How do you like to be shocked? : a quantitative online experiment into shock appeals, elaboration and suprise

(1)

HOW DO YOU LIKE TO

BE SHOCKED?

A quantitative online experiment into shock appeals,

elaboration and suprise.

MASTER THESIS PERSUASIVE COMMUNICATION

Nesrine Belayachi

Student number: 10627774 Supervisor: Dr. S.C.M. Welten Date: 16-06-2014

(2)

1

ABSTRACT

Keywords: SHOCK APPEALS – ADVERTISING – DISGUST APPEAL - MORAL OFFENSE

APPEAL- ELABORATION.

Shock appeals are emotional appeals that intentionally aim to shock or offend an audience in order to attract a high level of consumer attention. The aim of this study was to examine the differences in short-term effects that different types of shock appeals have on advertisement effectiveness; ad attitude, brand attitude, voluntarily ad exposure and purchase intention. Furthermore the study aimed to identify whether elaboration and suprise play a mediating role in this relationship.

Hypotheses were tested at the hand of a quantitative online experiment with 262 participants. The participants were asked to look at fictitious advertisements with one of three types of shock: a disgust appeal, a moral offense appeal or a control appeal. After which they completed a measure for elaboration followed by measures for suprise, advertisement attitude, brand attitude and purchase intention.

The results did not support the mediation model with elaboration and suprise. Furthermore the levels of shock appeal did not show differences for brand attitude, willingness to pay, purchase intention and voluntary exposure the ad.

(3)

2

INTRODUCTION

“I do not understand why a group of strangers have just brought me to tears in order to sell me a car” (Brown, 2013, Para. 5). Hyundai received many similar outraged reactions about their viral video featuring a man trying to commit suicide but failing because of the car’s 100% water emission. Hyundai used a moral offense appeal in their ad (Dahl, Frankenberger & Manchanda, 2003). A moral offense appeal is one of the seven types of shock-appeals; others are disgusting images, sexual references, profanity/obscenity, vulgarity, impropriety and religious taboos (Dahl et al., 2003). Shock appeals are a type of emotional advertising where one intentionally aims to shock its audience. Since people are constantly exposed to an excessive amount of advertisements each day, it requires more effort to get an advertisement noticed (Sandicki, 2011). The use of shock appeals is becoming wide-spread as there are strong indications that shock appeals significantly increase the audience’s awareness, recall and might positively influence short-term behavior (Bello, Pitts & Etzel, 1983; Dahl & Frankenberger, 2002; Parry, Jones, Stern & Robinson, 2013).

Regardless, current communications research does not offer clear insights on which type of shock appeal should be used in certain situations and why. This is a problem because using shock appeals is a risky strategy that can cause substantial brand damage as has

happened to Hyundai in the mentioned example (Brown, 2013; West & Sargeant, 2004). The present study aims to provide further scientific understanding of the concept of shock appeals in two main ways. Firstly by observing if there are short-term differences between the types of shock when it comes to their advertising effectiveness. Secondly this study aims to provide further scientific understanding of the concept of shock appeals by looking at the underlying

(5)

4 cognitive and affective processes that might mediate the relation of shock appeals with

advertising effectiveness.

It is relevant to study the differences in advertising effectiveness of the different types of shock appeals. This decade the use of shock appeals in marketing is at an all-time high, yet in general there is little quantifiable research regarding the differences between the types of shock (Noel, 2010). There is especially a lack of quantifiable research when it comes to the advertising effectiveness of the shock appeals. It has strong scientific value to conduct a study with a quantitative approach in order to clearly compare the different types of shock appeals with each other. Currently shock appeals do not have a prominent place in communication research, shock appeals are often mischaracterized in research as fear appeals (Dillard, 1994). Focusing on quantifying the advertisement effects of shock appeals might increase the

validity of future communication research concerning shock appeals. People might come to the conclusion that shock appeals are different emotional appeals than fear appeals.

Furthermore quantitative research allows us to apply the significant findings to a larger population and thus be able to create more precise and efficient marketing strategies. Identifying differences between the types of shock will also help marketers make better choices regarding which type of shock appeal they should use for what type of product.

Secondly, it is relevant to study the different underlying mediating processes that facilitate shock appeals. There is little information about the underlying cognitive and affective processes that an audience experiences when being subjected to shock appeals. Identifying the type of underlying conative processes of shock appeals has high scientific value as it helps determine whether there are different types of processes at play. Dependent on the findings it could indicate that the current categorization of shock appeals needs to be revised. If the types of shock appeals do not have the same underlying mechanism it suggests that the different types of shock-appeals might work in a different way. Subsequently the

(6)

5 current categorization and conceptualization of the shock appeals could be improved.

Identifying the underlying process also has practical importance; it will increase the effectiveness of campaigns that contain shock appeals. As it would become clear how

different types of shock appeals are processed and thus in which situation they would prove to be more effective.

To conclude the aim of this study is twofold. The first part is to examine and compare the effects of shock appeals on short-term measures of advertisement effectiveness. The second part is the intention is to identify the underlying cognitive and affective processes that mediate the relationship of shock appeals and advertising effectiveness. Leading to the following two research questions:

RQ1: To what extent is there a difference between the influences of the types of shock appeals on measures of advertising effectiveness (i.e., voluntarily exposure, ad attitude, brand attitude and purchase intention)?

RQ2: To what extent do underlying cognitive processes (elaboration) and affective processes (suprise) play a mediating role in the relationship of shock appeals with measures of advertising effectiveness (i.e., voluntarily exposure, ad attitude, brand attitude and purchase intention)?

