‘THE MODERATOR EFFECT OF PRIOR
KNOWLEDGE ON TYPE OF AD AND THE
CREDIBILITY OF THE MESSAGE’
The influences of type of ad and prior knowledge on message credibility and purchase intentions
2
‘THE MODERATOR EFFECT OF PRIOR
KNOWLEDGE ON TYPE OF AD AND THE
CREDIBILITY OF THE MESSAGE’
The influences of type of ad and prior knowledge on message credibility and purchase intentions
Master thesis, Msc Marketing, specialization Marketing Management University of Groningen, Faculty of Economics and Business
22-06-2015 Stephanie Pereboom Pembastraat 20 1339 RB Almere tel.: + 31 (0)6 155 00 342 e-mail: s.a.pereboom@student.rug.nl Student number: 2534991
3
MANGEMENT SUMMARY
Humor is important in everyday life – not only with friends, family and co-workers; but nowadays also for brands to advertise and bring across the campaign message. Regardless of the advantages humorous ads have, e.g., attracting attention from customers, increasing the favorability of the brand by consumers, using humor in advertising may not be that effective at all, since it can decrease message credibility. This is important since a decrease in message credibility results in a decrease in purchase intentions. So it is plausible to assume that there is a variable that influence the relationship between a humorous ad and message credibility, in order to make the relationship positive. What is known about message credibility is that it is positively influenced by prior knowledge, but it has never been researched as a moderator between the relationship of a humorous ad and message credibility. Also, until now, prior knowledge was mainly researched based on magnitude; low, moderate, and high prior knowledge. This research finds the valence of prior knowledge, in relation to humorous ads and message credibility, more important to explore since the valence explains future decisions based on stored knowledge, instead of the size of the stored knowledge. The aim of this study is to answer the following research question: “Does consumers’ prior knowledge (positively
or negatively) impact the effect of the type of ad (humorous or serious) upon the credibility of the ad’s message and does the latter increase purchase intentions?”.
This study has researched whether (1) a humorous ad, compared to a serious ad, in fact decreases message credibility; (2) whether positive prior knowledge, compared to negative prior knowledge, increases message credibility; (3) whether a humorous ad combined with positive prior knowledge, compared to the other conditions, increases message credibility; and (4) whether message credibility indeed increases purchase intentions.
4
The results showed that (1) message credibility is higher for a humorous ad than for a serious ad; (2) positive prior knowledge results in higher message credibility than negative prior knowledge; (3) message credibility is higher for a humorous ad combined with positive prior knowledge than for the other conditions; and (4) when message credibility is high versus low, purchase intentions increases.
5 PREFACE
Against all odds, I actually liked writing my Master thesis. I think it had a lot to do with the fact that, besides the topic having my personal interest, I could do it whenever and wherever I wanted as long as I made the deadlines.
People say that you learn the most when you have to do it yourself and I think that is absolutely true. And for that, and many other supportive guidance reasons, I would like to thank my first supervisor Debra Trampe. You never explicitly told me what to do or how to do it and although I did not enjoy that back then, I do appreciate it now. That is why I would also like to thank all the people who have posted their statistical tutorials on YouTube.
As the Dutch saying goes ‘all good things come to an end’, I am glad that I had the opportunity to study for my Master and I hope that this thesis will be the end of my student-era.
But for now, for you the reader, I hope you enjoy reading my thesis.
6 TABLE OF CONTENTS MANGEMENT SUMMARY ... 3 PREFACE ... 5 ABSTRACT ... 8 INTRODUCTION ... 9 THEORATICAL BACKGROUD ... 11 Humor in Advertisements ... 11 Influencers of effectiveness ... 11
The Credibility of a Message ... 13
Consumers’ Prior Knowledge about the Brand ... 14
Message Credibility on Purchase Intentions ... 16
METHODOLOGY ... 17
Participants and design ... 17
Procedure ... 17 Stimulus material ... 18 Prior knowledge ... 18 Type of ad ... 19 Pre-test ... 21 Data inspection ... 21 Sample description ... 21 Manipulation check ... 21 Measurements ... 22
Manipulation check perceived humor ... 22
Message credibility ... 22
Manipulation check brand evaluation ... 22
Control question. ... 23 Purchase intent ... 23 Suspicion check ... 23 RESULTS ... 23 Data inspection ... 23 Data distribution ... 23
7 Validity ... 24 Category involvement ... 24 Correlations ... 25 Sample description ... 26 Manipulation check ... 26 Hypothesis testing ... 27 Message credibility ... 27 Purchase intentions ... 30 DISCUSSION ... 32
Managerial and theoretical implications ... 33
Limitations and direction for future research ... 34
LITERATURE ... 36
APPENDIX ... 41
Appendix A. Questionnaire ... 41
8 ABSTRACT
Nowadays, using humor is a common way to advertise for brands. Various studies have shown that humor may not be beneficial for brands to use since, for example, it can decrease message credibility. That leaves the question, how come so many brands do use humor in their advertisement? Therefore, this study will examine the question: “Does consumers’ prior
knowledge (positively or negatively) impact the effect of the type of ad (humorous or serious) upon the credibility of the ad’s message and does the latter increase purchase intentions?”. In
contrast to other studies, this study shows that message credibility is higher for a humorous ad than for a serious ad. Also, message credibility is higher for positive prior knowledge, than for negative prior knowledge. And when the two variables are combined, humor and positive prior knowledge, it result in the highest message credibility of all four conditions. However, when prior knowledge is negative, message credibility is higher when combined with a serious ad compared to a humorous ad. And at last, this study shows that when message credibility is high versus low, purchase intentions increase. These insights result in some important managerial implications. For example, this study shows that it depends on the prior knowledge of the brand’s customers, whether the use of humor in an advertisement results in a higher message credibility and therefore in higher purchase intentions.
KEYWORDS Humorous ads, message credibility, prior knowledge, negative, positive, moderator effect, purchase intentions, experiment. RESEARCH THEME Affective Responses to Advertising
9
INTRODUCTION
The high interest of researchers in the effects of humor in advertisements is well understood. The use of humor in advertisements has become such a common practice (Eisend, 2009) that of every five television ads there is one using humor to convey the message (Beard, 2005). Research showed that using humor in ads has advantages for the brand, namely: attracting attention from consumers (Warren and McGraw, 2013), creating awareness, enhancing the attitudes towards the ad (Eisend, 2009), reducing negative cognitions (i.e., negative thoughts; Hoyer, MacInnis, and Pieters, 2013), and increasing the favorability of the brand by consumers (Hoyer, MacInnis, and Pieters, 2013). These advantages might only take place when humor is conducted in the right way.
Besides the many advantages humorous ads have, Eisend (2009) shows that there are also disadvantages of the use of humorous ads. For example, the effect of source credibility significantly reduces when humor is used in advertisements. Another remarkable disadvantage is that a humorous ad – when compared to a serious ad – reduces the credibility of the message (Suterland, 1982). And yet the message credibility is strongly related to influencing the consumers’ beliefs (Slater and Rouner, 1996) and purchase intentions (Davis, 1994; Goldsmith, Lafferty, and Newell, 2010).
