there is a domain match and a pleasant context will lead to lower target evaluation because of less happiness with the target object (product). However, when context and target do not match with each other (comparable with contrast effect mentioned by Meyers-Levy and Tybout (1997)), there is a domain mismatch and a pleasant context will lead to higher target evaluation because of greater happiness with the target object (product).
Said bluntly, Raghunathan and Irwin (2001) found that when context and target product do not match and thus seem inappropriate, a more pleasant context will result in higher target product evaluations, whereas with a match between context and target object (thus context and target object do seem appropriate), a less pleasant context will result in higher product evaluations.
Though, because Raghunathan and Irwin (2001) base their conclusions on the influence of four product context characteristics on happiness with a product and thus seem to build further on the findings from Meyers-Levy and Tybout (1997) instead of providing an alternative for it, the research finding from Meyers-Levy and Tybout (1997) will be the focal point in this thesis. However, it will be interesting to keep the findings of Raghunathan and Irwin (2001) in mind with the assessment of contextual cues and its influence on the judgment of recipients with regard to the message and its context.
contextual information, leading to a less positive judgment and contrast effects. Also the opposite, an assimilation effect, can be expected in other situations where the contextual information seems more appropriate. Based on the product nature (hedonic vs. utilitarian), which will be treated later on in this theoretical review, different effects can be expected.
2.3 Product involvement and nature
After the enlightenment above about the two elements of humor and context, hereafter the two main remaining elements of product involvement and product nature are being treated and looked upon. First an elaboration of product involvement and product nature is being done separately, whereupon these two variables are being linked together in light of this research in order to come to the hypotheses underlying the research question.
2.3.1 Product involvement
Product involvement is a concept which is widely acknowledged within the field of consumer behavior as an important aspect underlying consumers decisions. Within this thesis, the concept of product involvement is treated as a variable. Hereafter, the perspective on product involvement will be described, after that the different types of involvement, followed by a discussion on one of the most acknowledged models within the field of product involvement and message elaboration, the Elaboration Likelihood Model of Petty and Cacioppo (1981).
2.3.1.1 Perspective of product involvement
Product involvement is a widely acknowledged and important aspect underlying consumer interest and motivation when searching, using and buying certain products and services. There are different perspectives on product involvement. One is the perspective of a more product-like point of view, wherein a product is termed 'high involvement' when being more expensive mostly, and 'low involvement' when being less expensive, for example convenience goods.
the consumer. If personal relevance is low, than involvement is low, when personal relevance is high, than involvement is high.
2.3.1.2 Types of product involvement
Within the field of research of consumer behavior, research has led to a distinction between two types of product involvement, namely Situational Involvement and Enduring Involvement. Situational Involvement describes the type of product involvement which is only present in certain situations, for example the purchase of a certain product. Enduring Involvement is representing an enduring concern with a product that transcends situational influences (Houston and Rothschild, 1978; Laurent and Kapferer, 1985; Richins and Bloch, 1986; Rothschild, 1979). When taking the example of buying a car, which is mostly seen as a purchase which has some higher level of risk attached to it, consumers who normally experience a lower Enduring Involvement for cars (just because they're not interested in reading or learning about cars), will mostly express a higher level of Situational Involvement when buying a car. Consumers with higher Enduring Involvement for cars (expressed by reading, learning and talking about cars more) will also express a higher level of Situational Involvement when buying a car. So, one can be expressing lower Enduring Involvement for a certain product as such, but in a certain situation (for example a buying situation), if the perceived risk associated with this situation is high, Situational Involvement will also be high.
2.3.1.3 Elaboration Likelihood Model
A model strongly related to advertising and involvement of recipients is the Elaboration Likelihood Model (ELM) of Petty and Cacioppo (1981). In the ELM the level of involvement of consumers with a certain product and situation is connected to how certain individuals are processing cognitive information through one of two possible routes to persuasion. The ELM states that to persuade individuals with a certain message, different routes of persuasion work better, depending on whether the elaboration likelihood of the communication situation is high or low. Put in other words, if an individual is motivated (involved) with a product or situation, this person will more carefully elaborate on the arguments provided in the communication message that comes with the product or situation. If an individual is less motivated (involved) with a product or situation, this person will not evoke active thinking about the arguments or attributes in the communication message and will just accept or reject the message.