(7)

6

THEORETICAL FRAMEWORK

Shock appeals

Shock appeals are emotional appeals that intentionally aim to shock or offend an audience (Sandicki, 2011). Shock appeals are considered shocking due to the advertisement elements, rather than the advertised product (Krstic, 2007). Shock appeals are primarily used for two reasons, firstly to attract a high level of consumer attention. Secondly to generate free publicity, this happens through the buzz controversial advertisements create (Gardner, 1985; West & Sargeant, 2004). Anderson and Petterson (2004) have identified three core

components of shock appeals. The first component is distinctiveness. Distinctiveness is a difficult component to accomplish due to the high level of daily exposure to advertisements. An advertisement has to be perceived as distinct in order to attract attention. If an

advertisement is not unique enough, it will not be able to trigger audience effects. The second component is ambiguity. Ambiguity refers to the degree to which the advertisement, or the intent of the advertiser, can be interpreted in different ways. Ambiguous messages are

beneficial to keep the attention of the target audience, because ambiguous messages are more likely to be elaborately processed (Belayachi, 2013; Krstic, 2007). The third component is the transgression of norms and taboos. A shock appeal deliberately goes against shared ideals and societal values. Violating norms and taboos creates a suprise reaction. According to Dahl et al. (2003) suprise is what triggers the experience of shock. Shock appeals are surprising as they differ from what people expect (Belayachi, 2013; Anderson & Petterson, 2004). Shock appeals are emotional appeals rather than informational appeals because an emotion, suprise,

(8)

7 is the underlying mechanism of shock (Krstic, 2007). Dahl et al. (2012) identified seven types of shock appeals namely disgusting images, sexual references, profanity/obscenity, vulgarity, impropriety, moral offensiveness and religious taboos.

There are several suggestions that the findings of Anderson and Petterson (2004) do not equally apply to all types of shock appeals. Not all of the shock appeals have an equal perceived transgression of norms. Firstly the findings of Anderson and Petterson (2004) seem to fit with the appeal of moral offensiveness where the shock stems from suprise that is created due to the transgressions of morality norms. Of the seven shock appeals the moral offense appeal is expected to contain the highest norm violation, while disgust appeals has the lowest norm violation (Dahl et al., 2013). Suprise depends on the transgressions of norms, and suprise is the main facilitator of shock (Dahl et al., 2003). If disgust has a lower transgression of norms there probably also is a lower level of suprise, indicating a lower level of shock. Therefore we can state that moral offense appeals in all probability work differently than disgust appeals. Looking at the differences between the disgust appeal and moral offense appeal is therefore the most interesting. Hence in operationalization we will focus on comparing moral offense appeals with disgust appeals.

Moral offensiveness

The moral offensiveness appeal refers to an unconformity to established rules and patterns that are accepted in society. Dahl et al. (p199, 2003) operationalizes moral offensiveness as harming innocent people/animals, gratuitous violence or sex, alluding to people or objects that provoke violence (e.g., Hitler), violating standards for fair behavior (e.g., shooting a person in the back), putting children in provocative situations (e.g., sexual, violent), victim exploitation.

(9)

8 Disgust appeals

Dahl et al. (p199, 2003) have operationalized the disgusting images appeal, as images with references to blood, body parts, secretions, orifices, urinary/faecal, gases, odors, disease, parasites, bodily harm (e.g., dismemberment), death and decay.

Underlying processes

The elaboration likelihood model is a much discussed theory that focuses on the development and change of attitudes. The ELM differentiates between two routes to persuasion. The first route is the central route or the cognitive route in this case, where the audience actively and logically reflects on an idea. Levels of elaboration are usually high. The second route is the peripheral route, or in our case the affective route, is where the audience uses formerly established ideas and superficial “thinking” in order to process stimuli. When information is processed it does not involve high levels of elaboration (Perloff, 2010). Here we will outline the proposed underlying cognitive and affective processes of shock (see fig 1).

Cognitive process: Elaboration

High levels of norm transgression lead to high levels of elaboration (Loef & Verlegh, 2002). Moral offensiveness has the highest level of norm transgression. A perceived norm

transgression is incongruent with what people expect to see in an advertisement. Therefore the higher the norm transgression is, the higher the perceived incongruence. A perceived

(10)

9 incongruence leads to a high level of elaboration (MacInnis, Moorman & Jaworski, 1991). Therefore appeals with higher norm transgressions lead to higher levels of elaboration in comparison with appeals with lower norm transgressions. If the audience resolves the factor of incongruence, the exposure of consumers to these advertisements leads to higher

advertisement attitudes. An audience with high levels of elaboration devotes more cognitive resources to processing the advertisement (Huhmann & Mott‐Stenerson, 2008). Therefore audiences with high levels of elaboration will likely comprehended messages with shock appeals better. Comprehension of the incongruence in turn leads to more positive advertisement attitudes than non-comprehension (Loef & Verlegh, 2002).

Disgust appeals also have incongruent elements. However the strong emotion of disgust prevents high levels of elaboration. The feeling of disgust triggers a reflex of avoidance. People will likely detach themselves physically and mentally from the image (Nabi, 1998). The avoidance reflex does not allow for high levels of elaboration. Furthermore as the feeling of disgust is a mental reflex it suggests that disgust appeals are processed predominantly through the heuristic route and thus it is expected that there is a low-level of elaboration (Nabi, 1998). This leads to the following hypotheses.

H1a: Disgust appeals lead to less elaboration in comparison to Moral offense appeals.

(11)

10 Affective process: Suprise

Shock appeals work with suprise as the main emotion (Dahl, et al., 2003). Suprise is one of the basic human emotions (Scherer & Ekman, 1984). Suprise is a short-lived neutral emotion that varies in levels of intensity, from little-suprise to very-suprised. The intensity of suprise determines the reaction to the stimuli (Lindgreen & Vanhamme, 2003). The intensity of suprise is determined by the level of incongruence, this refers to the level of unexpectedness of what people are subjected to (Loef & Verlegh, 2002). In the case of shock appeals it refers to the unexpectedness of the advertisements used. Advertisements with elements that are highly incongruent with the expectations of the audience, such as moral offensiveness and disgust appeals, lead to high levels of suprise intensity. As a neutral emotion, suprise influences the emotions that closely follow it. Thus suprise, starts at neutral, can remain neutral or become positive or negative trough heuristic cues. If no emotions are quickly formed suprise has no effect (Lindgreen & Vanhamme, 2003)

Moral offensiveness is the shock appeal with the highest transgression of norms in comparison to the other shock appeals. A high transgression of norms is unexpected with thus would lead to a high level of suprise. Yet moral offensiveness does not carry another emotion. Thus the affective process of moral offensiveness is only facilitated by suprise. The other emotions and judgments are formed through the cognitive process (Huhmann & Mott‐ Stenerson, 2008). Suggesting that moral offensiveness is not strongly mediated through suprise as there are no emotions formed through the affective process.