10
Until now, studies mainly researched the effects of the magnitude (low, moderate and high) of prior knowledge (Britton and Tesser, 1982; Cacioppo et al., 1992; Chebat, Limoges, and Gélinas-Chebat, 1998; Hong and Sternthal, 2010; Sharifpour and Walters, 2014). In these and in many other studies, the importance of the valence of prior knowledge is not taken into account and this might have been crucial information to include. Research (Gu et al., 2011) in neuropsychology explains why: the magnitude of prior knowledge says nothing about the valence of prior knowledge. The magnitude and valence are interrelated, but are processed differently, such that valence “stores information about the past outcomes to improve future decisions” (p.3891), while magnitude only captures the size of that information. Thus, in many cases the valence of knowledge says more about prior knowledge than the magnitude of prior knowledge. This research makes the effort to study the difference between positive and negative prior knowledge. Besides, previous studies have researched – in a small quantity – the relationship between a humorous ad and message credibility, but it has never been researched whether (the valence of) prior knowledge could have a moderating role on the effect of the relationship between the type of advertisement and the message’s credibility.
For these reasons, this current research will answer the following research question: “Does consumers’ prior knowledge (positively or negatively) impact the effect of the type of
ad (humorous or serious) upon the credibility of the ad’s message and does the latter increase purchase intentions?”.
This research will make an academic contribution by researching a moderating relationship that has never been researched and may explain why the relationship between a humorous ad and message credibility – that initially is proven to be negative – can be positive. For marketing practices this research is relevant since it provides insights about possible advertising tactics. Such that, when it is proven that positive prior knowledge has an (in)direct effect on message credibility and purchase intent, it is useful for a brand to adapt their marketing strategy based on that finding. In translation, a brand should be careful when using humorous advertisements when their target customers have negative prior knowledge. Also, when it is confirmed that positive prior knowledge positively (vs. negatively) influences the relationship between a humorous ad and message credibility, it will indicate an important marketing tactic. Namely, that brands could try to influence purchase intentions by providing positive knowledge about the brand to their consumers.
11
participants and design, stimulus material, procedure, measurements, data reduction and pre-test. This is followed by the chapter results. At last the discussion chapter presents the theoretical and managerial implications, limitations and directions for future research.
THEORATICAL BACKGROUD
Humor in Advertisements
Looking at the definition of humor as such, Warren and McGraw (2013) define humor as arising “from the perception of something threatening or wrong, i.e., a violation” (p.2). When looking at humor in advertising, it “includes all forms of smile-inducing stimuli incorporated into a marketing communication for the purpose of achieving stated promotional objectives” (Duncan, 1980, p.302). This emotion can be evoked in various ways, such as by “puns, understatements, jokes, ludicrous situations, satire, and irony” (Sternthal and Craig, 1973, p.13), but it is the violation that plays a key role in humor. Ads that illustrate a violation generate greater involvement with the ad’s message and likeability (Brown, Bhadury, and Pope, 2010). An ad that is liked for its humorous message results in liking the product or brand (Slater and Rouner, 1998). Brown, Bhadury, and Pope (2010) show that consumers find ads humorous when the ad contains a surprise that is followed by a humorous response, e.g., benign violations. Although this may create buzz and pass on rates (Brown, Bhadury, and Pope, 2010), Sternhal and Craig (1973) show that only focusing on humor in the ad will cause consumers to only remember the ad and not the brand. Indeed, as confirmed by Krishnan and Chakravarti (2003), humorous ads are recalled more than non-humorous ads. In order for consumers to not only remember the ad, but also the brand, it is important for brand to make use of strong humor, i.e., “humor that is meaningfully linked to the brand claims” (Krishnan and Chakravarti, 2003, p.230) and to apply the humor in the ad to the offering or product (Sternthal and Craig, 1973). Besides, humor tends to be more suitable for low-involvement offerings in which the creation of positive feelings between the consumer and the ad is essential (Spotts, Weinberger, and Parsons, 1997) and when consumers have a low need for cognition (Zhang, 1996).
12
effect may be taken out of its context since men prefer and appeal more to aggressive and sexual humor, than any other group, and this type of humor is most often used in advertisements (Whipple and Courtney, 1981). Also, generally speaking, consumers who are open to be entertained, i.e., “have a need for humor”, or already have a positive attitude toward the brand that advertises (Cline, Altsech and Kellaris, 2003, p.34), tend to respond more favorably towards humorous ads. This will positively contribute to the effectiveness of the ad. Another factor that can influence the effectiveness of humorous ads is the type of medium used. Different types of media use humor, but it is more effective when it is used by media that can ‘bring the humor alive’, i.e., expressiveness (Hoyer, MacInnis and Pieters, 2013), such as a commercial on the television.
13
looked at the magnitude of prior knowledge and not of the valence of it, this research assumes that is it likely that the prior knowledge that the consumers had in Teixeira and Stipp (2013) research was positive instead of negative. This is expected because high knowledge was positively related to purchase intent and the chances are very small that when a respondent has high, mainly negative prior knowledge, he or she would have high purchase intentions. Since this assumption cannot be confirmed, it is therefore all the more important that this research does make the distinction between positive and negative prior knowledge.
The Credibility of a Message
14
sources are perceived as more credible and therefore have a higher influence on consumers decision making (Hoyer, MacInnis, and Pieters, 2013). Not only the use of non-marketing sources increases the speed in which the message is accepted and believed by consumers, but also the use of an expert, a trustworthy or knowledgeable person of a given topic, in an ad, has this effect (Sternthal, Dholakia, and Leavitt, 1978; Priester and Petty, 2003). This is the case since consumers believe these people have unbiased opinions (Hoyer, MacInnis, and Pieters, 2013).
Going back to the beginning of this subparagraph, humor can affect credibility. Sutherland (1982) researched the effects of the type of ad (serious or humorous) on perceived credibility. Sutherland (1982) let the respondents listen to a serious commercial and a humorous commercial, but varied the order in which they heard both ads. One group first heard the serious and then the humorous commercial and the other group first heard the humorous and then the serious commercial. After both commercials were heard, the respondents filled in a questionnaire. The results of that study show that respondents found the serious commercial more credible and believed it had more authority than the humorous commercial, regardless of the order that the ads were heard. This same effect is shown by another research preformed by Bryant, Brown, Silberberg, and Elliott (1981). Their study included students who viewed college textbooks that contained illustrations with humor. Likewise, the results showed that there is a negative relation between humorous ads and persuasiveness and this effect is mainly caused by a decrease in credibility. So, in line with these results, this research hypothesizes:
Hypothesis 1. Message credibility is higher for a serious ad compared to a humorous ad.
Consumers’ Prior Knowledge about the Brand
15
advertizes, consumers base their attitudes towards the brand on the ad that is presented by the brand. For this reason, research (Machleit and Wilson 1988; Machleit, Allen, and Madden 1993; Campbell and Keller, 2003) suggests that high prior knowledge leads to a greater difference in attitudes toward the ad and the attitude toward the brand compared to having low prior knowledge.