To form a brand attitude via the central route of processing requires message comprehension, learning and retention. The peripheral route associates positive or negative cues with the brand as a result of the execution or message without engaging in any extensive benefit-related thinking (Percy and Elliott, 2009). Thus, when involvement is high among consumers, the elaboration is also high; when elaboration is high, the central route to persuasion is likely to be followed. When involvement is low, the elaboration is also low, leading to consumers following the peripheral route. However, the peripheral route may also lead to persuasion by using a certain principle or decision-rule as their main guidance to persuasion. Another important prerequisite of processing the message and persuading individuals is the ability of consumers; the ability to process, which has to do with distraction, repetition, prior knowledge, message comprehensibility.
If individuals have the ability to process the persuasive communication in a right way, combined with the right motivation or involvement to process it, the central route to persuasion will be followed, which can lead to changes in attitudes because of elaboration and adopting the persuasive communication. As already mentioned, when individuals do not have the ability to process the message, and they lack enough motivation and involvement, they will take the peripheral route.
2.3.2 Product nature
Having described the principle of product involvement, hereafter the concepts of hedonic and utilitarian products will be assessed.
2.3.2.1 Hedonic products
2.3.2.2 Utilitarian products
Utilitarian goods or products are mainly bought for informational reasons, including instrumental and utilitarian reasons (Batra and Ahtola, 1990), thereby being described as primarily instrumental and functional (Hirschmann and Holbrook, 1982; Strahilevitz and Myers, 1998; Dhar and Wertenbroch, 2000). The main characteristic of pure utilitarian goods seems to be its functionality, to put it simple, a utilitarian product just has to perform the functions as can be expected from it. There are different utilitarian products which also have a certain level of hedonism associated with it. Reasons for associating hedonic cues might lie in the strive of producers of utilitarian products to distinguish themselves of other available utilitarian products in the market and hook in on a positive feeling consumers mostly bring about when being confronted with more hedonic products.
2.3.3 High involvement and utilitarian products
Starting off with utilitarian high involvement goods, one can expect, based first of all on the description of utilitarian goods, being "more instrumental and functional goods which are bought for more informational reasons" (Batra and Ahtola, 1990; Dhar and Wertenbroch, 2000; Hirschman and Holbrook, 1982; Strahilevitz and Myers, 1998), that a combination of humor and utilitarian products seems an inappropriate contextual cue (Meyers-Levy and Tybout, 1997). Taking this assumption as starting-point, one can further assume that consumers buying high involvement products (which is the situation under research in this thesis) have sufficient motivation to process the message and partial out the inappropriate information, thereby leading to a less positive judgment of the target object. Congruent with this line of reasoning, it is expected that using humor in advertising for high involvement utilitarian products will lead to a contrast effect, hence leading to the expectation that using humor as a contextual cue and associating it with a high involvement utilitarian product will lead to a less positive product judgment.
In search of further evidence underpinning this expectation, the research of Spotts et al (1997) on the usage in practice of humor within advertising came up with the low percentage of only 23% of all advertisements being used for high involvement utilitarian goods having used humor in it. This percentage is enormously lower than the level of humor used in advertisements for high involvement hedonic goods, with 62%. In every way, it seems that practitioners thus aren't too fond of using humor for this type of products either.
product is being preferred by consumers above a (moderately) incongruent alternative. Put in other words, in a high involvement situation, consumers don't like to buy something that seems incongruent with this situation. Assumption in this research is that humor and high involvement utilitarian goods seems an incongruent combination, which thus not seems to be preferred by consumers, based on Campbell and Goldstein (2001).
Batra and Stephens (1994) are debating against overall usage of humor within advertising in a high involvement situation. They performed research on affective responses to advertisements, thereby concluding that the evoking of affective responses to advertisements (e.g. evoking pleasure by using humor in advertisement) by itself may be insufficient for attitude changes in conditions where consumers are highly motivated (= high involved) to make careful, considered brand choices. This research thus doesn't make a distinction between the nature of products, being hedonic or utilitarian of nature, but it can be interpreted as advocating in general against the usage of humor in high involvement buying situations.
Somewhat earlier however, Johar and Sirgy (1994) have looked upon the differences in using utilitarian and value-expressive (e.g. humor) appeals for hedonic and utilitarian products with regard to its effectiveness. It came out that for utilitarian products the usage of utilitarian instead of more value-expressive appeals, like humor, was far more effective. These conclusions also advocate against the usage of humor for more high involvement utilitarian products.
Taking the Elaboration Likelihood Model (ELM) of Petty and Cacioppo (1981) into account, expectations described above that for high involvement utilitarian goods the usage of humor, which is more a peripheral cue, will not be appropriate and sufficient and will lead to less positive product judgments and evaluations is affined. Respondents are expected to follow the central route to persuasion, because they're high involved, and only usage of humor with the product then seems insufficient for enduring persuasion and will lead to less positive product evaluations. The variable of product nature, here utilitarian, influences if a contrast or assimilation effect (Meyers-Levy and Tybout, 1997) will persist. Expectancy with utilitarian high involvement products is that contrast will dominate.