Disgust appeals in contrast with moral offense appeals, work through disgust as a second emotion. When people are subjected to a disgust appeal they are suprised at first, yet disgusted almost immediately after. The feeling of disgust is a protection mechanism; elicitors of disgust evoke a reflex of distaste and nausea. It is a subjective feeling of revulsion, it evokes a predisposition to avoidance. The emotion of disgust is a reflex and thus we can state

(12)

11 that it works through heuristic cues (Nabi, 1998). After the experienced suprise the emotion of disgust is heightened. As a result disgust appeals lead to a higher level of reactance in

comparison to moral offensiveness (Leshner, Bolls, & Thomas, 2009). Furthermore the audience will have negative attitude towards the advertisement trough affect transfer due to the negative experienced emotion of disgust (Chaudhuri & Buck, 1995). This leads to the following hypotheses.

H2a: Disgust appeals lead to a higher level of suprise than moral offense appeals.

H2b: The effect of Shock on persuasion is mediated by suprise.

IV Shock: 0-Control 1-Disgust 2-Moral Mediation 1: Cognitive process (Elaboration): -Low -High DV’s:  Time of exposure  Ad attitude  Brand Attitude  Purchase intention

Figure 1: Conceptual model Mediation 2: Affective

process (Suprise): -Low -High

(13)

12

Advertising effectiveness

Shock appeals are used in advertisements to persuade the audience. When referring to advertising effectiveness, we are referring to the power of an advertisement to persuade the audience to change their beliefs and behavior (Mehta, 2000). There are several ways

advertising effectiveness can be measured. By observing how long the audience looks at the advertisement, their attitude towards the advertisement, their attitude towards the brand, how much people are willing to pay for the product and their explicit intention to purchase the product (Lavidge & Steiner, 1961). Firstly measuring how long people are willing to

voluntarily spend looking at an advertisement gives an indication as to how interested people are in the given ad (Krugman, Fox, Fletcher, Fischer, & Rojas, 1994). Secondly, the attitude towards the advertisement is a strong influencer of attitude towards a product. Attitude towards the advertisement is also an important factor when it comes to forming a purchase intention (Lavidge & Steiner, 1961). Thirdly, Brand attitudes are important as they measure how much people like a company, consumers show more loyalty to a brand if they have a more positive attitude (Dick & Basu, 1994). Fourthly, the amount people are willing to pay is also important as it indicate how much the product is “worth” in their eyes. Fifth and lastly the intention to purchase is the single most important indicator of behavioral change (Follows & Jobber, 2000).

Effects on voluntarily exposure

Once people pay attention to an advertisement, they process the advertisement. After which attitudes about the advertisement and brand are formed and the ad is stored in their memory. These attitudes in turn lead to action, such as purchase of a product (Lavidge & Steiner,

(14)

13 1961). However before a person can pay attention and process an ad he first has to look at the ad and see the ad (Ning, 2001). Voluntary exposure refers to the time people are willing to “expose” themselves to the advertisement before they look away or turn the page. It is important to observe how long people expose themselves to an advertisement as it indicates how much attention they will pay to the ad, but also how elaborately they will process the ad (Ning, 2001).

Shock appeals in general lead to high levels of attention and exposure in comparison with informational appeals and other types of emotional appeals (Dahl et al.,2003) However Nabi (1998) states that this does not apply for disgust appeals. Disgust appeals lead to a predisposition to avoidance. Disgusting images motivate people to turn away. Thus people do not look at a disgust appeal for long, this leads to less exposure. We can expect the audience to spend a longer time looking at a moral offense appeal than a disgust appeal. This leads to the following hypothesis:

H3: Disgust appeals lead to lower times of voluntary exposure in comparison to Moral offense appeals.

Effects on advertising attitude

Attitude is defined as an individual's internal evaluation of an advertisement (Spears & Singh, 2004). Shocking advertisements can stir up strong negative feelings among the consumers. An audience might form negative attitudes towards advertisements that contain shock appeals (Dahl et al., 2003; Parry et al., 2013). Emotional appeals that work through a low-elaboration route, such as disgust appeals, will likely cause the audience to experience negative feelings as affect transfer takes place (Chaudhuri & Buck, 1995). Affect transfer is when the emotion experienced during the ad will transfer towards the attitude of the ad. Thus people develop

(15)

14 negative attitudes towards advertisements when those advertisements provoke negative

feelings (Spears & Singh, 2004). When people are exposed to disgust appeals, they will most likely create a negative attitude towards the advertisement because disgust appeals provoke disgust, a negative feeling (Donovan, Henley, 1997).

Moral offensiveness on the other hand probably works through a high level of elaboration. Moral offensiveness works through a high level of norm transgression, thus audiences experience advertisements with moral offensiveness to have high incongruences with their expectations. This leads to high levels of audience elaboration as they aim to “resolve” the experienced incongruence. Petty et al. (1991) state that affect transfers also occur with messages with a high level of elaboration. However Huhmann and Mott‐Stenerson (2008) found that advertisements that stimulate high levels of elaboration and contain highly incongruent elements lead to positive attitudes towards the advertisement regardless of the negative feelings experienced. Furthermore if people are able to resolve the incongruence, thus in this case understand the clue used in the moral offense appeal, it would lead to a more positive attitude towards the advertisement. This indicates that people would experience a more negative attitude towards the advertisement if they would be exposed to a disgust appeal rather than a moral offense appeal. This leads to the following hypothesis:

H4: Disgust appeals lead to more negative attitudes towards the advertisement then Moral offense appeals.

Effects on brand attitude

Spears & Singh (2004) define attitude toward the brand as a personal evaluation towards a brand. Past research has found a relationship between advertisements and brand attitudes under a variety of conditions. Brand attitude is closely linked to the attitude towards the

(16)

15 advertisement (Gardner, 1985; MacKenzie & Lutz, 1989). When an advertisement is more positively evaluated, the brand is usually more positively evaluated as well and vice versa. The literature predicts that disgust appeals will lead to a more negative attitude towards the advertising than moral offense appeals. As affect transfer takes place the attitude towards the ad will transfer to the attitude towards the brand (Chaudhuri & Buck, 1995).We can expect that people who are exposed to a disgust appeal will likely have a more negative attitude toward the brand associated with the advertisement than people who would be exposed to a moral offense appeal. This leads to the following hypothesis:

H5: Disgust appeals lead to more negative attitudes toward the brand Bullit than Moral offense appeals.

