1
EBM867B20
Master Thesis Marketing
The effect of repeated exposure to
annoying advertising on brand attitude
The moderating effect of level of arousalJohannes Meijer 1437151
MSc. Business Administration (Marketing)
First Supervisor: dr. Debra Trampe Second Supervisor: dr. Hans Risselada
Faculty of Economics and Business University of Groningen
The Netherlands
2 Management summary
The purpose of this study is to determine whether the mere exposure effect holds for negatively perceived stimuli. Previous research towards the mere exposure effect is limited and mainly focused on the effect of positively and neutral perceived stimuli (Baker, 1999; Janiszewski, 1993). This experiment adds value to existing literature, because the focus is on annoying commercials. Beforehand, it was unclear how brand attitude was affected by repeated exposure towards annoying commercials.
The reason that it is unclear how the repeated exposure of annoying commercials influences the perception of brands among consumers, is because theories of mere exposure and classical conditioning contradict. Following the reasoning of classical conditioning, the expectation was raised that pairing a brand with a feeling of
annoyance would result in a negative evaluation of the advertised brand. In this case the feeling of annoyance is transferred from the commercial towards the brand, and
consequently results in a negative brand attitude (Baker, 1999) . On the contrary, mere exposure theory would predict the exact opposite effect. By being exposed to a stimulus several times, the stimulus would be perceived as less threatening and as such more approachable (Zajonc, 2001). Also, recognising the stimulus in itself is experienced as a rewarding experience. The positive reinforcement of recognizing the stimulus is not based on relevant characteristics of the stimulus, but simply occurs because the stimulus looks familiar. This effect is known as perceptual fluency and occurs
automatically when astimulus is shown repeatedly (Fang, Sing and Ahluwalia, 2007). Because the mere exposure effect is much less studied then theory of classical conditioning, this study focussed on the mere exposure effect.
3 easier. This is known as perceptual fluency, which results in an automatic positive
evaluation (Fang, Sing and Ahluwalia, 2007). Consequently, chances will increase that the rewarding experience of recognizing the stimulus takes place. Hence, level of arousal is included in this experiment.
In order to test whether the mere exposure effect holds for annoying commercials a 2x4 factorial design was conducted in which number of exposures and level of arousal were varied.
Beforehand it was expected that for initial exposures brand attitude would enhance for both the modest and the high arousal condition which is known as wearin effect
(Berlyne, 1970). Because arousal leads to enhanced perceptual fluency, it was
hypothesized that the wearin effect would be stronger for the high arousal condition. After the initial exposures it was expected that brand attitude would decrease for both conditions because of the occurrence of boredom which causes tedium (Campbell and Keller, 2003). This decrease after several repetitions is known as the wearout effect.
4 Preface
Everybody is familiar with extremely annoying commercials. Screaming children, women who get their white shirts really, really, really white again, or simply a message that insists through the use of bright colours and loud noises to buy a certain product. The one is even more annoying than the other, which raises the question why marketers choose for such an approach.
For me the psychology of the consumer is the most interesting part of marketing and therefore I enjoyed conducting this research. Our brain fools us all the time, and I could imagine that that this be the reason why consumers decide to buy products which are advertised in an extremely annoying way.
The occurrence of subconscious processes is something that fascinates me for several years now. Therefore, a study in which I could research the occurrence of subconscious processes like the mere exposure effect and classical conditioning was really interesting for me. Hence, I really enjoyed working on this thesis.
Furthermore, I would like to thank my supervisors for their support. Sometimes, in the process of writing a thesis, you get at a certain point in which an external view can help you to choose the right direction. This external view of Debra Trampe and Hans
5
Table of Contents
1. Introduction……… 7
1.1 Background and problem statement……… 7
1.2 Theoretical and managerial relevance Attitude………..10
1.3 Structure……….………. 11
2. Theoretical Framework……….………. 12
2.1 Brand Attitude……….….. 12
2.2 Attitude Formation……….. 12
2.3 Affective Classical Conditioning……… 13
2.4 Mere Exposure……… 13
2.5 Arousal……….. 17
2.6 Conceptual Model……… 19
3. Methodology ………... 20
3.1 Subjects and Design……… 20
3.2 Test Commercials………. 20
3.3 Independent Variable: Number of Exposures……… 22
3.4 Moderating Variable: Level of Arousal………. 22
3.5 Dependent Variable: Brand Attitude……….. 22
3.6 Procedure………. 22 3.7 Method of Analysis……… 23 4. Results ……….. 24 4.1 Data description……… 24 4.2 Manipulation Checks……….. 24 4.2.1 Brand Familiarity……….. 24
4.2.2 Perception of the Test Commercial……….25
6
4.3 Reliability Analysis Brand Attitude……….. 27
4.4 Hypothesis Test………. 27
5. Discussion an Recommendations……… 31
5.1 Discussion ……… 31
5.2 Theoretical Implications……… 32
5.3 Managerial Implications……….. 33
5.4 Limitations and Directions for Future Research……… 33
6. References ……… 35
7. Appendices ……….. 39
7.1 Appendix 1; Questionnaire………. 38
7
1. Introduction
1.1 Background and problem statement
Screaming women who get overly excited when they see the mailman who brings them their latest online purchases. This commercial of online retailer Zalando was regularly broadcasted in the Netherlands and awarded with the ‘loden leeuw’ in 2012, which is a price for the most annoying commercial (www.nos.nl). One might think that this is frustrating news for the advertisers who came up with this commercial. However, the opposite is true. Rosanne Thesing, PR-manager at Zalando, states that the commercials are developed to provoke consumers (www.marketingfacts.nl). This way of
communicating has resulted in a brand awareness of ninety-two per cent in the Netherlands, within two years after starting with the commercials. Furthermore, it is estimated that in 2013 Zalando and their main competitor Wehkamp together had a market share of 80 to 85 percent of the online fashion market in the Netherlands (www.marketingfacts.nl). Zalando does not have any physical shops, and consequently fully depends on their marketing activities in order to attract potential customers
towards their website. Therefore, the advertising campaign of Zalando can be seen as an enormous success.
Although many commercials are perceived as annoying, little research has been done towards this type of advertising. Instead, most research is focused on the effects of positive advertising (Baker, 1999; Janiszewski, 1993). By eliciting positive emotions of consumers, who are exposed towards an advertisement, advertisers aim on creating a positive feeling towards their product. The pairing of a neutral stimulus with a stimulus that creates a positive feeling, is known as classical conditioning (Zajonc, 2001). By creating a positive feeling among consumers, advertisers aim at enhancing attitudes of consumers towards their products.
8 because it is an effective way to develop positive attitudes towards a brand (Grosmann and Till, 1998). The use of positive images to condition responses towards brands has been proven in several studies (Shimp, Stuart and Engle, 1991; Stuart, Shimp, and Engle, 1987). Furthermore, attitudes that are formed through classical conditioning are
retained long enough to influence consumers at the time of purchase (Grosmann and Till, 1998). These findings are in accordance with the results of studies of Baeyens et al. (1988) and Zellner et al. (1983) who showed that classical conditioning is resistant to extinction. As such, classical conditioning of brands with stimuli that enhance positive emotions can positively influence sales.
When the principle of classical conditioning is taken into account, it is questionable why advertisers use annoying advertisement campaigns to promote their products. If
consumers have a feeling of annoyance towards a product, it can be expected that they will negatively evaluate that product. Therefore, using annoying advertising might negatively influence sales, because of the formation of negative attitudes towards the advertised brands or products.