All of the above leads to the following hypothesis with regard to high involvement utilitarian products:
2.3.4 High involvement and hedonic products
Taking this elaboration on theory further to hedonic high involvement goods, expectations are that a combination of this product type with humor seems to be an appropriate match (Meyers-Levy and Tybout, 1997), which will lead to an assimilation effect and more positive judgment of the target object (product). First of all, this expectation is based upon the description of hedonic goods, being "goods that provide more experiential consumption, fun, pleasure, and excitement which are purchased for consummatory, affective gratification" (Batra and Ahtola, 1990; Dhar and Wertenbroch, 2000; Hirschman and Holbrook, 1982; Strahilevitz and Myers, 1998).
Taking this as a basis, the search for further evidence leads again to the research of Spotts et al (1997) which proves that the usage of humor is considerably higher for high involvement hedonic than high involvement utilitarian goods, with a percentage of 62%. This has also arisen from the meta-analysis on humor in advertising performed by Eisend (2009), who concluded, based on past research which has been performed, that humorous advertising was significantly appropriate for high involvement hedonic goods.
Another, related research done by Scott, Klein and Bryant (1990) on patronage level for two types of events, a social event (which was determined as a more hedonic event) and a business event (which was determined as a more utilitarian event), provided proof for the usage of humor as a persuasion tool. When humor was involved in the flyers and promotional material for the social (hedonic) event, patronage levels were higher compared with the social (hedonic) event where the flyers and promotional materials did not involve humor.
Also, for the business (utilitarian) event, humor had no significant effect on the level of patronage, further directing proof towards hypothesis 1 stated above in this research that humor will not have a positive influence on more utilitarian products. The proof found for higher patronage levels for social (hedonic) events when using humor has been tested significantly and thus provides evidence that using humor for hedonic matters, in this research an event, strengthens the reactions on the target object. Transferring these outcomes for events to products, one might expect the same effects to persist.
making this statement also seem fitting for the field of humor within advertising. The works of Rossiter et al (1991) are highly valued and used in the field of advertising and academics, contributing to the importance one can attach to the statements provided. However, it deserves attention that the proof underpinning the writing of Rossiter et al (1991) are not empirically tested, but more or less a result of the extensive experiences in this field of work and research gained by the authors.
All of the above already points towards the expectation that humor for high involvement hedonic goods humor does seem quite appropriate to use in order to enhance product evaluations. The research of Johar and Sirgy (1991) on the difference in using utilitarian and value-expressive (for example humor) appeals for hedonic products, proved a far higher effectiveness when using value-expressive appeals than utilitarian appeals. This again provides some extra proof for the expectation just described. However, there also is room for counter-argumentation. Looking at the ELM of Petty and Cacioppo (1981) and recalling the consequences that flow from this model, one could say that based on the ELM, usage of humor for high involvement products seems insufficient in having consumers elaborating the communication message. As said, ELM expects that for high involvement products, respondents are expected to follow the central route to persuasion, making usage of humor seem insufficient for enduring persuasion, leading to less positive product evaluations. But, also these expected outcomes, based on the ELM, are debatable or at least can be described as offering basis for further deepening of the research. Also because the ELM lacks to provide a distinction between the product nature, being utilitarian or hedonic of nature. As said, the variable of product nature, here hedonic, influences if a contrast or assimilation effect (Meyers-Levy and Tybout, 1997) will persist. Expectancy with hedonic high involvement products is that assimilation will dominate.
All being said and described above does lead to the expectation that humor will be an appropriate combination with hedonic products, leading to a possibly better product evaluation. The fact that in this research the subject concerns high involvement hedonic and utilitarian products and the possible different effects of humor on these products, seems to provide an interesting angle of different expectations compared with the ELM which have to be investigated.
Summing all of the above up and taking the different argumentation mentioned in the available literature into account, the following hypothesis with regard to high involvement hedonic products will be tested in this research:
2.3.5 Low involvement and product nature
Next to high involvement, expectations about the role of using humor in advertisements for more low involvement products, both hedonic and utilitarian, can be made.
Based on the elaborations on available literature and research from the past, one might expect that, irrespective of the product nature being hedonic or utilitarian, consumers that are being low involved with a product, will follow the peripheral route to persuasion (Petty and Cacioppo, 1981). The characteristics of the peripheral route makes humor as a contextual cue, or peripheral cue if you like, seem sufficient in order to have people elaborate on the communication message or advertisement and be persuaded by its peripheral cues.