Effects on purchase intention

Purchase intention refers to the level of likelihood of making a purchase. Whereas attitudes are summary evaluations, intentions represent an explicit motivation plan to carry out

behavior. Purchase intention can be defined as the intention of an individual to buy a specific product or brand. A behavioral intention reflects a person’s decision to perform certain behavior. Purchase intention is the best predictor of buying behavior (Fishbein & Ajzen, 1975). Ad attitude is generally a relatively good predictor of ad attitude. However Bello, Pitts & Etzel, (1983) found that purchase intention is not related to brand attitude and ad attitude when people are offended. They state that the audience might even be lightly amused by the ad they still could form strong reservations regarding buying the product. They found that the brands that purposefully use offensive content were evaluated as negative. People often would not purchase products from a brand that uses these types of appeal out of principle. This

(17)

16 suggests that people will less likely buy a product that was advertised with a moral offense appeal than a product that is advertised with a disgust appeal. This is the case since moral offence appeals are per definition offensive while disgust appeals generally are not offensive. This also suggests that people are willing to pay more for ads containing disgust appeals rather than shock appeals. This leads to the following hypotheses:

H6a: Disgust appeals lead higher purchase intentions in comparison to Moral offense appeals.

H6b: Disgust appeals lead to higher willingness to pay in comparison to Moral offense appeals.

METHODOLOGY

Study design and procedure

The research sample for the quantitative online experiment consisted of 262 participants. The participants were randomly assigned to one of the three shock appeal conditions; Control group (29.77 %), Disgust appeal (34 %), Moral offense appeal (36.3 %). The age of the participants ranged between 18 and 68 years (Mage = 23.42, SD = 6.01). The participants were currently living in 22 different countries. 39.6 % male (n = 103) 60 % female (n = 156) and 0.4 % other (n =1) participants took part in the survey. 48.9 % of the participants had a university degree or higher, 20.4 % had a college degree, 9.2 % had a practical collage degree, 20 % finished High School and 1.2 % had no high school degree.

The participants were collected online, over the course of one week, as most shocking advertisements also are prevalent in the online sphere (Eckler & Bolls, 2011). In order to

(18)

17 encourage participation and to avoid non-responses an incentive was offered for participating. Therefore, all participants were offered a chance to win one of several 15 euro gift certificates of choice. The survey was distributed predominantly through email, LinkedIn, Facebook and on an energy drink fan-forum (www.myenergydrinks.com). The participants were of a minimum age of 18 in order to avoid ethical complications by subjecting minors to explicit shocking content. Preceding the experiment there was a banner warning the participants of the explicit content. After which the participants were asked to look at a fictitious advertisement for the energy drink brand Bullit with either a disgust appeal, a moral offense appeal or a control condition(see Appendix A). Subsequently participants were asked to complete the measure for elaboration followed by measures for advertisement attitude, brand attitude, purchase intention and their willingness to pay. At the end of the survey some demographical questions about age, gender, education and country of residence were asked.

A total number of 128 questionnaires were deleted from a total of 390 applicants. 126 questionnaires were excluded from analysis due to incompletion, questionnaires were

considered incomplete when not all dependent measures were submitted. Two additional cases were excluded from the analysis as these participants had remarked they had not seen the ad because they accidentally clicked continue, the timer data confirmed these remarks, and as they belonged to the same group (control) it was appropriate to exclude this data.

Independent variables

Shock appeal. The participants were only subjected to an advertisement with either a

disgusting appeal, moral offensive appeal or the control appeal. Moral, disgust and control were transformed into one variable called shock.

(19)

18 Dependent variables

Voluntary exposure. The measure of voluntary exposure was derived from Wang (2009).

Participants were exposed to an advertisement with the following task “Please look at the following advertisement as if you were flipping through a magazine”. The time of exposure was measured with the use of a timing element that measured the amount of time in seconds that expired until a participant clicked on continue.

Ad attitude. The attitude scales were adapted from previous research (Malcheit & Wilson,

1988). Attitude towards the advertisement was measured with 5 different items on a 7-point Likert scale. The items the participants rated were:”I perceive the advertisement as

Bad/Good” , ”I perceive the advertisement as Negative/Positive” , ”I perceive the advertisement as Unlikable/Likable” , ”I perceive the advertisement as

Unfavorable/Favorable” and ”I perceive the advertisement as Unpleasant/Pleasant”. These items loaded on one factor (EV = 4.15; R2 =0.89; Cronbach’s Alpha = 0.95 that proofed to be reliable.

Brand attitude. Brand attitude towards Bullit is assessed using pre-existing scales from

previous research (Malcheit & Wilson, 1988). The five items were all measured using a 7-point Likert scale. The items the participants rated were:”I perceive the brand Bullit as

Bad/Good”, ”I perceive the brand Bullit as Negative/Positive”, ”I perceive the brand Bullit as Likable/Unlikable”, ”I perceive the brand Bullit as Unfavorable/Favorable” and “I perceive the brand Bullit as Unpleasant/Pleasant”. These items loaded on one factor (EV = 3.90; R2 =0.88; Cronbach’s Alpha = 0.93 that proofed to be reliable.

Purchase intention. Intention to purchase was measured in two ways, by looking at the

willingness to pay and their intent to purchase. Firstly the willingness to pay for Bullit was measured using Baker, Levy and Grewals’ (1992) “Willingness to buy” scale: a measure

(20)

19 asking the participants to “please indicate how many euros you would be willing to pay for a 250ml can of Bullit in the supermarket?” Where people could indicate how much they would they would be willing to spend. The intention to purchase was also measured using three items derived from Baker, Levy and Grewals (1992) all three items were measured on a 7-point Likert scale ranging from 1 (Strongly disagree) to 7 (Strongly agree): “If I would see Bullit in the supermarket I would buy it” , “I intend to buy a Bullit energy drink this month” And “The next time I intend to buy an energy drink I will probably buy Bullit” These items loaded on one factor (EV = 2.49; R2 =0.72; Cronbach’s Alpha = 0.90 that proofed to be reliable.

Mediation

Suprise. Suprise is measured on an explicit level trough a one item measurement derived from

Dunning (1993). The participants were asked to rate the following item: “I was suprised by the advertisement” which was rated on a 7-point Likert scale ranging from 1 (Strongly agree) to 7 (Strongly disagree).