9 Within a marketing context, repeated exposure has the effect that the risk that is
associated with choosing a brand is reduced (Baker, 1999). Hence, consumers might purchase products and brands which have an exposure advantage. These findings stimulate companies to choose a marketing strategy in which their brand name is exposed heavily.
Interestingly, the effects of classical conditioning and mere exposure seem to be contradictory for negatively perceived stimuli. From the perspective of classical
conditioning theory, it is expected that repeatedly pairing an unconditioned brand name with a negatively perceived commercial will lead to a negative brand evaluation.
However, mere exposure theory suggests that simply being exposed several times towards annoying advertisements would create a feeling of safety and as such creates positive brand evaluations. Previous research focused on the effects of repeated exposure towards positively perceived stimuli (Baker, 1999; Zajonc, 2001). In these studies classical conditioning and mere exposure both lead to a more positive brand evaluation. However, these theories contradict when related to the repeated exposure of annoying advertising. As such, this research adds novel insights into the mechanisms underlying the mere exposure effect.
This study focusses on the mere exposure effect, rather than the theory of classical conditioning. According to Grossman and Till (1998) classical conditioning has been referred to as the most widely studied mechanism of the influence of advertisements on brand attitude. Mere exposure has been studied much less in an advertising context (Baker, 1999). Furthermore, the mere exposure effect has never been tested for negatively perceived stimuli.
10 commercials that elicit negative feelings. After being exposed towards this type of advertisements subjects negatively perceived the brands that were advertised.
Because little research has been done towards the mere exposure effect in a marketing context and because the mere exposure effect has never been tested for negatively perceived stimuli, this study examines whether brand attitude is enhanced when the number of exposures of annoying commercials increases.
Another aspect that is taken into account in this study is the effect of level of arousal that is generated by annoying commercials. Intuitively, is expected that the extent to which a commercial evokes arousal influences the relationship between number of exposures and brand attitude. When a commercial evokes higher levels of arousal perceptual fluency is enhanced, which results in a positive influence on the relationship between number of exposures and brand attitude.
This leads to the following research question:
What is the effect of repeated exposure to annoying advertisement on brand attitude and is this effect moderated by the level of arousal?
1.2 Theoretical and managerial relevance
11 From the perspective of advertisers this study is interesting, because it shows whether or not it is beneficial to use annoying advertising. Although the mere exposure effect is proven for neutral perceived stimuli, it is unclear what the effect is for negatively perceived stimuli. This study will show whether or not it is beneficial for companies to make use of annoying advertisements and whether or not it is beneficial to expose consumers extensively towards these types of advertisements. Furthermore, it will show whether it is beneficial to use commercials that are either modest or high arousal
evoking.
1.3 Structure
This paper is structured as follows; in the following chapter relevant literature is
12
2. Theoretical Framework
Within this the chapter the theories are discussed that are relevant for this study. After the theories are mentioned, the hypotheses will be discussed which are tested in the following chapters.
2.1 Brand attitude
According to Crano and Prislin (2006) attitudes are evaluative judgments towards a stimulus that represents an evaluative integration of cognition and affect experienced in relation to an object. With regard to brands, attitudes determine how a brand is
evaluated by consumers (Simonin and Ruth, 1998).
2.2 Attitude formation
With commercial campaigns advertisers aim to be persuasive towards consumers in order to influence the attitudes that they have towards the advertised product, which eventually determines which brand an individual purchases (Baker, 1999).
The formation of attitudes towards a subject can occur in several ways (Crano and Prislin, 2006). With regard to attitude formation, a distinction can be made between attitudes that are formed consciously, and attitudes that are formed unconsciously. Conscious formation of attitudes takes place when an individual is motivated and has the ability to deliberate about the stimulus that is presented (Baker, 1999). One way in which advertisers can create positive feelings toward their products is to communicate that the advertised brand outperforms competitors on several aspects (Baker, 1999). For example, when an advertisement claims that a car is safer, faster, and cheaper than their main competitors, positive attitudes towards this car are formed through
conscious reasoning.
13 conscious awareness. Unconscious formation of attitudes takes place through classical conditioning and the mere exposure effect. These effects are described below.
2.3 Affective classical conditioning
With classical conditioning repeatedly pairing a stimulus that is unconditioned with a stimulus that is conditioned, will cause the stimulus which is conditioned to
automatically evoke a conditioned respons (Baker, 1999).
In order for classical conditioning to be effective, it is necessary that individuals who are exposed to advertising have sufficient attention towards the advertisement. However, it is not required that individuals are aware that that they are learning or that the
advertisement has a persuasive effect (Allen and Janiszewski, 1989). For example, classical conditioning can be effective when consumers are aware of the fact that an advertisement is funny, even though they are unaware that being exposed to a funny advertisement influences their perceptions and behaviour towards a brand (Baker, 1999).
2.4 Mere exposure
14 According to Fang, Sing and Ahluwalia (2007) the mere exposure effect can be explained because the stimulus that is presented several times is processed more easily. This is described as perceptual fluency, which leads to an automatic positive evaluation. A positive feeling is generated because a stimulus that is recognized indicates familiarity, which is a signal of a harmless situation. This is confirmed by Zajonc (2001) who states that the absence of aversive consequences is perceived as a signal of safety.
Furthermore, Winkielman and Cacioppo (2001) state that successfully recognizing a stimulus in itself is a rewarding experience. Fang, Sing and Ahluwalia (2007) describe this effect as the misattribution model. Instead of judging a stimulus on relevant aspects, the stimulus is liked, simply because the individual has seen it before.
Within a marketing context, being exposed several times towards a brand can give the brand a competitive advantage over its competitors, because the exposure gives consumers a comfortable feeling when approaching the brand (Baker, 1999).
Furthermore, Zajonc (2001) states that the mere exposure effect occurs even when a stimulus is processed subliminally. This implies that an individual does not have to consciously process the stimulus that is shown. Studies of Murphy, Monohan, and Zajonc (1995) and Zajonc (1980) show that subliminal stimuli that are flashed frequently are better liked than subliminal stimuli that are flashed less frequently. As such, these studies show that even when participants are not aware of being exposed towards stimuli, the number of exposures positively influences the attitude which they hold towards these stimuli.
15 experiment showed that attitudes towards these words were mediated by the number of exposures.
Another example is shown in a study of Becknell, Wilson and Baird (1963) where nonsense syllables were repeatedly shown and afterwards addressed to one of several boxes which contained nylon stockings. The frequency of exposure of the nonsense syllables had a significant effect on brand preference and brand choice.
Within these two examples the Turkish words or the nonsense syllables in itself do not cause an emotional effect. People who do not know the meaning of Turkish words do not feel happy or excited by seeing them. Therefore the increased liking towards the brands can only be caused by the repeated exposure. This is different for annoying advertising where the stimulus in itself will cause a negative emotional response.