As mentioned before, prior knowledge can differ in magnitude, but it can also differ in valence. This latter will be the focus of this research since I want to know what the influences are of prior knowledge with regards to past outcomes and future decisions (Gu et al., 2011), rather than the size of prior knowledge.
Klebba and Unger (1983) have researched the effects of positive and negative prior knowledge on credibility in a field setting about cars. They asked participants several questions to determine whether the participant viewed the Mustang as more positive or negative than the Pinto. The results showed that the Mustang, on all elements, was perceived more positively than the Pinto. These outcomes were used with another group of respondents during a comparison test. The other group of respondents was asked to fill in a number of questions about Lee Iacocca and Chrysler Corporation regarding (source) credibility and advertising. The last question referred to whether the participant associated the Lee Iacocca/Chrysler with the Mustang or with the Pinto (comparison test). If the respondent compared the Lee Iacocca or Chrysler to the Pinto this would mean that that respondent would have negative prior knowledge, since the first group unanimously rated the Pinto as negative and the Mustang as positive. Besides, the results showed that participants who associated the Lee Iacocca/ Chrysler with the Pinto (i.e., had negative prior knowledge), rated the advertisement as less credible. The effect for positive prior knowledge was not significant. However, Austin and Dong (1994) demonstrate that positivity leads to higher credibility, thereof inferring that positivity also means positive prior knowledge. Therefore, it is expected:
Hypothesis 2. Message credibility is higher when people have positive prior knowledge compared to negative prior knowledge.
16
credibility. In addition to those studies, this research assumes that if a consumer evaluated a brand negatively and afterwards an ad is shown to that consumer, it is likely that the negative evaluation then accounts for negative prior knowledge in evaluation of the ad of that brand. So, assuming that in that case negative evaluations are equal to negative prior knowledge, this research predicts that negative prior knowledge strengthens the already (predicted) negative relation between a humorous ad and message credibility. Thus, making a humorous ad less credible when consumers have negative prior knowledge. The same studies (Belch and Belch, 2007; Eisend, 2011; Hoyer, MacInnis, and Pieters, 2013) demonstrate that consumers who have a positive attitude are more likely to accept the message. This research predicts the same effect as above: when a consumer has a positive attitude towards a brand it is likely that this consumer has positive prior knowledge when he or she is, later on, confronted with an ad of that brand. But in this case it is also expected that the positive prior knowledge can ‘turn’ the negative relation between the humorous ad and message credibility into a positive relation. Hence, this research expects:
Hypothesis 3. Prior knowledge moderates the relation between type of ad and message credibility, such that when prior knowledge is positive (vs. negative) the relation between a humorous ad and message credibility is positive (vs. negative).
Message Credibility on Purchase Intentions
Message credibility and purchase intentions are frequently researched (Kamins and Assael, 1984; Grewal, Kavanoor, Fern, Costley, and Barnes, 1997; Martin, Lang, and Wong, 2003; Bohner, Einwiller, Erb, and Siebler, 2003; Coulter and Punj, 2004; Eisend, 2006). Goldsmith, Lafferty, and Newell (2010) show that corporate credibility has a positive effect on purchase intent. This effect is explained by consumers who take their “perceptions of the trustworthiness and expertise of a company” into account when they judge the company’s products on quality and these judgments determine whether the consumers want to buy the product or not (p.46). In addition, in earlier research of Davis (1994), he already found evidence that consumers were positively influenced by the “good citizenship” of a company on their purchase intentions. So in line with these results, this research hypothesizes a similar effect for message credibility:
17
In order to test whether the hypotheses can be supported, the following conceptual model (figure 1) is designed: FIGURE 1 Conceptual model H1 H4 H3 H2 METHODOLOGY
Participants and design
In total, 106 people participated in the experiment, most of them were women (59.4%), had an age between 18 and 29 (57.5%), and an annual income lower than €30,000 (61.3%). They completed a questionnaire, in Dutch, that was spread via the e-mail. Participants were randomly assigned to a 2 (type of ad: humorous vs. serious) x 2 (prior knowledge: positive vs. negative) between-participants, factorial design.
Procedure
A message was spread via the internet (e.g., Facebook, LinkedIn, and e-mail) asking for participants. It said that when people were interested to participate and expected not be distracted within the coming minutes, they could send their e-mail address and the link with the questionnaire was immediately sent to them. Participants were assigned to a condition (based on the time of sending an e-mail saying that they wanted to participate) and exposed to the manipulated independent variables: prior knowledge (either positive or negative) and type of ad (either humorous or serious) of a non-existing brand for a low-involvement offering. This procedure was chosen so people could (1) be randomly assigned to an experiment group and (2) fill-in the questionnaire at a moment that they choose to fill it in and was a moment without distraction. When participants received the questionnaire they were not told what the study was about, only that it was performed for writing a Master thesis. Also, it was told that
Type of ad
serious/humorous
Prior knowledge
positive/negative
18
the questionnaire was completed anonymously and that the duration of filling in the questionnaire was about five minutes.
Half of the participants were given positive prior knowledge and the other half was given negative prior knowledge. This was done by giving participants some information about the brand Tarex (i.e., prior knowledge) and it was told that they should read the information carefully since they could not go back and read the text again. After that, participants were shown three images of the product Tarex, but the color of the product varied (this was changed by the researcher in PhotoShop). The product was either blue, red or green (see figure 2). The participants were asked which product color had their preference. This ‘assignment’ had no function for this research, but served as an unrelated filler task, to mislead participants in guessing what the subject of the research was. Thereafter, participants were shown one of two versions of the advertisement of Tarex (humorous vs. serious) and it was told that the ad they were seeing was the one which was currently being used by Tarex to advertise with. Likewise, the participants were told to view the ad carefully since they would answer some questions afterwards and they were unable to view the advertisement again. They first answered some descriptive questions and after that the research questions. When all twenty questions were filled-in, the participant was thanked for their cooperation.
FIGURE 2 Unrelated filler task
Option 1 Option 2 Option 3
Stimulus material
19
group received information about the brand that was positive and the other group received information that was negative (see table 1). In the previous chapter of this research is stated that other research has found information that can serve as prior knowledge, e.g., a company’s reputation (Goldberg and Hartwick, 1990). Therefore, this research has used that information in the creation and manipulation of negative and positive prior knowledge.
TABLE 1
Manipulating prior knowledge
Positive prior knowledge Negative prior knowledge In order for you to able to answer the
following questions, I will tell you something about the brand, Tarex. Tarex is a German brand that entered the German market three years ago (in 2012) and is now considering expanding to the Dutch market. In Germany Tarex is known for its mouthwash and also for its fair way of treating its customers. This can partly be explained by the satisfied or reimbursed policy that Tarex is following. Independent research has shown that customers rate Tarex, number 1 in the product category of mouthwash. Besides, in Germany Tarex is winning market share very fast, due to their reliable and high quality products.
In order for you to able to answer the following questions, I will tell you something about the brand, Tarex. Tarex is a German brand that entered the German market three years ago (in 2012) and is now considering expanding to the Dutch market. In Germany Tarex is known for its mouthwash and also for its recently received bad publicity concerning their poor way of treating their customers. Independent research tested the products of Tarex and the results were extremely disappointing. Not only did Tarex score low on reliability, their product also scored low on quality. Therefore, in Germany, Tarex is now losing market share very fast.