Elaboration. The level of elaboration that the participants experienced was measured by using

Cacioppo and Petty’s (1981) Thought-Listing Technique. The participants were tasked with the following request “Please list every thought that went through your mind as you looked at

the ad. You can fill in as many or as few thoughts as you experienced.” The participants were

presented with 15 separate text entry lines. The number of filled in text entry’s represent the number of thoughts and thus indicate the level of elaboration, more thoughts is more

(21)

20 Control variables

In in order to increase the internal validity of the study several potential control variables were measured. The control variables in this study were measured by asking participant’s age, gender and education level. Prior knowledge was measured in this study by asking if the participants had prior knowledge of Bullit. Perceived quality of the ad was measured with 3 items all measured on a 7 point Likert scale: “I perceive the advertisement as

Unbiased/Biased” “I perceive the advertisement as Convincing/Unconvincing” and “I

perceive the advertisement as Professional/Unprofessional”. These items loaded on one factor (EV = 1.89; R2 =0.57; Cronbach’s Alpha = 0.60. After consideration “I perceive the

advertisement as Unbiased/Biased” was removed from the quality of ad scale due the low Cronbach alpha coefficient. After deletion Cronbach’s Alpha = 0.83 proofed to be reliable.

Manipulation check

Disgust appeal. As a manipulation check for the disgust appeal the following three items were

used, all the items were measured using a 7-point Likert scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). “The advertisement was difficult to look at”, “the advertisement is disgusting”, “and the advertisement is repugnant”. The factor analysis showed that all items combined created one factor for moral offense appeal (EV = 2.26, r2 =

.69; Cronbach’s Alpha = 0.84 that proofed to be reliable.

Moral offense appeal. As a manipulation check for the moral offense appeal the following

two items were used, all the items were measured using a 7-point Likert scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). “The advertisement is morally wrong” (M=4.12

(22)

21 correlation between the two variables (r = 0.85, n =262, p > 0.001).” The moral offense scale has good internal consistency, with a Cronbach alpha coefficient of (α=.916).

RESULTS

Randomization

ANOVA analyses were conducted in order to ensure that age, gender, education, year of birth, prior knowledge and perception of quality did not differ per condition (see fig 2). Year born, gender, education, country of residence and prior knowledge did not have a significant mean differences for the different experimental conditions. Therefore we do not have to control for these variables in future analysis.

Ad quality was not equal in all 3 experimental conditions (F (2,259) = 9.69, p < 0.001. A Tukey HSD post hoc test demonstrated that the mean of Ad quality differed significantly between the control group (Mdifference= - 0.79, p = 0.003) and the disgust appeal group .ad

quality also differed significantly between the control group and the moral offense appeal group (Mdifference= - 1.03, p < 0.001). Depending on the group means yet there are strong

indications that the validity of the items was low, they correlated strongly with attitude measures, therefore there will not be controlled for the perceived ad quality, the hypothesis will still be tested as planned.

(23)

22

Table 1: Randomisation checks participant characteristics per experimental condition.

Experimental condition

Control Disgust Moral offense F (2,259) Sig.

M (SD) M (SD) M (SD) Year born 70.65 (4.56) 70.76 (5.24) 69.90(7.59) 0.55 0.082 Gender 4.43 (1.41) 4.29 (1.74) 4.71 (1.46) 5.94 0.206 Education 4.22 (1.14) 3.80 (1.26) 3.91 (1.23) 2.62 0.075 Country of residence 123.94 (30.64) 124.71 (18.52) 125.77 (27.06) 0.11 0.898 Prior knowledge 1.28 (0.45) 1.15 (0.36) 1.25 (0.44) 2.54 0.080 Ad Quality 4.58a (1.69) 5.37b (4.47) 5.61b (1.54) 9.96 < .001

Note. Means per row with different letters differ significantly with p<0.05 in Tukey Post Hoc tests.

Manipulation check

In order to see if the manipulation was successful an ANOVA with a Tukey HSD post hoc test was conducted with the shock condition as independent variable and the measures of disgust and moral offensiveness as dependent variables (see table 2). For the moral offense appeal condition the Levene’s test for homogeneity of variances showed that we have violated the homogeneity of variance assumption. Therefore we will use Welch and Brown-Forsythe test and a Games-Howell post hoc test for the moral offense appeal condition.

Participants in the disgust appeal condition rated highest on the disgust measure (M = 4.64, SD=1.73), differing significantly with the control condition, while not with the moral offense appeal condition F (2,259) = 4.15, p = 0.017. Participants in the Moral Offense appeal condition (M = 4.04, SD=1.92) did not differ significantly with the control condition and the disgust appeal condition F (2,259) = 0.97, p = 0.376. The lack of significant

(24)

23 differences indicates that the manipulation check was unsuccessful. After careful deliberation the decision was made that the hypotheses will still be tested as planned and ad quality will not be included as covariate, for in-depth clarification see discussion section.

Table 2: Mean scores for evoked emotion per experimental condition.

Control Disgust Moral offense F (2,259) Sig. M (SD) M (SD) M (SD)

Disgust 3.88a (1.77) 4.64b (1.73) 4.03a,b(1.92) 4.15 0.017 Moral offense 3.97(1.97) 4.18 (1.77) 4.40 (2.38) 0.97 0.376

Bold means targeted variable.

Analyses

Elaboration

A one-way ANOVA with a Tukey HSD post hoc test was conducted with shock as independent variable and elaboration as dependent variable (see table 3). The ANOVA showed that there were no significant differences in elaboration between the three conditions: F = (2,259) = 0.33, p= 0.78. Tukey’s Post Hoc showed that the control condition (M = 3.88, SD = 1.97), does not differ significantly with the disgust appeal condition (M = 3.87, SD = 1.65, p = 1.000). The control condition (M = 3.88, SD = 1.97) also does not differ

(25)

24 significantly with the moral offense appeal condition (M = 4.07, SD = 1.91, p =0.781).

However the disgust appeal condition (M = 3.87, SD = 1.65) did not differ significantly from the moral offense appeal condition (M = 4.07, SD = 1.91, p = 0.750). There are no significant findings to support that hat disgust appeals lead to less elaboration in comparison to moral offense appeals, we therefore reject H1a.

Suprise

A one-way ANOVA with a Tukey HSD post hoc test was conducted with shock as

independent variable and suprise as dependent variable (see table 3). The ANOVA showed that there was a significant difference in suprise between the three conditions: F = (2,259) =

4.69, p= 0.010. Tukey’s Post Hoc showed that the control condition (M = 3.88, SD = 1.97),

differed significantly with the disgust appeal condition (M = 3.87, SD = 1.65, p = 0.009). The control condition (M = 3.88, SD = 1.97) however does not differ significantly with the moral offense appeal condition (M = 4.07, SD = 1.91, p = 0.597). And the disgust appeal condition

(M = 3.87, SD = 1.65) did not differ significantly from the moral offense appeal condition (M = 4.07, SD = 1.91, p = 0.089). Therefore there are no significant findings to support that disgust appeals lead to more negative attitudes towards the advertisement than moral offense appeals, we therefore reject H2a.