However, it is unclear what happens if negatively valenced stimuli are presented repeatedly. Until now research has not shown whether brands included in annoying advertising are liked better or worse after being shown several times. Theories of classical conditioning and mere exposure contradict with regard to what is expected. When the principle of classical conditioning is taken into account, it is expected that being exposed to annoying advertising will lead to a negative brand evaluation. When an individual is confronted with an annoying advertisement the feeling of irritation will be paired with the brand which is shown. Consequently, a negative brand evaluation will occur. However, the mere exposure effect predicts precisely the opposite effect. By being repeatedly exposed to an annoying commercial, the commercial and the brand name will become more familiar and hence, less threatening. Furthermore, a consumer will experience that the risks which are associated with purchasing this brand will be reduced. Simply being exposed repeatedly towards the stimulus makes the stimulus more accessible, and therefore liked more. This feeling of familiarity and the
16 Previous research towards classical conditioning has shown that pairing a negatively valenced stimulus with a neutral valenced stimulus will lead to a negative evaluation of the neutral stimulus after several exposures (Levy and Martin, 1975). Within their research Levy and Martin (1975) paired pictures which subjects experienced as neutral, with pictures which they disliked. After pairing the two pictures with each other, the picture which originally was seen as neutral, was evaluated negatively after several exposures.
Furthermore, Gresham and Shrimp (1985) referred to classical conditioning as “the most widely discussed mechanism of the influence of advertising on consumers brand
attitudes.” In their study, Gresham and Shrimp (1985) paired supermarket products with positively and negatively valenced commercials in order to generate conditioned
responses. Moreover, classical conditioning in general has been studied extensively in order to explain the effects of advertising (Gorn, 1982; Janiszewski and Warlop, 1993; Pitt and Abratt, 1988).
However, empirical studies of the mere exposure effect in an advertising context are limited. Furthermore, the few studies which focused on this effect in advertising (Baker, 1999; Campbell and Keller, 2003; Janiszewski, 1993; Obermiller, 1985) used neutrally valenced stimuli. Hence, it is unknown what happens when annoying advertising is repeatedly shown. To provide insights into this gap in the literature, this study will focus on the mere exposure effect rather than the principle of classical conditioning.
17 resulted in a more negative evaluation of the brand. As such, the positive relationship between number of exposures and brand attitude changed into a negative relationship after two exposures.
These findings are in accordance with Berlyne’s (1970) two factor theory. This theory proposes a two part process in which repetition influences the response towards messages. During initial levels of repetition brand attitude is enhanced. This phase is known as wearin, in which initial levels of repetition create a feeling of safety and as such, lead to a more positive evaluation of the stimulus (Campbell and Keller, 2003). After initial levels of repetition brand attitude will decrease because of the occurrence of boredom. This phase is known as wearout (Berlyne’1970).
When the reasoning of the two factor (Berlyne, 1970) theory is followed with regard to this experiment, it is expected that wearin effect occurs from the first to second exposure (Campbell and Keller, 2003). Repeated exposure serves in this phase to increase positive habituation in the absence of negative consequences (Campbell and Keller, 2003).
This results in the following hypotheses:
H1: The number of exposures towards annoying advertising has a positive effect on brand attitude from the first to the second exposure.
2.5 Arousal
Arousal can be described in several ways. According to Singh and Churchill (1987) an individual who experiences arousal feels tense, alert and energized. With regard to advertising arousal is relevant, because it provides an individual with energy for psychological activity (Singh and Churchill, 1987).This psychological activity will
18 According to Singh and Churchill (1987) psychological events like the experience of arousal create a short term memory trace in the brain, which lasts for a consolidation period. During this period of consolidation, a long-term memory trace is formed. Meanwhile, retrieval is inhibited in order to retain the trace and protect it against disruption (Singh and Churchill, 1987). This is known as action decrement theory. Singh and Churchill (1987) also state that higher levels of arousal lead to memory traces which last longer, and as such, to greater long-term memory. These findings are confirmed by Sharot and Phelps (2004), who show that arousal influences memory through enhancing attention and perception during encoding and by affecting memory retention.
Intuitively, it is expected that the level of arousal that is generated by the commercial will influence the relationship between the number of exposures and brand attitude. Because more arousal leads to a stronger memory trace, perceptual fluency will be enhanced. The previous section described that because the stimulus is processed more easily, a positive feeling towards the stimulus is created. As such, it is expected that more arousal leads to more perceptual fluency and therefore has a positive effect on the relationship between the number of exposures and brand attitude. Hence, the following hypothesis is formulated:
H2: The effect of repeated exposures to annoying advertising on brand attitude are stronger for initial repetitions when high levels of arousal are experienced than when modest levels of arousal are experienced.
19 and Keller, 2003).
Although it is expected that wearout eventually occurs for all levels of arousal, wearout probably occurs after less exposures for modest levels of arousal, than for high levels of arousal. When high levels of arousal are experienced, consumers will feel less boredom, and consequently do not think about counterarguments. The following hypothesis is therefore proposed:
H3: The wearout effect towards annoying advertising occur later when subjects experience high levels of arousal, than when subjects experience modest levels of arousal.
2.6 Conceptual Model
From the literature described above, a conceptual model is derived. This conceptual model is shown in figure 1.
20
3. Methodology
In order to provide support for the hypotheses a survey was conducted.This section starts with describing the research design that was used. Subsequently, information about the participants is given. Furthermore, the variables and way of manipulating is described, followed by the procedure of participating with the survey.
3.1 Subjects and design
The experiment was tested through exposing participants towards annoying television commercials. The experiment was a 2 x 4 factorial design in which the number of exposures (1,2,3 or 4) and the type of advertisement (modest arousal versus high arousal) were varied. In this design different people were exposed to different experimental manipulations, which implies a between group design (Field, 2010).
Each participant of the experiment was randomly assigned to one of eight conditions. For every condition a minimum of 20 individuals participated. However, due to random assignment of the participants towards one of the questionnaires, not every condition contained an equal amount of subjects. In total 195 subjects participated in the experiment. The data was collected through asking people in the environment of the researcher to fill in the questionnaire. Furthermore, the researcher asked friends and family to further spread the questionnaire. This way of collecting data is known as snowball sampling (Biernacki and Waldorf, 1981).
3.2 Test commercials
An important aspect with regard to executing the experiment was the selection of two commercials which were suitable to measure the effects that are stated in the
21 participants. As such, the brand attitude that was measured was fully determined by the exposure towards commercials which were shown, without being influenced by pre-existing attitudes towards the brand. Furthermore, the commercials had to be perceived as being annoying and a difference in the level of generated arousal had to be present. Intuitively, commercials of the Asian brand Pretz fulfilled all these requirements. Pretz is a kind of cookie that is sold mainly on the Asian market. In the Netherlands it is available at special Asian stores, but not in the mainstream supermarkets. As such, it was
expected that most respondents did not know the brand.
Through a pre-test two commercials of Pretz were selected that best met the criteria that are stated above. A pre-test was conducted in which 19 students were asked to rate five different commercials of the brand Pretz. They were asked about their opinion about the commercials which were shown (Funny, Boring, Annoying, Other). Further, in order to measure experienced level of arousal, they were asked to rate each commercial on a four-item differential scale, anchored by stimulated-relaxed, slow-fast, dull-jittered and sleepy-wide awake (Bearden, Netemeyer and Haws, 2011).
From the five commercials that were show, one commercial was selected as modest arousal evoking commercial, and one commercial was selected as high arousal evoking commercial. The pre-test showed that 74 per cent thought that the modest arousal evoking commercial was annoying, while 79 per cent of the respondents had the opinion that the high arousal evoking commercial was annoying. Further, it was shown that the modest arousal commercial (M = 2.81, sd = 0.88) and the high arousal
22 3.3 Independent variable: number of exposures
Within the experiment the number of exposures was manipulated. The test commercial was shown 1, 2, 3 or 4 times. The number of exposures was similar to the number of exposures that were used in a study of Campbell and Keller (2003) in which the mere exposure effect was proven within an advertising context.