20 FIGURE 3 Humorous ad
21 Pre-test
Data inspection. The two advertisements were immediately tested to determine whether the humorous ad, that was used in the experiment, was really perceived as humorous and the serious ad as serious. There were no missing data in the pre-test or were there outliers. To test the data distribution, a Shapiro-Wilk test (Shapiro and Wilk, 1965; Razali and Wah, 2011) is performed with total perceived humor as dependent variable. The null hypothesis is that the data are normally distributed. The results show that the null hypothesis can be rejected, p = .00, this means that the data of the total perceived humor are non-normally distributed. When the data is non-normally distributed it means that no parametric tests (e.g., t-test, ANOVA) can be performed, since these tests assume a normal distribution. Therefore, when a parametric test in this research is performed on data that is non-normally distributed, a non-parametric test (e.g., Kruskal-Wallis, Mann-Whitney) will follow to see whether the outcomes differ.
Furthermore, the reliability of the measures of the variable total perceived humor was tested with the Cronbach’s alpha and showed that the measures are related and consistent with each other (α = .98 > critical value of .60; Reynaldo and Santos, 1999).
Sample description. The pre-test was done among forty people with similar characteristics as the participants who participated in the experiment (see table 2). Three one-way ANOVA’s with as independent variable gender, income or age and the type of ad (humorous or serious) as dependent variable, shows that due to the random assigning of participants, there was no difference between participants in each condition on the descriptive variables; gender, F = .40, p = .53, income, F = .47, p = .76, and age, F = .36, p = .87.
TABLE 2
Descriptive variables of the participants
N Gender Age Income
Pre-test 40 female (60%) 18-29 (50%) < €30.000 (45%) Experiment 106 female (59%) 18-29 (58%) < €30.000 (61%) Total 146 female (60%) 18-29 (55%) < €30.000 (57%)
22
1.51), t(39) = 19.52, p = .00. Since the data is non-normally distributed, a Mann-Whitney test is performed with type of ad as the independent variable and the total perceived humor as dependent variable. The results show that the humorous ad was perceived as more humorous (mean rank = 30.50) than the serious ad (mean rank = 10.50), U = .00, p = .00. Therefore, both ads were used in the experiment.
Measurements
The experiment was divided into seven parts (see Appendix B). First, it measured the demographic variables which were age, gender and income. These variables were taken from Ou et al. (2014).
Manipulation check perceived humor. The second part contained the questions that measured the total perceived humor (α = .88), referring to the independent variable, type of ad. This included six questions measuring how humorous, interesting, enjoyable, funny,
serious and dull the advertisement was. All these questions were measured on a seven-point
semantic differential scale, varying from, for example, “not humorous at all” to “extremely humorous” (scale adapted from Bryant, Brown, Silberberg, and Elliot, 1981; Zhang, 1996). Since not all the questions had the same scale, they were adjusted to the example given above. This was done so the participant might not get the feeling that one question was more important than another, due to a difference in scale. Also the question about the seriousness of the advertisement was developed by the researcher of this study. Both the serious and dull scale needed to be reverse coded in the data set, so that all the high values represented humorous perceptions of the ad and the low scores represented the serious perceptions of the ad.
Message credibility. The third part measured the credibility of the message (α= .92) with the following two questions: (1) How credible was the advertising? (“not credible at all” – “extremely credible”); and (2) Compared to most commercials you see, did you find the Tarex advertisement? (“deceptive” – “honest”), (“misleading” – “sincere”) (scale adapted from Bryant, Brown, Silberberg, and Elliot, 1981; Goldberg and Hartwick, 1990).
23
– “good”), (“dislike” – “like”), and (“not nice” – “nice”) (scale adapted from Hong and Sternthal, 2010). Since the participants’ evaluation of the brand depended on the information they were given during the experiment, these questions were used to measure prior knowledge.
Control question. To research the participant’s involvement with the product category, the fifth part involved the question: How likely are you to purchase mouthwash? (“unlikely” – “likely”) (scale adapted from Teixeira and Stipp, 2010).
Purchase intent. The sixth part measured the participant’s purchase intention, with the question: How likely are you to purchase this brand Tarex? (“unlikely” – “likely”) (scale adapted from Teixeira and Stipp, 2010).
Suspicion check. The final question concerned a suspicion check and asked whether the participant had an idea what the study was about. Not one participant guessed the purpose of the study and therefore, no data needed to be excluded from the research.
RESULTS Data inspection
Data distribution. The data showed no missing values or outliers. To test the data distribution, another Shapiro-Wilk test (Shapiro and Wilk, 1965; Razali and Wah, 2011) is performed with total perceived humor, total brand evaluation, and message credibility as dependent variable. Again, the null hypothesis is that the data are normally distributed. For total perceived humor the null hypothesis can be rejected, p = .01. For total brand evaluation the null hypothesis cannot be rejected, p = .06, as well as for message credibility, p = .10. This means that the data for the total perceived amount of humor are non-normally distributed, but the data for the total brand evaluation and message credibility are. As explained before, the implications of the non-normally distributed data means that when a parametric test is performed a non-parametric test will follow to see whether the distribution made a difference in the outcomes. In this research that will be done when tests include the variable total perceived humor.
24
Therefore, a new variable was computed which indicated whether the participant did or did not tell their annual income. An one-way ANOVA with the new variable (told income: yes or no) as independent variable and total perceived humor, total brand evaluation, and message credibility as dependent variables, showed that the two groups did not significantly differed on total perceived humor, F(1,105) = 2.36, p = .13, total brand evaluation F(1,105) = .38, p = .54 or message credibility F(1,105) = .06, p = .82. These results were confirmed with the Mann-Whitney test with the new variable (told income: yes or no) as independent variable and total perceived humor, total brand evaluation, and message credibility as the dependent variables, total perceived humor, U = 390.50, p = .17, total brand evaluation, U = 474.50, p = .62 or message credibility U = 514.50, p = .93. Yet, the option was reported in the file as missing data.
Validity. The validity of the measures is tested with the factor analysis. The factor analysis is appropriate since the Kaiser-Meyer-Olkin Measure of Sampling Adequacy is .76 for total perceived humor, .81 for total brand evaluation, and .72 for message credibility (>.5), the Bartlett’s test of sphericity is significant for all three variables (p = .00) and the communalities are all higher than .40. If the factor analysis exceeds the critical value of .50 it means that that variable explains the component. The three measures of message credibility all explain the same component: ‘credible’ (.89), ‘honest’ (.96), and ‘sincere’ (.94).
Category involvement. The mean of category involvement was calculated for all four conditions. As table 3 shows, the mean of category involvement was the highest for the condition where participants saw the serious ad and read the positive prior knowledge and the lowest for participants who saw the serious ad and read the negative prior knowledge.