Hence, for low involvement product situations the following hypothesis can be formulated:
H3: in low involvement product situations the usage of humor in advertisement will have a more positive influence on product evaluation than the usage of non-humor, irrespective of the product nature
3. Methodology
3.1 Introduction
In this methodology part, the research procedure which will be followed in order to investigate the research question and the related hypotheses, all being depicted in the conceptual model which can be found in chapter 1, is being described.
3.2 Research design
3.2.1 Conclusive causal research
Research designs can be broadly classified as being exploratory or conclusive (Malhotra, 2007). Exploratory research has as its main objective to provide insights into and comprehension of a problem that is confronting the situation, while conclusive research is more designed to assist the decision makers in determining, evaluating, and selecting the best course of action in a given situation, they are conclusive in nature in that they are used as input for managerial decision making (Malhotra, 2007).
Because of the fact that in this research the objective is to test certain hypotheses and examine presumed relationships between different variables, conclusive research is the research practiced here. Within the field of research there are three basic research designs available; exploratory, descriptive and causal. In this research the objective is to determine cause and effect relationships, one independent variable (humor advertisement) is being manipulated, while the control over the other moderating variables (product involvement and product nature) is guaranteed. All of this is congruent with and leads to the conclusion that the research design here concerns conclusive causal research.
To guarantee a thorough application of a conclusive causal research, the sample used has to be large enough, and representative for the population.
3.2.2 Experiment
advertisement on product evaluation is a research which is investigating the causal relations between these dependent and independent variables, certainly considering the fact that the independent variable here, humor advertisement, is being manipulated in the research, congruent with experimental settings.
Hence, an experimental design is being applied in this research. Although there are several types of experimental designs, in this research the post-test only design is chosen while the possibility of biasing participants in the research when using pre-test post-test design was too high and the effects of using humor in advertisement on product evaluation lends itself good for post-testing only.
3.3 Participants and sample
In this research both men and women are being approached to participate. The argumentation underlying the choice for having both men and women participate in this research is based on the fact that advertisement is aimed at both genders also. Furthermore, both men and women are known with the concept of humor and are being confronted with it in their everyday lives. Bluntly said, for the purpose of this research the usage of a sample consisting of both men and women seems more contributive than choosing for any one of these genders. Snowball sampling technique will be used in order to recruit quickly a sufficient group of participant; first a group of family, friends, and acquaintances will be asked to participate in the research, whereby they are asked to forward the research link to their friends, family, and acquaintances in order to come to a broad and differentiated sample.
The aim is to compose a sample consisting of people from different age, sex, income levels, and level of education to come up with a highly differentiated sample. After having constructed this sample, participants will be randomly assigned to one of the conditions in this research which will be described later on.
Per experimental group a minimum of 20 participants is needed in order to meet academic standards on experimental research and obtain data that can be considered as reliable.
3.4 Variables
3.4.1 Description and manipulation of the variables
Within this experiment, the following variables will be involved: - Independent variable: Humor advertisement
- Dependent variable: Product evaluation
- Moderating variables: Product involvement and Product nature
All of the above mentioned variables will be described with regard to the how and why behind manipulating these variables in this research in order to test the different hypotheses.
3.4.1.1 Independent variable: Humor advertisement
The independent variable used in this experiment is humor advertisement. This advertisement will be presented as print ads to the respondents. The independent variable will be manipulated by offering different subgroups of respondents conditioned advertisements in which no humor is being used, while others are being presented conditioned advertisements in which humor will be used. This manipulation of the independent variable is in line with several other research performed in the past on the effectiveness of advertisement, like for example the work of Strick, et al (2009) on the role of humor in advertisement when merely associating it with a product, in which they manipulate the independent variable humor ad by directing some participants to the condition in which the ad contains humor and others to the condition in which the (identical) ad does not contain humor. Also the work of Cline and Kellaris (1999) on the joint impact of humor and ad strength on attitudes manipulates its variables similarly, by assigning respondents to one of two mock-up magazines; one of them containing non-humorous ads, the other containing humorous ads.
T-Test. Novel products are chosen to prevent people being biased upfront because of having already developed certain pre-set attitudes towards products or brands.