3.4 Moderating variable: level of arousal
The level of arousal that was generated was manipulated through showing two different commercials. The first commercial was modestly annoying and as such generated a modest level of arousal. The second commercial was extremely annoying and hence generated a high level of arousal.
3.5 Dependent variable: brand attitude
Brand attitude was measured in order to determine the success of repeatedly showing the annoying advertisement. According to Campbell and Keller (2003) brand attitude can be measured through using a three-item, seven-point differential scale. The items were anchored by low quality-high quality, unappealing-appealing, and unpleasant-pleasant (Campbell and Keller, 2003).
3.6 Procedure
The respondents were randomly assigned to one of eight conditions. Instead of telling the true purpose of the research, participants were told that the experiment focussed on testing the effect of showing television commercials. No additional information was given with regard to the goal of testing the effect of repeated exposure.
23 which three different commercials were shown. Besides the test-ad of Pretz, these clips contained filler ads. Depending on the condition to which respondents belonged, the test-ad was shown either 1,2,3 or 4 times. Afterwards participants were asked to answer the questions that measured the brand attitude of the several brands which were shown. Brand attitude was measured with a three-item, seven-point differential scale, anchored by low quality-high quality, unappealing-appealing, and unpleasant-pleasant (Campbell and Keller, 2003). The items were averaged (Cronbach’s alpha = 0.869). Only the results of the test-ads were analysed afterwards. The results of the filler ads were neglected. Subjects further completed manipulation checks and then indicated gender, age, education and family size .
3.7 Method of analysis
In order to test the hypotheses an Anova is conducted to determine the main effect and the interaction effect. Contrast analyses show whether a quadratic trend is present for the modest and high arousal condition, and is used to test whether the wearin and wearout effects are significant for the modest and high arousal condition. In this way the first and third hypotheses are tested. Further, when the Anova shows that the main effect and the interaction effect are significant a regression analysis can be conducted in order to determine whether the beta’s for the modest arousal commercial and the high arousal commercial differ significantly from each other. In this way, the second
24 4. Results
The following chapter discusses the results of the data analysis. Firstly, a description of the data is given. Further, manipulation checks and a reliability analysis of brand attitude were performed followed by tests of the hypotheses.
4.1 Data description
A total of 195 subjects participated with the experiment. Missing value analysis showed that 28 cases missed more than 56 per cent of the variables. According to Hair et al. (2010) cases that contain more than 50 per cent of missing data should be deleted. After deleting these 28 cases 167 cases remained that could be used for the analysis. The remaining cases included less than 4 percent of missing data. Hair et al. (2010) state that missing data of individual cases which remain under 10 percent can be ignored.
Of the respondents 56.3 per cent were male and 43.7 per cent were female. The respondents were between 18 and 71 years old with an average age of 34 years old. With regard to education, 79.6 per cent was highly educated. Furthermore, 74.9 per cent lived either in a one or two persons household.
4.2 Manipulation checks
In order to check whether the manipulations had worked out as planned, several variables were analysed in other to determine the success of the manipulation. Further, it was tested whether manipulations differed per condition which would imply that the condition to which a respondent belonged had a confounding effect.
4.2.1 Brand familiarity
25 important that a commercial was used that was unfamiliar for a majority of the
respondents. Respondents were asked whether or not they knew the brand Pretz before participating with the experiment. Descriptives showed that 97.6 per cent of the respondents was not familiar with the brand Pretz before participating with the
experiment. As such, it is assumed that their evaluation of the brand was fully determined by being exposed towards the commercial which was used in the experiment.
In order to test whether the condition to which a respondent belonged influenced brand familiarity a logistic regression was conducted. Results showed that there was no
significant main effect between number of exposures (W = 0.000, p>0.1) and brand familiarity. Also, no significant main effect was present between the level of arousal and brand familiarity (W = 0.000, p>0.1). Further, no significant effect appeared for the interaction between the number of exposures and level of arousal on opinion about the commercial (W = 0.574, p>0.1). Hence, the conclusion can be drawn that the condition to which a participant belonged did not have a confounding effect on brand familiarity of respondents.
4.2.2 Perception of the test-commercials
26 variables were transferred in new variables which showed whether or not individuals thought a commercial was annoying. After that, a logistic regression was conducted to test whether the condition to which a respondent belonged influenced their opinion about the commercial. Results showed that for the modest arousal evoking commercial there was no significant main effect between number of exposures (W = 0.420, p>0,1) and opinion about the commercial. Also, no significant main effect was shown between the level of arousal and opinion about the commercial (W = 0.015, p>0.1). Further, no significant effect was present for the interaction between the number of exposures and level of arousal on opinion about the commercial (W = 0.033, p>0.1). It is concluded that the condition to which a participant belongs did not have a confounding effect on the opinion of respondents about the low arousal evoking commercial.
Results showed further that for the high arousal evoking commercial there was no significant main effect between number of exposures (W = 0.067, p>0.1) and opinion about the commercial. Further, no significant main effect was present between the level of arousal and opinion about the commercial (W = 0.011, p>0.1). Also, no significant effect appeared for the interaction between the number of exposures and level of arousal on opinion about the commercial (W = 0.518, p>0.1). Hence, the condition to which a participant belonged did not have a confounding effect on the opinion about the high arousal commercial of respondents.
4.2.3 Level of arousal
As described in chapter three, respondents were exposed to either the modest arousal or the high arousal evoking condition. In order to determine whether the two
27 A reliability analysis showed that the Cronbach’s alpha of the experienced level of arousal was 0.891. According to Malhotra (2010) a Cronbach’s alpha higher than 0.6 indicates internal consistency. The Cronbach’s alpha that was found here was higher than 0.6 which implies that the construct of arousal was measured in this experiment. Furthermore, an Anova was conducted in order to test whether the condition to which a respondent belonged influenced the level of experienced arousal. Results showed that the main effect of level of arousal on experienced level of arousal was significant (F = 25.91, p < 0.001). This implies that the modest arousal condition (m = 5.02, SD = 1.36) differed significantly from the high arousal condition (m = 5.65, SD = 1.21). Further, there was no significant effect between number of exposures (F = 0.156, p>0.1) and experienced level of arousal. Also, no significant effect appeared for the interaction between the number of exposures and level of arousal on experienced level of arousal (F = 0.114, p>0.1). It can be concluded that the condition to which a participant
belonged did not have a confounding effect on the experienced level of arousal.
4.3 Reliability analysis brand attitude
As described in chapter three brand attitude was measured with a three-item, seven-point differential scale, anchored by low quality-high quality, unappealing-appealing, and unpleasant-pleasant. A new variable was created by averaging the items which were measured. The Cronbach’s alpha of the variable brand attitude was 0.869. This Cronbach’s alpha is larger than 0.6 and the conclusion can be drawn that that the variable brand attitude is internally consistent.
4.4 Hypothesis tests
28 H1: The number of exposures towards annoying advertising has a positive effect on brand attitude from the first to the second exposure.
H2: The effect of repeated exposures to annoying advertising on brand attitude are stronger for initial repetitions when high levels of arousal are experienced than when modest levels of arousal are experienced.
H3: The wearout effect towards annoying advertising occur later when subjects experience high levels of arousal, than when subjects experience modest levels of arousal.