TABLE 3
Mean category involvement among the four conditions Mean Std. Deviation
Humorous & positive 3.70 2.07
Serious & positive 4.00 1.30
Humorous & negative 3.44 1.87
Serious & negative 3.11 1.55
25
Correlations. To ensure that the descriptive and control variables: age, gender, income, and category involvement do not affect the independent variables: type of ad and prior knowledge or the dependent variables: message credibility and purchase intentions this was tested with a Spearman correlation (table 4). The Spearman was chosen because, except for message credibility, all data are categorical. As the results showed, age (type of ad p = .31, prior knowledge p = .79, message credibility p = .87; purchase intentions p = .83), gender (type of ad p = .44, prior knowledge p = .92, message credibility p = .20; purchase intentions
p = .91) and income (type of ad p = .74, prior knowledge p = .59, message credibility p = .58;
purchase intentions p = .54) are not significantly different from zero (p > 0) on the (in)dependent variables. This same effect accounts for category involvement and the independent variables (type of ad p = .12, prior knowledge p = .16). There is a small correlation between category involvement and message credibility (correlation = .21, p = .03) and between category involvement and purchase intentions (correlation = .50, p = 00).
TABLE 4
Correlations and (p-values) between descriptive/control variables and (in)dependent variables
Age Gender Income Category involvement
Type of ad .10 (.31) -.07 (.44) -.03 (.74) .15 (.12) Prior knowledge -.03 (.79) .01 (.92) -.05 (.59) .14 (.16) Message credibility -.02 (.87) -.13 (.20) -.54 (.58) .21 (.03) Purchase intentions .02 (.83) .01 (.91) -.06 (.54) .50 (.00)
26
p = .00, but it showed that (on average) the higher the participants’ category involvement, the
higher the participants’ purchase intentions. So, it was not necessary to control for category involvement.
Sample description
As previously mentioned, a total of 106 people participated in the experiment. These participants had similar demographic characteristics as the participants in the pre-test (see table 2, page 21). During the experiment participants were randomly assigned to one of the four conditions. Three one-way ANOVA’s with as independent variable gender, income or age and the four conditions (type of ad*prior knowledge) as dependent variable, shows that due to the random assigning of participants, there was no difference between participants in each condition on the descriptive variables: gender (M = 1.59, SD = .49), F = .56, p = .46, income (M = 2.01, SD = 1.42), F = .73, p = .54, and age (M = 2.96, SD = 1.34), F = .53, p = .76.
Table 5 shows the descriptive statistics of the participants in each condition. It demonstrates that due to the random assigning of participants each conditions contains participants with similar demographic characteristics, with exception of the condition humorous ad and negative prior knowledge. In this conditions fewer women participated, compared to the other three groups. There is no explanation for this occurrence, other than coincidence.
TABLE 5
Descriptive statistics per condition
Condition N Gender Age Income
Humor & positive 27 female (66%) 18-29 (59%) < €30.000 (78%) Serious & positive 27 female (63%) 18-29 (55%) < €30.000 (56%) Humorous & negative 25 female (44%) 18-29 (56%) < €30.000 (52%) Serious & negative 27 female (63%) 18-29 (59%) < €30.000 (59%) Total 106 female (59%) 18-29 (58%) < €30.000 (61%)
Manipulation check
27
ANOVA with type of ad (humorous or serious), prior knowledge (positive or negative), and the interaction effect of type of ad and prior knowledge as independent variables and total perceived humor as dependent variable, show that the humorous ad (M = 4.85, SD = 1.07) and the serious ad (M = 2.43, SD = .86) also significantly differ from each other in the experiment,
F(1,105) = 178.41, p = .00. Such that the humorous ad is found more humorous than the
serious ad. The results show that prior knowledge has no significant effect on the total perceived humor, F(1,105) = .13, p = .80, neither does the interaction effect of type of ad and prior knowledge on total perceived humor, F(1,105) = 1.99, p = .20. This means that besides the type of ad, prior knowledge and the interaction of type of ad and prior knowledge has no significant effect on total perceived humor.
The Mann-Whitney test with the type of ad as independent variables and the total perceived humor as dependent variable, confirms that the humorous ad and serious ad significantly differ from each other, 2(3) = 71.41, p = .00.
The two-way ANOVA with type of ad, prior knowledge, and the interaction effect of type of ad and prior knowledge as independent variables and total brand evaluation as dependent variable, show that the positive prior knowledge (M = 4.5, SD = 1.27) and negative prior knowledge (M = 2.82, SD = 1.26) significantly differ from each other in the experiment,
F(1,105) = 49.69, p = .00. Such that the positive prior knowledge results in a higher brand
evaluation than the negative prior knowledge. Type of ad has a no significant effect on the total brand evaluation, F(1,105) = .34, p = .70. Also, there is also no significant interaction effect of type of ad and prior knowledge on total brand evaluation, F(1,105) = 1.20, p = .40.
In sum, type of ad and prior knowledge are manipulated correctly and are not confounding with each other.
Hypothesis testing
Message credibility. The first test that is conducted is a two-way ANOVA with type of ad, prior knowledge, and the interaction effect of type of ad and prior knowledge as independent variables and message credibility as dependent variable (see table 6). The results of this full model show that the main effect of type of ad on message credibility is significant,
F(1,105) = 7.65, p = .01. The descriptive statistics (table 7) show that this relation is such that
28
TABLE 6
Two-way ANOVA with interaction of type of ad and prior knowledge
Df Mean Square F Sig. Partial Eta
Squared Corrected Model 3 17.70 9.03 .00 .21 Type of ad 1 14.98 7.65 .01 .07 Prior Knowledge 1 20.95 10.69 .00 .10 Type of ad * Prior knowledge 1 16.56 8.45 .00 .08 TABLE 7
Descriptive statistics of message credibility
Type of ad Prior knowledge Mean Std. Deviation N
Humorous Positive 4.83 1.11 27 Negative 3.15 1.18 25 Total 4.02 1.41 52 Serious Positive 3.28 1.54 27 Negative 3.19 1.67 27 Total 3.23 1.59 54 Total Positive 4.06 1.54 52 Negative 3.17 1.44 54 Total 3.62 1.55 106
The two-way ANOVA also shows that the main effect of prior knowledge on credibility is significant, F(1,105) = 10.69, p = .00, such that message credibility is higher for positive prior knowledge (M = 4.06, SD = 1.54) than for negative prior knowledge (M = 3.17,
SD = 1.44) (see table 7). In total, 10% of the variability in credibility is being accounted for
by prior knowledge (partial eta squared = .10). This result is in line with hypothesis 2, which says that message credibility is higher when people have positive prior knowledge, compared to negative prior knowledge.
29
Meaning that type of ad and prior knowledge strengthen each other/interact. This result is in line with hypothesis 3, which said that prior knowledge moderates the relation, such that when prior knowledge is positive (vs. negative) the relation between a humorous ad and message credibility is positive (vs. negative). In total, 8% of the variability in credibility is being accounted for by type of ad and prior knowledge (partial eta squared = .08).
When type of ad and prior knowledge are combined, it results in the amount of message credibility that is shown in the bar chart below (figure 5).