3.4.1.2 Dependent variable: Product evaluation
The dependent variable in this experiment is product evaluation. Product evaluation will be used because in this experiment it will be of most interest to see what the impact and influence will be of using humor or no humor in the ads on the evaluations of respondents concerning the products that are depicted in the ads. One might for example like the advertisement when humor is being used, resulting in positive attitude towards the ad (Aad), but of more importance for firms, marketers and producers of different goods will be what the impact of using humor in the ad on the product advertised in the ad itself will be. Likert scales, widely used rating scale that requires the respondents to indicate a degree of agreement or disagreement with each of a series of statements about the stimulus objects (Amoo and Friedman, 2000), are being used to measure the evaluations of respondents concerning the products. Product evaluation will be measured by using the scale of Duncan and Nelson (1985) for measuring Attitude toward and liking of the product/brand, containing statements like "People like myself would probably not like ..." and "I would expect that people using ... would be satisfied". Furthermore, product evaluation will be measured by adding statements about the quality of the product, derived from the scale of Peroshius and Monroe (1987) on Quality (Product). This scale contains statements like "This ... appears to be of good quality" and "This ... seems to be dependable".
All items used in order to measure product evaluation are based on a 7-point Likert scale in order to increase the possibilities for comparison and measurement.
3.4.1.3 Moderating variable: Product involvement
your evaluation of how well the product might fit your needs, the product might be adjusted". This approach outlined above is congruent with the way in which Batra and Stephens (1994) have conducted their experiment and manipulated the product involvement variable.
Secondly, the products chosen in the high and low involvement conditions will be products that are broadly acknowledged as typical for the product category to create a maximized high or low involvement product situation. For example, for high involvement products a washing machine (high involvement utilitarian) and a holiday (high involvement hedonic) is chosen. For low involvement a chocolate bar (low involvement hedonic) and toilet paper (low involvement utilitarian) is chosen.
For manipulation checks, all respondents in the experiment have to score the product they saw in their ad on level of involvement, based on the scale of Ratchford (1987) on Involvement. This scale measures the rating of respondents on their most recent choice of a brand of the product just like they had just seen in the ad, with regard to the "importance of the decision", "amount of thought required for the decision", and "amount to lose when choosing the wrong brand" on a 7-point Likert scale. A T-Test will be used to check for significant differences between the groups concerning their scores on level of involvement
3.4.1.4 Moderating variable: Product nature
The second moderating variable in this experiment is product nature, being hedonic or utilitarian. This variable will be manipulated by using typical hedonic products and assign these to participants in a hedonic condition, and assign typical utilitarian products to participants in a utilitarian condition (which products are being used is already described above).
A manipulation check will be performed by using the scale of Batra and Ahtola (1991) on Hedonic and Utilitarian consumer attitudes. Respondents are asked to score the product they have seen in the ad on 8 items, four hedonic items (Pleasant-Unpleasant/Nice-Awful/Agreeable-Disagreeable/Happy-Sad) and four utilitarian items (Useful-Useless/Valuable-Worthless/Beneficial-Harmful/Wise-Foolish), with a 7-point Likert scale. A T-Test will be done to see if the groups with hedonic products significantly score higher on the hedonic parts in comparison to the groups with utilitarian products and vice versa.
3.4.2 Experimental design
High Involvement
Low Involvement
Hedonic Utilitarian Hedonic Utilitarian
Humor advertisement
Condition 1 Condition2 Condition3 Condition4
Non-humor advertisement Condition 5 Condition 6 Condition 7 Condition 8 Figure 2: Experimental design
3.5 Questionnaire and procedure
Participants in this research will be sent a hyperlink which randomizes the eight different questionnaires which were made for every condition by email. By using this randomization, every respondent has an equal chance to be assigned to one of every eight conditions. The questionnaire starts with a neutral introduction on the experiment (however not explaining the true purposes of the research) and an explanation of the procedure.
The core of the questionnaire contains one advertisements of a product (high or low involvement, hedonic or utilitarian, with humor or without humor, depending on the assigned condition). After been shown the ad, respondents will be shown a page with different questions about the advertisement, which they are asked to fill out. In the end demographic variables of the respondents are measured. After having answered the questions, respondents are automatically sending their questionnaire results back to the researcher by clicking on the 'send' button. See Appendix 1 for the used questionnaire and Appendix 2 for all ads used.
The procedure with regard to the usage and assignment of the participants will be as follows. First of all, respondents are being selected among different age groups, sexes, income levels and more to create a differentiated base sample. After the selection process, participants are being randomly assigned to one of the eight possible sample groups by using a hyperlink that will be sent to participants which automatically randomly assigns participants to one of the eight conditions. Participants are being sent the hyperlink to the questionnaire by email and are asked to thoroughly read the instructions and follow these from beginning to end. After having answered all questions, participants send their responses to the researcher by clicking on the 'send' button.
3.7 Plan of analysis
H1 H2
@