Table 3: ANOVA shock on Mediators

Control Disgust Moral offense F (2,259) Sig. M (SD) M (SD) M (SD)

Suprise 4.04a (1.78) 4.93b (1.94) 4.33a,b (2.01) 4.70 0.010 Elaboration 3.88(1.97) 3.88 (1.65) 4.07 (1.91) 0.33 0.718

(26)

25 Mediation Suprise and elaboration

To test whether the relationship of shock on advertising effectiveness is mediated by suprise and elaboration Hayes‘ and Preacher’s (2013) multiple mediation regression models

(INDIRECT) are used to estimate the mediation effects. The shock variable is recoded with the method of unweighted contrast coding.

Voluntary exposure. We looked at whether the effect of shock on voluntary exposure is

mediated by suprise and elaboration. Elaboration and suprise did not mediate the relation between the independent variable shock and the dependent variable voluntary exposure; b = 0.91, t = 0.88, p = 0.378. Results based on 10000 bootstrapped samples with a 95%

confidence also indicated an insignificant mediation effect of -0.14 as the interval range

contained 0.

Ad attitude. We looked at whether the effect of shock on ad attitudes is mediated by suprise

and elaboration. Elaboration and suprise appear to have a significant mediation on the relation between the independent variable shock and the dependent variable ad attitude; b = 1.14, t = 5.31, p < 0.001. However results based on 10000 bootstrapped samples with a 95%

confidence indicated an insignificant mediation effect of -0.01 as the interval range contained

0.

Brand attitude. We looked at whether the effect of shock on brand attitudes is mediated by

suprise and elaboration. Elaboration and suprise have no significant mediating role on the relation between the independent variable shock and the dependent variable brand attitude; b= - 0.31, t = -0.14, p = 0.886. The results based on 10000 bootstrapped samples confirmed with

a 95% confidence an insignificant mediation effect of -0.01 as the interval range contained 0.

Purchase intention. We looked at whether the effect of shock on purchase intention is

(27)

26 role on the relation between the independent variable shock and the dependent variable

purchase intention; b = -0.22, t=-1.03, p = 0.306. The results based on 10000 bootstrapped samples with a 95% confidence confirmed an insignificant mediation effect of

-0.008 as the interval range contained 0.

Willingness to pay. We looked at whether the effect of shock on willingness to pay is

mediated by suprise and elaboration. Elaboration and Suprise have no significant mediating role on the relation between the independent variable shock and the dependent variable Willingness to pay; b = -0.01, t = -013, p = 0.902. The results based on 10000 bootstrapped samples with a 95% confidence confirmed an insignificant mediation effect of -0.038 as the interval range contained 0.

The mediation regression models indicate that there is no statistical support to indicate that suprise and elaboration mediate the relationship between shock and advertisement

effectiveness, we therefore reject H1b. The mediation regression models indicate that there is no statistical support to indicate that the effect of shock on persuasion is mediated by

elaboration, we therefore reject H2b.

Voluntary exposure

A one-way ANOVA with a Tukey HSD post hoc test was conducted with shock as independent variable and voluntary exposure as dependent variable (see table 4). The

ANOVA showed that there were no significant differences in voluntary exposure between the three conditions: F (2,259) = 0.43, p = 0.649. Tukey’s Post Hoc showed that the voluntary exposure in the control condition (M= 12.66, SD=8.80) did not differ significantly with disgust appeal condition (M= 12.02, SD=8.90): p = 0.983. The voluntary exposure in the

(28)

27 control condition also does not differ significantly with the moral offense appeal condition (M

= 14.43, SD=18.94): p = 0.664. Lastly the voluntary exposure in the disgust appeal condition

did not differ significantly from voluntary exposure moral offense appeal condition. (M = 4.04), SD=1.92): p = 0.655. Therefore we find no statistical support that disgust appeals lead to lower times of voluntary exposure in comparison to moral offense appeals, thus we reject H3.

Table 4: ANOVA shock on DV’s

Control Disgust Moral offense F (2,259) Sig. M (SD) M (SD) M (SD) Voluntary exposure 12.66 (8.80) 12.02 (8.90) 14.43(18.94) 0.43 0.649 Ad attitudes 4.74a (1.47) 5.89b (1.26) 5.88b (1.45) 18.56 <0.001 Brand attitudes 4.07 (1.43) 4.67 (1.46) 5.01 (1.28) 1.66 0.192 Purchase intention 5.84 (1.37) 5.58 (18.52) 5.54 (7.78) 0.62 0.542 Willingness to pay 0.70 (0.61) 0.65 (0.51) 0.61 (0.54) 0.57 0.568

Ad attitudes

A one-way ANOVA with a Tukey HSD post hoc test was conducted with shock as

independent variable and ad attitude as dependent variable (see table 4). The ANOVA showed that there is a significant difference in ad attitude between the three conditions: F = (2,259) =

18.56, p= 0.00, η2 =0.13. The effect size indicates a medium to strong differences between the means. Tukey’s Post Hoc showed that the control condition (M = 4.74, SD = 1.47), differed significantly from the disgust appeal condition (M = 5.98, SD = 1.26, p > 0.001).

(29)

28 The control condition (M = 4.74, SD = 1,47) also does differed significantly with the moral offense appeal condition (M = 5.88, SD = 1.45, p > 0.001). However the disgust appeal condition (M = 5.98, SD = 1.26) did not differ significantly from the moral offense appeal condition (M = 5.88, SD = 1.45, p = 0.997). There are no significant findings to support that disgust appeals lead to more negative attitudes towards the advertisement then moral offense appeals, we therefore reject H4.