FIGURE 5
Bar chart mean message credibility
30
serious ad and positive prior knowledge (M = 1.54, p = .00), such that message credibility is, with 1.54, higher for a humorous ad in combination with positive prior knowledge, than for a serious ad in combination with positive prior knowledge. The same effect also accounts for the humorous ad in combination with positive prior knowledge and the humorous ad in combination with negative prior knowledge (M = 1.68, p = .00) and the serious ad in combination with negative prior knowledge (M = 1.64, p = .00). The conditions serious ad and positive prior knowledge (p = 1.00), humorous ad and negative prior knowledge (p = 1.00), and serious ad and negative prior knowledge (p = 1.00) do not significantly differ from each other. So, after the post hoc test, it is shown that the interaction effect only holds for a difference between the humorous ad in combination with positive prior knowledge and the other three conditions, but not for the latter three conditions among each other.
TABLE 8
Post Hoc of credibility per condition
Mean difference Sig. Humor & positive Serious & positive 1.54 .00
Humorous & negative 1.68 .00 Serious & negative 1.64 .00 Serious & positive Humor & positive -1.54 .00 Humorous & negative .14 1.00 Serious & negative .10 1.00 Humorous & negative Humor & positive -1.68 .00
Serious & positive -.14 1.00 Serious & negative -.04 1.00 Serious & negative Humor & positive -1.64 .00
Serious & positive -.10 1.00 Humorous & negative .04 1.00
31
output of the regression analysis shows that R2 = .34 (adjusted R2 = .34). Although the R square is not that high, it still means that 34% of the variability in purchase intentions, can be accounted for by message credibility. Furthermore, the output shows that message credibility and purchase intentions are correlated (Pearson Correlation = .59). As is shown in table 9, message credibility is significant (p = .00), which suggests that for each unit that message credibility increases, purchase intentions increases with .72, F(1,105) = 54.42, p = .00. Table 9 also shows that the constant is not significant (p = .25), this means that when message credibility is zero, purchase intentions is not significantly different from zero. So, in line with hypothesis 4, when message credibility is high compared to low, purchase intentions increase.
TABLE 9
Regression analysis of message credibility and purchase intentions Unstandardized Coefficients T Sig. B Std. Error Constant .46 .38 1.16 .25 Message credibility .72 .10 7.38 .00 TABLE 10
ANOVA of message credibility and purchase intentions
Df Mean Square F Sig.
Regression 1 130.79 54.42 .00
Residual 104 2.40
Total 105
32 TABLE 11
Overview of (not) supported hypotheses
Hypothesis Supported
1. Message credibility is higher for a serious ad compared to a humorous ad. Not supported 2. Message credibility is higher when people have positive prior knowledge
compared to negative prior knowledge.
Supported
3. Prior knowledge moderates the relation, such that when prior knowledge is positive (vs. negative) the relation between a humorous ad and message credibility is positive (vs. negative).
Supported
4. High (vs. low) message credibility increases purchase intentions. Supported
DISCUSSION
33
Besides, the product used in the survey is a low-involvement product (mouthwash). As Spotts, Weinberger, and Parsons (1997) explain, humor is more suitable of low-involvement offerings. A last explanation can be that the participants were open to be entertained and/or had a need for humor, since that results in more favorable responses.
In addition to existing studies (Britton and Tesser, 1982; Cacioppo et al., 1992; Chebat, Limoges, and Gélinas-Chebat, 1998; Hong and Sternthal, 2010; Sharifpour and Walters, 2014) that examined the effect of the magnitude of prior knowledge on message credibility, this research shows that the valence of prior knowledge also influences message credibility. In line with hypothesis 2, this study shows that positive prior knowledge has a significant higher effect on message credibility, when compared to negative prior knowledge.
The results of this study show that there is an interaction effect of prior knowledge on the relation between type of ad and message credibility, such that message credibility is highest for a humorous ad in combination with positive prior knowledge.
The importance of this study is shown with the last hypothesis that showed that when message credibility is high, compared to low, purchase intentions are also high. And therefore, it can be said that this study provides some additional implications to influence purchase intentions.
Managerial and theoretical implications
Theoretical implications. This study contributes to the current theory by showing that the valence of prior knowledge has an significant impact on message credibility. Subsequently, this research shows that prior knowledge plays a moderating role in the relationship between type of ad and message credibility. At last, this study contradicts current theories, by showing that humorous ads have a higher message credibility than serious ads.
Managerial implications. Based on the findings of this study, there are meaningful implications for marketing managers. Most importantly, the results show that purchase intentions can be stimulated by a high message credibility. Managers should therefore always aim for the highest message credibility they can get.
34
message credibility – and therefore lowest purchase intentions – of all four conditions measured in this paper.
Although the difference is very small, in case of a brand with customers who are known for having negative prior knowledge, that brand could better make use of a serious advertisement, since message credibility will be slightly higher. This does not mean that brands should not try first to provide positive information about their brand among their customers, so that it in a later stage can accounts for positive prior knowledge, and then use a humorous ad to advertise with.
When it is unknown whether customers have negative or positive prior knowledge or the brand is new to the market, a humorous ad also results in the highest message credibility, compared to a serious ad.
Limitations and direction for future research
A first limitation of this study can be found in the sample size. While the sample size that was used to apply the data was sufficient for this research, it should be checked whether these effects also hold among a larger sample size. Also, the sample size should include participants that better represent the ‘average’ Dutch person, since now the age and income level are lower than what is average for the Netherlands.
Another limitation of this study is that it only involved printed media. Research (Hoyer, MacInnis, and Pieters, 2013) has shown that humor is more effective when it is used by media that can bring the humor alive, for example a television commercial. So future research should focus on media that has a higher level of expressiveness and vividness than printed media, so it can determine whether the results still hold under the conditions of other types of media.
While a majority of previous research showed that message credibility is higher for a serious ad, compared to a humorous ad, this research found the opposite effect. Therefore, more research will be necessary to investigate this difference. It may be possible that serious ads will result in a higher message credibility than humorous ads when the medium type differ, or regards a different kind of product, or a different kind of message. So there are enough factors to research that might differently influence the relationship between the type of ad and message credibility.
35
36
LITERATURE
Alden, D. L., Hoyer, W. D., and Lee, C. (1993), “Identifying Global and Culture-Specific Dimensions of Humor in Advertising: A Multi-National Analysis,” Journal of
Marketing, 57 (April), 64–75.
Austin, E. W. and Dong, Q. (1994), “Source v. Content Effects on Judgments of News Believability,” Journalism Quarterly, 71 (Winter), 973–983.
Beard, F. K. (2005), “One hundred years of humor in American advertising,” Journal of
Macromarketing, 25(1), 54–65.
Belch, G. E. and Belch., M. A. (2007), Advertising and Promotion: An Integrated Marketing
Communications Perspective (7th ed.). New York, NY: McGraw-Hill.
Bohner, G., Einwiller, S., Erb, H., and Siebler, F. (2003), “When Small Means Comfortable: Relations Between Product Attributes in Two-Sided Advertising,” Journal of
Consumer Psychology, 13(4), 454–463.
Britton, B. K. and Tesser, A. (1982), “Effects of Prior Knowledge on Use of Cognitive Capacity in Three Complex Cognitive Tasks,” Journal of Verbal Learning and Verbal
Behavior, 21, 421–438.