The hypotheses were tested through conducting an Anova. Analysis of brand attitude showed an insignificant main effect of number of exposures (F(3.167) = 1.719, p = 0.165). As such, there is no proof for a main effect between number of exposures and brand attitude. Furthermore, there was an insignificant main effect of arousal towards brand attitude (F(1,167) = 1.722, p = 0.190). Also, the interaction between number of exposures and level of arousal was insignificant (F(3, 167) = 1.093, p = 0.354). Hence, no evidence is provided that level of arousal influences the relationship between number of exposures and brand attitude.
Consequently, all three hypotheses were rejected. As such, no proof is found with regard to the existence the mere exposure effect for either the modest arousal condition or the high arousal condition. Further, no significant patterns of wearin and wearout were found for either condition. Also, no conclusions can be drawn with regard to whether the slopes of both condition differ for the wearin period and whether wearout occurs later for the high arousal condition, than for the modest arousal condition. Hence, all three hypothesis are rejected.
29 between the number of exposures, level of arousal and brand attitude. This plot is shown below.
Figure two: plot of results
The plot showed that for the modest arousal condition the pattern of wearin and wearout is present. Anova analysis revealed a directional quadratic trend for this condition (F(3, 77) = 1.093, p = 0.349). The means of the different conditions are shown in table 1.
Table 1: means and standard deviations for evaluation measure
30 (β= - 0.668, t = -1.765, p = 0.086 ).
Further, for the high arousal condition no quadratic trend is visible. Regression analysis showed a significant decreasing linear trend (β= -0.251, t = -2.133 p = 0.036). This implies that for the high arousal condition there is a negative relationship between the number of exposures and brand attitude. The results of the regression analyses can be found in table 2.
31
5. Discussion and recommendations
In order to answer the research question this chapter discusses the main findings of the results section and will relate these findings to the theories that are discussed in the theoretical part of this report. Furthermore, implications for theory and practice are given, followed by limitations and recommendations for future research.
5.1 Discussion
The main goal of this study was to investigate whether the mere exposure effect holds for negatively perceived stimuli. Previous research towards the mere exposure effect is limited and aimed on testing the effect of positively and neutral perceived stimuli. Theories of classical conditioning and mere exposure contradict with regard to the expected effect of repeated exposure of annoying advertising. The occurrence of mere exposure was predicted for both the modest and the high arousal evoking condition. Further, patterns of wearin and wearout were expected for both conditions.
32 For the high arousal condition the findings were different. Results showed that the number of exposures had a significant negative influence on the brand attitude of the high arousal condition. This implies that the mere exposure effect is not present for the high arousal condition. Instead, for the high arousal condition it can be concluded that the principle of classical condition determines the perception of the brand. The feeling of annoyance of this commercial is transferred towards the brand, leading to a negative brand evaluation.
According to Campbell and Keller (2003) tedium stimulates individuals to think of counterarguments and hence, negatively influences brand attitude. Although the high arousal evoking commercial probably is not experienced as boring, it might stimulate individuals to think of counterarguments towards the product. By seeing the high arousal commercial several times, extensive feelings of annoyance are evoked, which are difficult to handle for the individuals who watched the commercial. This leads to negative thoughts about the product and a direct transfer of the feelings of annoyance towards the brand which is advertised. Hence, this results into a strong feeling of disliking the product.
5.2 Theoretical implications
33 from the commercial towards the brand, and as such leads to a negative evaluation of the brand which is promoted.
5.3 Managerial implications
Also for managerial purposes, this study provides valuable implications. Annoying commercials are widely used, even though little research has been done towards this type of advertising. Remarkably enough, the lack of research does not motivate
marketing managers to be careful with using annoying commercials. It seems that most of the time the choice for an annoying commercial is determined by a feeling that annoying commercials are remembered more easily, and therefore leads to more sales. Although this might be the case for modest arousal evoking annoying commercials, this study clearly provides evidence that high arousal evoking annoying commercials
negatively influence the perception of consumers towards the brand. Therefore, in practice it is important that when annoying commercials are used, these commercials do not generate high levels of arousal.
5.4 Limitations and directions for further research
With regard to this experiment, several limitations are present. The sample size
consisted of 195 participants. Although this is considered to be suitable for analyzing the data, a larger sample size would lead to more generalizable results.
Furthermore, a large proportion of the respondents of the sample were relatively young and highly educated, which might imply that the sample is not completely
representative for the population.
34 such a short time period. The results might have been different when the commercials were repeated within several days or weeks, instead of within five minutes. However, the experimental conditions did not provide the opportunity to test it this way.
Furthermore, in this experiment the brand Pretz was used. Pretz is a type of snack and therefore can be considered as a low involvement product. Therefore, the variation of brand attitude among consumers is limited. Even someone who positively reacts
towards repeated exposures of the annoying commercial, will not be extremely positive towards the brand, because the advertised brand is not really important for the
consumer. The results might be different for high involvement products. The effects of annoying advertising can be stronger for these types of products. Therefore, testing the repeated exposure for high involvement products can result in different findings. For example, the direction of wearin and wearout which this study showed for the modest arousal commercial, might be more clear for high involvement products.
Further, another interesting extension of this experiment can be to test whether the mere exposure occurs when commercials are used for brands which are familiar for consumers. This research focused on the effect towards an unfamiliar brand. However, the results might be different when familiar brands are tested, because pre-existing attitudes can influence the effectiveness of repeatedly showing the annoying commercial.