Brand attitude

A one-way ANOVA with a Tukey HSD post hoc test was conducted with shock as independent variable and brand attitude as dependent variable (see table 4). The ANOVA showed that there were no significant differences in brand attitude between the three

conditions: F = (2,259) = 1.66, p = 0.19. Tukey’s Post Hoc showed that the control condition (M = 4.05, SD = 1.43) did not differ significantly with the disgust appeal (M = 4.67, SD = 1.46), p = 0.989. Also the control condition does not differ significantly with the moral offense appeal (M = 5.01, SD = 1.28), p = 0.316. Lastly the disgust appeal condition did not differ significantly from the moral offense appeal condition. (M = 4.04, SD=1.92), p = 0.225. There is no statistical support that disgust appeals lead to more negative attitudes toward the brand Bullit than moral offense appeals, we therefore reject H5.

Purchase intentions

Purchase intentions. A one-way ANOVA with a Tukey HSD post hoc test was conducted with shock as independent variable and purchase intention as dependent variable (see table 4). The ANOVA showed that there were no significant differences in purchase intention between

(30)

29 the three conditions: F = (2,259) = 0.62, p= 0.54. Tukey’s Post Hoc showed that the control condition (M = 5.84, SD = 1.37), did not differ significantly with the disgust appeal condition

(M = 5.58, SD = 1.51, p = 0.535). The control condition (M = 5.84, SD = 1.37) also does not differ significantly with the moral offense appeal condition (M = 5.46, SD = 1.78, p = 0.686). Lastly the disgust appeal condition (M = 5.58, SD = 1.51) did not differ significantly from the moral offense appeal condition (M = 5.46, SD = 1.78, p = 0.962). There is no statistical support that disgust appeals lead to higher purchase intentions in comparison to moral offense appeals, we therefore reject H6a.

Willingness to pay. A one-way ANOVA with a Tukey HSD post hoc test was conducted with shock as independent variable and willingness to pay as dependent variable (see table 4). The ANOVA showed that there were no significant differences in purchase intention between the three conditions: F = (2,259) = 0.57, p= 0.58. Tukey’s Post Hoc showed that the control

condition (M = 0.69, SD = 0.61), did not differ significantly with the disgust appeal condition

(M = 0.648, SD = 1.46, p = 0.837). The control condition (M = 0.69, SD = 0.61) also does not differ significantly with the moral offense appeal condition (M = 0.61, SD = 0.54, p = 0.837). Lastly the disgust appeal condition (M = 0.648, SD = 1.46) did not differ significantly from the moral offense appeal condition (M = 0.61, SD = 0.54, p = 0.869). There is no

statistical support that disgust appeals lead to a higher willingness to pay in comparison to moral offense appeals, we therefore reject H6b.

CONCLUSION

Discussion

This study set out to find out to what extent there is a difference between the effects of different types of shock appeals on advertising effectiveness. Specifically the study looked at

(31)

30 the differences between disgust appeals, moral offense appeals on brand attitude, ad attitude and purchase intention when used in a commercial context. Firstly we expected that disgust appeals would lead to more negative attitudes towards the advertisement than moral offense appeals, findings did not confirm this assumption. The findings also did not confirm that disgust appeals lead to more negative attitudes toward the brand than moral offense appeals. Furthermore the fact that disgust appeals would lead to higher purchase intentions in

comparison to moral offense appeals also was not supported by the research findings.

Additionally the claim that disgust appeals would lead to a higher willingness to pay for a can of Bullit in comparison to moral offense appeals was not supported by the findings. It was also expected that disgust appeals would lead to less seconds of voluntary advertisement in comparison to moral offense appeals, again the results show no significant indication that this is the case. Therefore in this study did not find a significant difference between the type of shock used and its effect on advertising effect.

The second aspect that this research aimed to find out is to what extent the level of elaboration and suprise plays a mediating role in the relationship of shock appeals with brand attitude, ad attitude the willingness to pay, voluntary exposure and purchase intention. The findings indicate that there is no significant difference between moral offense appeals and disgust appeals when it comes to the level of elaboration and suprise they experience. Furthermore there are also no significant indicators that elaboration and suprise have a mediating role in the relationship of shock and its effect on advertising effectiveness.

Limitations

Firstly, the manipulation checks were not significant. The means indicated that people felt more disgusted at the disgust appeal than the when presented with the control or moral offense appeal. However the difference was not significant. Similar findings apply to the

(32)

31 moral offense appeal, people rated the moral offense appeal as being the most morally

offensive however the mean differences were not significant. This could indicate that the chosen moral offense ad and disgust appeal ad were not proper representations of these shock appeal. However as the ads were built in complete accordance with Dahl’s (2003)

operationalization it is also possible that there other explanations. Firstly the manipulation tests are measured at the end of the experiment, this could mean that if the emotions experienced were short-lived that the validity of the measure might be compromised as the participants could not accurately remember the emotions they experienced when they were exposed to the advertisement. Secondly literature did not specify that shock appeals lead to certain emotions and assessments rather it only focuses on ad conceptualizations, therefore using participant assessment as a manipulation check for shock appeals might have been incorrect.

Secondly there was a significant difference in how professional and credible the participant rated the different levels of shock. However looking at the results there are indications that there might have been problems with the validity of these control items. The participants rated the disgust appeal and moral offense appeal advertisements as less

professional and less credible than the control ad. However there is a strong possibility that they did not rate the quality of the ads but rather the content. Especially considering that the control advertisement was considered significantly more professional and credible while the only difference with the other ads was the background picture (see Appendix A). This

indicates that the measure of ad quality is probably a measure of ad attitude. Therefore it was decided that the differences in how professional and credible the conditions were rated would likely not affect the findings.

Thirdly, the lack of ecological validity of this online experiment is a limitation to the study. The experiment creates an artificial setting where the participants are forced into a high

(33)

32 involvement situation. While in reality advertisements are often not intensively studied

(Underwood, 1990). The different types of shock could have unconscious effects on the participants that were not able to measure this way.

Implications and future research

This study has been a first exploration regarding the differences between the types of shock. The literature showed strong indications that the workings and advertising effectiveness of shock appeals differ from each other. That this study did not find significant results to confirm mediation effects and differences between the types of shock does not mean that there are no differences regarding the effects and processes of the different shock appeals. This study focused on explicit short term effects, where people had to self-report their own measures. There are two logical steps for future research into shock appeals. First future research needs to focus on measuring the implicit and unconscious effects of the different shock appeals, the fact that people do not know that they are influenced by an advertisement does not mean that they are not (Dijksterhuis, Aarts & Smith, 2005). Second, future research should focus on long term effects of shock appeals; it could also be that people are affected trough shock advertising on the long run rather than the short term. Shock appeal research is necessary and important because in practice one always needs to think about how advertising appeals are processed and thus what effect are they likely to cause.