Brown, M. R., Bhadury, R. K., and Pope, N. K. LI. (2010), “The Impact of Comedic Violence on Viral Advertising Effectiveness,” Journal of Advertising, 39 (Spring), 49–65. Bryant, J. Brown, D. Silberberg, A. R. and Elliott, S. M. (1981), “Effects of Humorous
Illustrations in College Textbooks,” Human Communication Research, 8 (Fall), 43-57. Campbell, M. C. and Keller, K. L. (2003), “Brand Familiarity and Advertising Repetition
Effects,” Journal of Consumer Research, 30 (Fall), 292–304.
Cacioppo, J. T., Marshall-Goodell, B. S., Tassinary, L. G., and Petty, R. E. (1992), “Rudimentary Determinants of Attitudes: Classical Conditioning Is More Effective When Prior Knowledge about the Attutude Stimulus Is Low than High,” Journal of
Experimental Social Psychology, 28, 207-233.
Chebat, J., Limoges, F., and Gélinas-Chebat, C. (1998), “Limits of the Effect of Advertisement Framing: The Moderating Effects of Prior Knowledge and Involvement,” Advances in Consumer Research, 25, 324–333.
37
Coulter, K. S. and Punj, G. N. (2004), “The Effects of Cognitive Resource Requirements, Availability, and Argument Quality on Brand Attitudes,” Journal of Advertising, 33 (Winter), 53–64.
Davis, J. J. (1994), "Consumer Response to Corporate Environmental Advertising, "Journal
of Consumer Marketing,” 11 (2), 25-47.
Duncan, C. P. (1979), “Humor in Advertising: A Behavioral Perspective,” Journal of the
Academy of Marketing Science, 7 (4), 285–306
Eisend, M. (2006), “Two-sided Advertising: A Meta-analysis,” International Journal of
Research in Marketing, 23, 187–198.
– – – (2009), “A Meta-Analysis of Humor in Advertising,” Journal of the Academic
Marketing Science, 37, 191-203.
– – – (2011), “How Humor in Advertising Works: A Meta-Analytic Test of Alternative Models,” Marketng Letters, 22 (2), 115-132.
Field, A., Miles, J., and Field, Z. (2012), Discovering Statistics Using R (1st ed.) London: Sage Publications Ltd.
Goldberg M. E. and Hartwick, J. (1990), “The Effects of Advertiser Reputation and Extremity of Advertising Claim on Advertising Effectiveness,” Journal of Consumer Research, (Fall), 172–179.
Goldsmith, R. E., Lafferty, B. A., and Newell, S. J. (2000), “The Impact of Corporate Credibility and Celebrity Credibility on Consumer Reaction to Advertisements and Brands,” Journal of Advertising, 29 (Fall), 43–54.
Grewal, D., Kavanoor, S., Fern, E. F., Costley, C., and Barnes, J. (1997), “Comparative Versus Noncomparatative Advertising: A Meta-Analysis,” Journal of Marketing, 61 (Fall), 1–15.
Gu, R., Lei, Z., Broster, L., Wu, T., Jiang, Y., and Luo, Y. (2011), “Beyond Valence and Magnitude: A Flexible Evaluative Coding System in the Brain,” Neuropsychologia, 49 (14), 3891–3897.
Hong, J. and Sternthal, B. (2010), “The Effects on Consumer Prior Knowledge and Processing Strategies on Judgments,” American Marketing Associations, 47 (Spring), 301–311.
38
Kamins M. A. and Assael, H. (1984), “Two-Sided Versus One-Sided Appeals: A Cognitive Perspective on Argumentation, Source Derogation, and the Effect of Disconfirming Trial on Belief Change,” Journal of Marketing Research, 14 (Winter), 29–39.
Klebba, J. M. and Unger, L. S. (1983), “The Impact of Negative and Positive Information on Source Credibility in a Field Setting,” Advances in Consumer Research, 10 (1), 11-16. Krishnan, H. S. and Chakracarti, D. (2003), “A Process Analysis of the Effects of Humorous
Advertising Executions on Brand Claims Memory,” Journal of Consumer Psychology, 13 (3), 230–245.
Lee, Y. H. and Lim, E. A. C. (2010), “When Good Cheer Goes Unrequited: How Emotional Receptivity Affects Evaluation of Expressed Emotion,” Journal of Marketing
Research, 47 (Winter), 1151 – 1161.
Liao, Q. V. and Fu, W. (2011), “Age difference in credibility judgments of online health information,” Association for Computing Machinery, 21 (1), 1 – 23.
Machleit, K. A. and Wilson, R. D. (1988), “Emotional Feelings and Attitude Towards the Advertisement: The Roles of Brand Familiarity and Repetition,” Journal of
Advertising, 17 (3), 27–35.
– – –, Allen, C. T., and Madden, T. J. (1993), “The Mature Brand and Brand Interest: An Alternative Consequence of Ad-Evoked Affect,” Journal of Marketing, 57 (Fall), 72– 82.
Martin, B. A. S., Lang, B., and Wong, S. (2003), “Conclusion Explicitness in Advertising,”
Journal of Advertising, 32 (Winter), 57–65.
Myers, R. (1990), Classical and modern regression with applications (2nd ed.). Boston, MA: Duxbury.
Mitchell, A. A. & Dacin, P. A. (1996), “The Assessment of Alternative Measures of Consumer Expertise,” Journal of Consumer Research, 23 (Winter), 219–239.
Ou, Y. C., Vries, de, L., Wiesel, T., and Verhoef, P. C. (2014), “The Role of Consumer Confidence in Creating Customer Loyalty,” Journal of Service Research, 17 (3), 339-354.
Priester, J. R. and Petty, R. E. (2003), “The Influence of Spokesperson Trustworthiness on Message Elaboration, Attitude Strength, and Advertising Effectiveness,” Journal of
Consumer Psychology, 13 (4), 408–421.
Razali, N. M. and Wah, Y. B. (2011), “Power comparisons of Shapiro-Wilk, Kolmogorov-Smirnov, Lilliefors and Anderson-Darling tests,” Journal of Statistical Modeling and
39
Reynaldo, J. & Santos, A. (1999), “Cronbach's Alpha: A Tool for Assessing the Reliability of Scales,” Extension Journal Inc, 37.
Rouet, J. F., Britt, M. A., Mason, R. A., and Perfetti, C. A. (1996), “Using multiple sources of evidence to reason about history,” Journal of Educational Psychology, 88 (3), 478– 493.
– – –, Favart, M., Britt, M. A., and Perfetti, C. A. (1997), “Studying and Using Multiple Documents in History: Effects of Discipline Expertise,” Cognition and Instruction, 15 (1), 85–106.
Shapiro, S. S. and Wilk, M. B. (1965), “An analysis of Variance Test for Normality (Complete Samples),” Biometrika, 52 (3/4), 591-611.
Sharifpour, M. and Walters, G. (2014), “The Interplay Between Prior Knowledge, Perceived Risk and the Tourism Consumer Decision Process: A Conceptual Framework,” The
Marketing Review, 14 (3), 279–296.