35 6. References
Academic journals
Allen, C.T., & Janiszewski, C.A. (1989), “Assessing the Role of Contingency Awareness in Attitudinal Conditioning with Implications for Advertising Research,” Journal of
marketing research, 26(2), 30-43
Baker, W.E. (1999), “When can Affective Conditioning and Mere Exposure Directly Influence Brand Choice?,” Journal of Advertising 28 (4), 31 – 46
Beckell, J.C., Wilson, W.R., & Baird, J.C. (1963), “The Effect of Frequency of Presentation on the Choice of Nonsense Syllable,” The Journal of Psychology 56(1), 165 - 170
Berlyne, D.E. (1970), “Novelty, Complexity, and Hedonic Value,” Perception and Psychophysics, 8(2), 279- 285
Beayens, F., Combez, G., Van den Berg, O., & Eelen, P. (1988), “Once in Contact Always in Contact: Evaluative Conditioning is Resistant to Extinction,” Advances in Behavioral Research Therapy, 10, 179 – 199
Biernacki, P., and Waldorf, D. (1981), "Snowball sampling: Problems and techniques of chain referral sampling." Sociological methods and research 10(2), 141-163
Campbell, C.M., & Keller, K.L. (2003), “Brand Familiarity and Advertising Repetition Effects,” Journal of Consumer Research, 30 (3), 292 - 304
Crano, W.D. & Prisilin, R. (2006), “Attitudes and Persuasion,” Annual Review of Psychology, 57, 345 – 347
Fang, X., Singh, S. & Ahluwalia, R. (2007), “An Examination of Different Explanations for the Mere Exposure Effect,” Journal of Consumer Research, 34(1), 97 – 103
Gorn, G.J. (1982), “The effects of Music in Advertising on Choice Behavior: A Classical Conditioning Approach,” Journal of Marketing 46, 94 - 101
Gressham, L. G. & Shimp, A.T. (1985), “Attitude Toward the Advertisement and Brand Attitudes: A Classical Conditioning Perspective,“ Journal of Advertising, 13(1), 10 - 17
36 Janiszwewski, C. (1993), ”Preattatentive Mere Exposure Effects,“ Journal of Consumer Research, 20(3), 376 – 392
Janiszewski, C. & Warlop, L. (1993), ”The Influence of Classical Conditioning Procedures on Subsequent Attention to the Conditioned Brand,” Journal of Consumer Research, 20 (9), 171 - 188
Levy, A.B. & Martin I. (1975), “Classical Conditioning of Human ‘Evaluative’ Responses,” Behaviour research and therapy ,13(4), 221 - 226
Murphy, S.T., Monahan, J.L. & Zajonc, R.B. (1995), “Additivity of Non-conscious Affect Combined Effects of Priming and Exposure,” Journal of Personality and Social
Psychology, 69(4), 589 - 602
Olson, M.A. & Fazio T.H. (2001), “Implicit Attitude Formation Through Classical Conditioning,” Psychological Science 12(5), 413 – 417
Obermiller, C. (1985), “Varieties of Mere Exposure: the Effects of Processing Style and Repetition on Affective Responses,” Journal of Consumer Research, 6(9). 93 – 100 Pitt, L.F. & Abratt, R. (1988), “Music in Advertisements for Unmentionable Products – A Classical Conditioning Experiment,” International Journal of Advertising, 7, 130 - 137 Shimp, T.A., Stuart, E.W. & Engle, R.W. (1991), “A program of Classical Conditioning Experiments Testing Variations in the Conditioned Stimulus and Context,” Journal of Consumer Research, 19(6), 1 - 12
Singh, S.N., & Churchill, A.G. (1987), “Arousal and Advertising Effectiveness,” Journal of Advertising, 16(1), 4-10
Sharot, T., & Phelps, E.A. (2004), “How Arousal Modulates Memory: Disentangling the Effects of Attention and Retention,” Cognitive, Affective & Behavioral Neuroscience, 4(3), 294 - 306
Simonin, B.L. & Ruth, J.A. (1998), “Is a Company Known by the Company it Keeps? Assessing the Spilover Effects of Brand Alliances on Consumer Brand Attitudes,” Journal of Marketing Research 35(1), 30 - 42
37 Stuart, E.W., Shimp T.A. & Engle, R.W. (1987), “Classical Conditioning of Consumer Attitudes: Four Experiments in an Advertising Context,” Journal of Consumer Research, 14(12), 334 – 351
Winkielman, P. & Cacioppo, J.T. (2001), “Mind at Ease Puts a Smile on the Face: Psychophysiological Evidence that Processing Facilitation Increases Positive Affect,” Journal of Personality and Social Psychology, 81(6), 989 - 1000
Zajonc, R. B. (1968), “Attitudinal Effects of Mere exposure,” Journal of personality and Social Psychology Monographs, 9(6), 1 - 27
Zajonc, R. B. (1980), “Feeling and Thinking Preferences Need no Inferences,“ American psychologist, 35(2), 151 – 175
Zajonc, R.B. (2001), “Mere Exposure: a Gateway to the Subliminal,” Current Directions in psychological science, 10(6), 224 – 228
Zajonc, R. B., & Rajecki, D. W. (1969). “Exposure and Affect: a Field Experiment,” Psychonomic Science, 17, 216 – 217
Zellner, D. A., Rozin P., Aron, M., & Kulish, C. (1983). “Conditioned Enhancement of Human’s Liker for Flavor by Pairing with Sweetness,” Learning and Motivation, 14 (3), 338 – 350.
Books
Bearden, W.O., Netemeyer, R.G & Haws, K. L. (2011), “Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research,” Sage publications, London, UK
Field, A. (2009), “Discovering Statistics Using Spss,” Sage Publications, London, UK Hair, J.F., Black, W.C., Barry, J.B., & Anderson, R.E. (2010), “Multivariate Data Analysis,” Pearson Education, New York, US
39
7. Appendices
7.1 Appendix 1; questionnaire
This appendix contains the first version of the questionnaire in which the high arousal test commercial was show one time. The other versions differed with regard to the number of times the test commercials were shown. Furthermore, half of the
questionnaires contained the modest arousal test commercial instead of the high arousal test commercial.
Beste deelnemer,
Allereerst ontzettend bedankt dat u mijn vragenlijst wil invullen. U helpt mij hier enorm mee met afstuderen!
Mijn onderzoek richt zich op het effect van reclames. Ik vraag u daarom een aantal reclames te bekijken en vervolgens een aantal vragen hierover te beantwoorden. Het is van belang dat u de filmpjes met geluid bekijkt. Deelname aan het onderzoek duurt ongeveer vijf minuten. Uw gegevens zullen uiteraard vertrouwelijk en anoniem verwerkt worden.
40 1)
Ik wil u vragen bovenstaand merk te beoordelen op de volgende kenmerken. De vakjes geven de mate aan waarin u het merk beoordeelt op het kenmerk. (1 = erg lage kwaliteit, 2 = redelijk lage kwaliteit, 3 = enigszins lage kwaliteit, 4 = neutraal, 5 = enigszins hoge kwaliteit, 6 = redelijk hoge kwaliteit, 7 = erg hoge kwaliteit)
Lage Kwaliteit 0 0 0 0 0 0 0 Hoge Kwaliteit
Onaantrekkelijk 0 0 0 0 0 0 0 Aantrekkelijk
Onplezierig 0 0 0 0 0 0 0 Plezierig
2)
Bij het invullen van de volgende vragen wil ik u vragen ervan uit te gaan dat het product verkrijgbaar is bij een winkel bij u in de buurt. Ook deze vragen zijn gericht op het product dat hierboven is weergegeven.
Ik ga dit product 0 0 0 0 0 0 0 Ik ga dit
nooit kopen product zeker
Kopen
Ik heb de intentie 0 0 0 0 0 0 0 Ik heb niet de
om dit product te intentie om
kopen dit product te
41 In het kopen van dit geïnteresseerd
product in het kopen
van dit
product
Hieronder volgt het tweede reclameblokje. Ik wil u vragen deze te bekijken en vervolgens onderstaande vragen te beantwoorden. (Porsche – Coca Cola – Nike)
3)
Ik wil u vragen bovenstaand merk te beoordelen op de volgende kenmerken. De vakjes geven de mate aan waarin u het merk beoordeelt op het kenmerk. (1 = erg lage
kwaliteit, 2 = redelijk lage kwaliteit, 3 = enigszins lage kwaliteit, 4 = neutraal, 5 = enigszins hoge kwaliteit, 6 = redelijk hoge kwaliteit, 7 = erg hoge kwaliteit)
Lage Kwaliteit 0 0 0 0 0 0 0 Hoge Kwaliteit
Onaantrekkelijk 0 0 0 0 0 0 0 Aantrekkelijk
Onplezierig 0 0 0 0 0 0 0 Plezierig
4)
42
Ik ga dit product 0 0 0 0 0 0 0 Ik ga dit
nooit kopen product zeker
Kopen
Ik heb de intentie 0 0 0 0 0 0 0 Ik heb niet de
om dit product te intentie om
kopen dit product te
kopen Ik ben geïnteresseerd 0 0 0 0 0 0 0 Ik ben niet In het kopen van dit geïnteresseerd
product in het kopen
van dit
product
Hieronder volgt het derde reclameblokje. Ik wil u vragen deze te bekijken en vervolgens onderstaande vragen te beantwoorden. (Iphone – Cartier – Sony)
5)
Ik wil u vragen bovenstaand merk te beoordelen op de volgende kenmerken. De vakjes geven de mate aan waarin u het merk beoordeelt op het kenmerk. (1 = erg lage
43
Lage Kwaliteit 0 0 0 0 0 0 0 Hoge Kwaliteit
Onaantrekkelijk 0 0 0 0 0 0 0 Aantrekkelijk
Onplezierig 0 0 0 0 0 0 0 Plezierig
6)
Bij het invullen van de volgende vragen wil ik u vragen ervan uit te gaan dat het product verkrijgbaar is bij een winkel bij u in de buurt. Ook deze vragen zijn gericht op het product dat hierboven is weergegeven.
Ik ga dit product 0 0 0 0 0 0 0 Ik ga dit
nooit kopen product zeker
Kopen
Ik heb de intentie 0 0 0 0 0 0 0 Ik heb niet de
om dit product te intentie om
kopen dit product te
kopen Ik ben geïnteresseerd 0 0 0 0 0 0 0 Ik ben niet In het kopen van dit geïnteresseerd
product in het kopen
van dit
product
44 7)
Ik wil u vragen bovenstaand merk te beoordelen op de volgende kenmerken. De vakjes geven de mate aan waarin u het merk beoordeelt op het kenmerk. (1 = erg lage
kwaliteit, 2 = redelijk lage kwaliteit, 3 = enigszins lage kwaliteit, 4 = neutraal, 5 = enigszins hoge kwaliteit, 6 = redelijk hoge kwaliteit, 7 = erg hoge kwaliteit)
Lage Kwaliteit 0 0 0 0 0 0 0 Hoge Kwaliteit
Onaantrekkelijk 0 0 0 0 0 0 0 Aantrekkelijk
Onplezierig 0 0 0 0 0 0 0 Plezierig
8)
Bij het invullen van de volgende vragen wil ik u vragen ervan uit te gaan dat het product verkrijgbaar is bij een winkel bij u in de buurt. Ook deze vragen zijn gericht op het product dat hierboven is weergegeven.
Ik ga dit product 0 0 0 0 0 0 0 Ik ga dit
nooit kopen product zeker
Kopen
Ik heb de intentie 0 0 0 0 0 0 0 Ik heb niet de
om dit product te intentie om
kopen dit product te
kopen Ik ben geïnteresseerd 0 0 0 0 0 0 0 Ik ben niet In het kopen van dit geïnteresseerd
45 van dit
product
9)
Geef aan of u volgende merken gezien hebt in de commercial blokje die u zojuist gezien heeft Wel Niet Heineken 0 0 Adidas 0 0 Pretz 0 0 Volvo 0 0 Coca Cola 0 0
Ik wil u vragen ook de volgende twee commercials te bekijken en onderstaande vragen te beantwoorden. Beide commercials duren ongeveer 15 seconden.
Hier volgt het eerste filmpje (Pretz modest Arousal)
10)
Wat vindt u van bovenstaande commercial?
0 Grappig
46 0 Irritant
0 Geen van bovenstaande antwoorden
11)
Elk paar woorden geeft een gevoelsdimensie aan. Sommige kunnen wellicht ongebruikelijk lijken, maar over het algemeen ervaart u bij het zien van het filmpje waarschijnlijk meer het ene dan het andere gevoel. Ik wil u dan ook vragen welk gevoel u ervaart na het zien van het bovenstaand filmpje.
De rondjes geven aan in hoeverre u de emotie ervaart bij het zien van het filmpje. De schaal vormt een tegenstelling dus: 1 = heel erg ontspannend, 2 = behoorlijk
ontspannend, 3 = een beetje ontspannend, 4 = neutraal, 5 = een beetje stimulerend, 6 = behoorlijk stimulerend 7 = heel erg stimulerend
Ontspannend 0 0 0 0 0 0 0 Stimulerend
Traag 0 0 0 0 0 0 0 Snel
Saai 0 0 0 0 0 0 0 Gejaagd
Slaperig 0 0 0 0 0 0 0 Klaarwakker
Hier volgt het tweede filmpje (Pretz High Arousal)
12)
47 0 Grappig
0 Slaapverwekkend 0 Irritant
0 Geen van bovenstaande antwoorden
13)
Elk paar woorden geeft een gevoelsdimensie aan. Sommige kunnen wellicht ongebruikelijk lijken, maar over het algemeen ervaart u bij het zien van het filmpje waarschijnlijk meer het ene dan het andere gevoel. Ik wil u dan ook vragen welk gevoel u ervaart na het zien van het bovenstaand filmpje.
De rondjes geven aan in hoeverre u de emotie ervaart bij het zien van het filmpje. De schaal vormt een tegenstelling dus: 1 = heel erg ontspannend, 2 = behoorlijk
ontspannend, 3 = een beetje ontspannend, 4 = neutraal, 5 = een beetje stimulerend, 6 = behoorlijk stimulerend 7 = heel erg stimulerend
Ontspannend 0 0 0 0 0 0 0 Stimulerend
Traag 0 0 0 0 0 0 0 Snel
Saai 0 0 0 0 0 0 0 Gejaagd
Slaperig 0 0 0 0 0 0 0 Klaarwakker
14)
Kende u voor deelname aan dit onderzoek het merk Pretz?
48 15)
Hieronder volgend nog een aantal algemene vragen
Helemaal niet Heel erg
Ik sta open voor nieuwe culturen 0 0 0 0 0 0 0 Ik ben altijd op de hoogte van wat 0 0 0 0 0 0 0 er gebeurt in de wereld
Ik sta open voor nieuwe ervaringen 0 0 0 0 0 0 0 Nieuwe dingen vind ik fascinerend 0 0 0 0 0 0 0
Mensen zien mij als een 0 0 0 0 0 0 0
onvoorspelbaar persoon
Ik verkies een stabiel leven boven 0 0 0 0 0 0 0 een onvoorspelbaar leven vol
veranderingen
Ik houd van verrassingen 0 0 0 0 0 0 0
Als kind dacht ik er vaak over na 0 0 0 0 0 0 0 om de wereld te verkennen
Ik houd van nieuwe dingen en 0 0 0 0 0 0 0
veranderingen in mijn dagelijkse routine
Ik vind het leuk om taken uit te 0 0 0 0 0 0 0 voeren waarbij ik oplossingen voor
problemen kan bedenken
Ik voer liever een complexe, 0 0 0 0 0 0 0
moeilijke taak uit, dan dat ik een makkelijke taak uitvoer.
Ik vind het moeilijk goed na te 0 0 0 0 0 0 0 blijven denken onder druk
Ik houd van een leven vol puzzels 0 0 0 0 0 0 0 die ik moet oplossen
Ik houd ervan te experimenteren 0 0 0 0 0 0 0 met nieuwe dingen
Ik houd ervan nieuwe verschillende 0 0 0 0 0 0 0 dingen te doen
Wanneer ik een nieuw merk zie in 0 0 0 0 0 0 0 de winkel, koop ik het vaak om het
te proberen
Wat is uw geslacht?