Slater, M. D. and Rouner, D. (1996), “How Message Evaluations and Source Attributes May Influence Credibility Assessment and Belief Change,” Journalism & Mass
Communication Quarterly, 73 (Winter), 974-991.
Spotts, H. E., Weinberger, M. G. and Parsons, A. L. (1997), “Assessing the Use and Impact of Humor on Advertising Effectiveness: A Contingency Approach,” Journal of
Advertising, 16 (Fall), 17–32.
Sternthal, B. and Craig, S.(1973), “Humor in Advertising,” Journal of Marketing, 37 (Fall), 12–18.
– – –, Dholakia, R., and Leavitt, C. (1978), ''The Persuasive Effect of Source Credibility: Tests of Cognitive Response," Journal of Consumer Research, 4 (Spring), 252-260. Sutherland, J. C. (1982), “The Effects of Humor on Advertising Credibility and Recall,”
Educational Resource Information Center.
Teixeira, T. S. and Stipp, H. (2013), “Optimizing the Amount of Entertainment in Advertising,” Journal of Advertising Research, 53 (Fall), 286–296.
Warren, C. and McGraw, A. P. (2013), “When Humor Backfires: Revisiting the Relationship Between Humorous Marketing and Brand Attitude,” Marketing Science Institute
Working Paper Series, 13.
Weinberger, M. G. and Spotts, H. E. (1989), “Humor in U.S. versus U.K. TV Commercials: A Comparison,” Journal of Advertising 18 (2), 39–44.
Whipple, T. W. and Courtney, A. C. (1981), “How Men and Women Judge Humor,” (in eds.)
40
Wineburg, S. (1991), “Historical Problem Solving: A Study of the Cognitive Processes Used in the Evaluation of Documentary and Pictorial Evidence,” Journal of Educational
Psychology, 83 (1), 73–87.
41 APPENDIX Appendix A. Questionnaire
I am conducting a survey for my Master thesis. The survey contains 19 multiple choice questions and 1 final open question. Filling in the questionnaire will take 5 minutes of your time and will be completely anonymous. Thank you in advance for your cooperation.
1. What is your age?
1. Younger than 18 years old 2. Between 18 and 29 3. Between 30 and 39 4. Between 40 and 49 5. Between 50 and 64 6. 65 years or older
2. What is your gender? 1. Man
2. Women
3. Could you indicate your gross income per year? 1. Lower than €30.000,- per year
2. Approximately €30.000,- per year
3. Between €30.000,- and €60.000,- per year 4. €60.000 ,- or higher
5. I would rather not say this
The following questions are referring to the brand Tarex. In order for you to be able to answer these questions, some basic information about the brand is provide next. Please read this information carefully, you will not be able to go back to this information.
42
People’s perceptions are important in advertisement. You will now see three options, please give the answer that represents your feeling the best.
Insert unrelated filler task humorous or serious options
4. In which option do you think the product is more noticeable (humorous ad)?/ Which product color has your preference (serious ad)?
1. Option 1 2. Option 2 3. Option 3
Tarex is currently advertising their mouthwash (in Germany). Please take a close look at their advertisement. Afterwards you will go to the questions and you will not be able to go back to this advertisement.
Insert humorous or serious advertisement
The following questions concern the advertisement that you just saw. Please indicate your answer from a scale from 1 to 7, where 1 and 7 each represents two extremities (antonyms) and 4 represents neutral.
5. How humorous was the advertising, according to you?
1 2 3 4 5 6 7
Not humorous at all Neutral Extremely humorous
6. How interesting was the advertising, according to you?
1 2 3 4 5 6 7
Not interesting at all Neutral Extremely interesting
7. How enjoyable was the advertising, according to you?
1 2 3 4 5 6 7
43
8. How funny was the advertising, according to you?
1 2 3 4 5 6 7
Not funny at all Neutral Extremely funny
9. How serious was the advertising, according to you?
1 2 3 4 5 6 7
Not serious at all Neutral Extremely serious
10. How dull was the advertising, according to you?
1 2 3 4 5 6 7
Not dull at all Neutral Extremely dull
11. How credible was the advertising?
1 2 3 4 5 6 7
Not credible at all Neutral Extremely credible
12. Compared to most commercials you see, did you find the Tarex advertisement:
1 2 3 4 5 6 7
Deceptive Neutral Honest
13. Compared to most commercials you see, did you find the Tarex advertisement:
1 2 3 4 5 6 7
Misleading Neutral Sincere
The following questions are about the brand Tarex. Please indicate your answer form a scale from 1 to 7, where 1 and 7 each represents two extremities (antonyms) and 4 represents neutral.
14. What do you think about the brand Tarex?
1 2 3 4 5 6 7
44 15. What do you think about the brand Tarex?
1 2 3 4 5 6 7
Bad Neutral Good
16. What do you think about the brand Tarex?
1 2 3 4 5 6 7
Dislikeable Neutral Likeable
17. What do you think about the brand Tarex?
1 2 3 4 5 6 7
Not nice Neutral Nice
18. How likely are you to purchase mouthwash?
1 2 3 4 5 6 7
Not at all likely Neutral Very likely
19. How likely are you to purchase this brand Tarex?
1 2 3 4 5 6 7
Not at all likely Neutral Very likely
20. What do you think the study is about? Fill-in
45 Appendix B. Measurement and Latent Variables
Measurements variables Latent variable
Control variables 1. What is your age?
(“younger than 18 years old” –“65 years or older”) 2. What is your gender?
(“man” – “woman”)
3. Could you indicate your gross income per year?
( “lower than €30.000,- per year” – “€60.000 ,- or higher”) (Ou, de Vries, Wiesel, and Verhoef, 2014)
Independent variable
1. How humorous was the advertising? (“not humorous” – “humorous”) (Zhang, 1996)
2. How interesting was the advertising?
(“not interesting at all” – “extremely interesting”) (Bryant, Brown, Silberberg & Elliot, 1981) 3. How enjoyable was the advertising?
(“not enjoyable at all ” – “extremely enjoyable”) (Zhang, 1996)
4. How funny was the advertising? (“not funny at all” – “extremely funny”) (Zhang, 1996)
5. How serious was the advertising? (“not serious at all” – “extremely serious”)
6. How dull was the advertising? (“not dull” – “dull”)
(Zhang, 1996)
Moderating/independent variable
1. What do you think about the brand Tarex? (“unfavorable” – “favorable”)
(“bad” – “good”) ( “dislike” – “like”) (“not nice” – “nice”) (Hong and Sternthal, 2010) Dependent variables
1. How credible was the advertising?
(“not credible at all” – “extremely credible”) (Bryant, Brown, Silberberg, and Elliot, 1981)
Type of ad
Prior knowledge
46
2.Compared to most commercials you see, did you find the Tarex advertisement?
(“deceptive” – “honest”) (“misleading” – “ sincere”) (Goldberg and Hartwick, 1990)
1. How likely are you to purchase mouthwash? (“not at all likely” – “very likely”)
2. How likely are you to purchase this brand Tarex? (“not at all likely” – “very likely”)
(Teixeira and Stipp, 2010)
Category involvement