0 Man
0 Vrouw
49 Wat is uw opleidingsniveau? 0 Geen opleiding 0 Bassischool 0 VMBO / Mavo 0 HAVO / VWO 0 MBO 0 HBO 0 Universiteit
Uit hoeveel personen bestaat uw huishouden (een studentenhuis wordt niet als een huishouden gezien) 0 1 Persoon 0 2 Personen 0 3 Personen 0 4 Personen 0 5 Personen
50
Appendix 2; extra analyses moderator effect
Chapter 1: introduction
1.1 Introduction
Within the report no significant effects of repeatedly showing the low arousal annoying commercial were found. Further, also the moderating effect of level of arousal on the effect between number of exposures and brand attitude was insignificant. However, for high arousal evoking annoying commercials a negative effect was found between repeated exposure and brand attitude. This should be a warning for advertisers,
because it shows that this type of commercials does not have the effect that advertisers intend to achieve. Instead, it leads to a negative evaluation of the brand that is
advertised. However, it is still unclear whether the repeated exposures of highly arousal evoking commercials also results in less sales of the advertised products. Therefore, this extension will focus on the effect of repeated exposure of high arousal annoying
commercials on purchase intention of consumers.
Because there is a negative relationship between number of exposures and brand attitude, it is expected that number of exposures also negatively influences purchase intention. Furthermore, it is possible that consumers with different personality traits react differently towards the repeated exposure of highly arousing commercial.
51 positively influenced by novelty seeking behavior. This implies that the negative effect of repeated exposure on purchase intention is stronger for individuals who score high on novelty seeking behavior.
This results in the following model:
Figure 1: Conceptual model including moderator
1.2 Structure of the extension
The following paragraph discusses the variables and the selection of cases that are suitable for the analysis. Further, in chapter two the simple regression analysis of number of exposures on purchase intention is conducted. In chapter three several assumptions are tested. Furthermore, the moderator variable novelty seeking is added to the model in order to test whether this moderator variable significantly influences the relation between number of exposures and purchase intention. Further, a
conclusion of the findings of this extension is given.
1.3 Description of variables
52 1.3.1 Number of exposures
The high arousal commercial is shown 1,2,3 or 4 times.
1.3.2 Novelty seeking
In order to measure novelty seeking, three items were measured on a 7 point Likert scale, anchored by; not agree at all - totally agree. The following items were measured (Bearden, Netemeyer and Haws, 2011):
I like to try new products and services I enjoy looking for new experiences
When I go shopping, I enjoy buying new products and brands
1.3.3 Purchase intention
In order to measure purchase intention, respondents were asked to rate the test-brand on a three-item differential scale that was developed by Spears and Singh (2004). The items that were used to measure purchase intention were measured on a 7-point Likert scale and anchored by; never – certainly, intention – no intention, interested – not interested.
1.4 Creating new variables
Although previous research has proven that the questions that were used measure the constructs of novelty seeking and purchase intention, reliability analyses were
53 questions actually measured the constructs of novelty seeking and purchase intention. Hence, the questions were averaged in order to create the variables purchase intention and novelty seeking.
1.5 High arousal condition
Because this analysis is focused on testing the effects of the high arousal condition, cases were selected that were actually exposed towards the high arousal evoking commercial. A total of 88 respondents were exposed towards the high arousal evoking commercial, and the results of these participants were analyzed in the rest of this appendix.
Chapter 2: Simple regression analysis
First of all, in this chapter the principles of simple regression are explained. Afterwards, a regression model is conducted in which the simple regression equation is tested.
2.1.1 Simple model
When a regression is conducted, the following model is conducted:
predicted outcome = β0 + β1X1 + ε
β0 = intercept
54 2.1.2 Simple regression for the first hypothesis
As described before, the most important finding of the original experiment was that number of exposures negatively influenced brand attitude for the high arousal condition. Intuitively, it is expected that if a negative relationship exists between number of exposures and brand attitude, there is probably also a negative relationship between number of exposures and purchase intention.
Consequently, the following hypothesis is stated:
H1: The number of exposures of a high arousal evoking commercial negatively influences purchase intention.
A simple regression analysis was conducted and showed a negative significant effect between number of exposures and brand attitude (β=-0.432, t = -3.038, p = 0.003 ). The results of the simple regression can be found in table one. Proof was provided that the number of exposures of a high arousal evoking commercial negatively influences purchase intention. When an individual is exposed an extra time towards a commercial, purchase intention will decrease with 0.432. The p-value of simple regression analysis shows whether or not the predictor variable (number of exposures) significantly influences the dependent variable (purchase intention). The p-value in this model is 0.003 which is smaller than the critical p-value of 0.05(Field, 2010). Hence, it can be concluded that the predictor variable is significantly influencing the dependent variable (Field, 2010). Consequently, the first hypothesis is accepted.
55 The regression analyses resulted in the following equation:
Purchase intention = 3.806 -0.432 Number of exposures + ε
2.2 Assessing goodness of fit
As described above, the regression model is estimated through using the method of ordinary least squares. Based on the variables that are included in the model, the best fitting regression equation is conducted. However, the variables that are included might not be fully responsible for determining the dependent variable. In this example, only the number of exposures is included in the model as predicting variable of purchase intention. However, there might be more variables that can be accountable for the variation in purchase intention. Therefore, it should be investigated how much of the variation in purchase intention is explained by number of exposures. This can be done by looking at the R square of the regression model. The R square for the regression model is 0.098, which means that number of exposures accounts for 9.8 per cent of the variation in purchase intention (Field, 2010). This implies that there are other variables that also influence purchase intention. In the following chapter the variable novelty seeking will be addedin order to improve the model.
Chapter 3: expanding the model
3.1 Description of new model
In order to improve the model the variable novelty seeking is included in the model. Like described in the introduction, intuitively it is expected that novelty seeking has a
56 repetition. Hence, it is expected that novelty seeking positively influences the negative relationship between number of exposures and purchase intention. As such, it is expected that the negative effect of repeated exposures on purchase intention is stronger for people who score high on novelty seeking. Consequently, the following hypothesis is stated:
H2: Novelty seeking positively influences the relationship between number of exposures and purchase intention.
3.2 Computing new variable
In order to test whether the relationship between number of exposures and purchase intention is moderated by the variable novelty seeking, an interaction variable between number of exposures and novelty seeking is computed.
3.3 Improvement of the model
Firstly, the improvement of the model is tested. Before conducting the regression, the independent variables number exposures and novelty seeking are included. The second step also includes the interaction between these two variables.
57 more variance when the variable novelty seeking is added towards the basic model.
3.4.1 Multicollinearity
An important assumption that has to be checked before performing a regression analysis is that the independent variables included in the model are not correlated with each other. If they are significantly correlated with each other multicollinearity is said to be present. Multicollinearity implies that there is a direct relationship between some or all of the independent variables. If this is present in the model, it is difficult to assess the separate effects of independent variables on the dependent variable because this effect is partly caused by other independent variables. Collinearity statistics provided by the regression analysis are used to check if multicollinearity is present.
The focal values indicating multicollinearity are the VIF-scores of each of the
independent variables. VIF-scores are preferred over simple collinearity coefficients because VIF-scores indicate the collinearity coefficient of an independent variable over all other independent variables, whereas the simple collinearity coefficient only indicate the degree of collinearity between two independent variables (on a one-to-one basis, as opposed to a one-to-all based used in calculating the VIF-scores). As such VIF-scores provide a more complete picture of the presence (or absence) of multicollinearity. The VIF-scores indicate high multicollinearity of the interaction variable and variable
exposure (VIF < 10), and moderately high multicollinearity of variable novelty seeking (4 < VIF < 10) (Field, 2010). The correlation table shows that only variables exposure and novelty seeking are not significantly correlated. The exact values of the VIF-scores of the different variables can be found in table 2.