(34)

33

REFERENCES

Andersson, S., Hedelin, A., Nilsson, A., & Welander, C. (2004). Violent advertising in fashion marketing. Journal of Fashion Marketing and Management, 8, 96–112. Andrews, J. C., Durvasula, S., & Akhter, S. H. (1990). A framework for conceptualizing and

measuring the involvement construct in advertising research. Journal of Advertising,

19, 27-40.

Babbie, E. (2007). The Basics of Social Research. Belmont: Thomson Wadsworth.

Baker, J., Levy, M., & Grewal, D. (1992). An experimental approach to making retail store environmental decisions. Journal of Retailing, 68, 445-460.

Belayachi, N. (2013). The effect of different shock appeals on purchase intention: An online

experiment. Unpublished research proposal, University of Amsterdam.

Bello, D. C., Pitts, R. E., & Etzel, M. J. (1983). The communication effects of controversial sexual content in television programs and commercials. Journal of Advertising, 12, 32-42.

Brown. K. (2013). Suicide car ad removed after Twitter outrage. Retrieved May 6, 2014, from http://www.thetimes.co.uk/tto/news/medianews/article3748965.ece.

Cacioppo, J. T., & Petty, R. E. (1981). Social psychological procedures for cognitive response assessment: The thought-listing technique. Cognitive assessment, 2, 309-342.

Chaudhuri, A., & Buck, R. (1995). Affect, Reason, and Persuasion Advertising Strategies That Predict Affective and Analytic‐Cognitive Responses. Human Communication

(35)

34 Dahl, D., Frankenberger, D., & Manchanda, R. (2003). Does It Pay to Shock? Reactions to

Shocking and Nonshocking Advertising Content among University Students. Journal

of Advertising Research, 43, 268-280.

Dens, N., & De Pelsmacker, P. (2010). Consumer response to different advertising appeals for new products: The moderating influence of branding strategy and product category involvement. Journal of Brand Management, 18, 50-65.

Dick, A. S., & Basu, K. (1994). Customer loyalty: toward an integrated conceptual framework. Journal of the Academy of Marketing Science, 22, 99-113. Dijksterhuis, A., Aarts, H., & Smith, P. K. (2005). The power of the subliminal: On

subliminal persuasion and other potential applications. The new unconscious, 15, 77 106.

Dillard, J. P. (1994). Rethinkin the Study of Fear Appeals: An Emotional Perspective.

Communication Theory, 4, 295-323.

Donovan, R. J., & Henley, N. (1997). Negative outcomes, threats and threat appeals:

Widening the conceptual framework for the study of fear and other emotions in social marketing communications. Social Marketing Quarterly, 4, 56-67.

Dunning, T. (1993). Accurate methods for the statistics of suprise and coincidence.

Computational linguistics, 19, 61-74.

Eckler, P., & Bolls, P. (2011). Spreading the virus: emotional tone of viral advertising and its effect on forwarding intentions and attitudes. Journal of Interactive Advertising, 11, 1-11.

Follows, S. B., & Jobber, D. (2000). Environmentally responsible purchase behaviour: a test of a consumer model. European Journal of Marketing, 34, 723-746.

(36)

35 Gardner, M. P. (1985). Does attitude toward the ad affect brand attitude under a brand

evaluation set? Journal of Marketing Research, 5, 192-198.

Hayes, A. F., & Preacher, K. J. (2013). Statistical mediation analysis with a multicategorical independent variable. British Journal of Mathematical and Statistical Psychology. Huhmann, B. A., & Mott‐Stenerson, B. (2008). Controversial advertisement executions and

involvement on elaborative processing and comprehension. Journal of Marketing

Communications, 14, 293-313.

Krstic, T. (2007). Attitudes toward shock advertising of Western-European and Serbian

university students with regard to public health context (anti-smoking and anti HIV/AIDS campaigns). University of Nottingham.

Krugman, D. M., Fox, R. J., Fletcher, J. E., Fischer, P. M., & Rojas, T. H. (1994). Do adolescents attend to warnings in cigarette advertising? An eye-tracking approach.

Journal of Advertising Research, 34, 39-39.

Lavidge, R. J., & Steiner, G. A. (1961). A model for predictive measurements of advertising effectiveness. The Journal of Marketing, 2, 59-62.

Leshner, G., Bolls, P., & Thomas, E. (2009). Scare'em or disgust'em: The effects of graphic health promotion messages. Health communication, 24, 447-458.

Lindgreen, A., & Vanhamme, J. (2003). To suprise or not to suprise your customers: the use of suprise as a marketing tool. Journal of Customer Behaviour, 2, 219-242.

Loef, J., & Verlegh, P. W. (2002). Cognitive and Affective Consequences of Two Types of

(37)

36 Machleit, K. A., & Wilson, R. D. (1988). Emotional feelings and attitude toward the

advertisement: The roles of brand familarity. Journal of advertising, 17, 27-35. Manchanda, R., Dahl, D., & Frankenberger, K. (2002). Shocking ads: do they work. Advances

in consumer research, 29, 230-231.

Mehta, A. (2000). Advertising attitudes and advertising effectiveness. Journal of Advertising

Research, 40, 67-71.

Nabi, R. L. (1998). The effect of disgust‐eliciting visuals on attitudes toward animal experimentation. Communication Quarterly, 46, 472-484.

Ning, C. (2001). The effect of advertisement exposure and brand maturity on the mode of information processing. Journal of Chinese Psychology Acta Psychologica Sinica, 5, 10-20

Noel, C. P. (2010). Shock advertising: Theories, risks, and outcomes analyzed using the case of Barnardo's. Student Pulse, 2, 10-15.

Parry, S., Jones, R., Stern, P., & Robinson, M. (2013). ‘Shockvertising’: An exploratory investigation into attitudinal variations and emotional reactions to shock advertising.

Journal of Consumer Behavior, 12, 112-121.

Pollay, R. W. (1986). The distorted mirror: Reflections on the unintended consequences of advertising. Journal of Marketing, 50, 18-36.

Sandicki, O. (2011). Shock tactics in advertising and implications for citizen-consumers.

International Journal of Humanities and Social Science, 1, 42–50.

Scherer, K. R., & Ekman, P. (1984). On the nature and function of emotion: A component process approach. Approaches to emotion, 2293, 317-318.

(38)

37 Spears, N., & Singh, S. N. (2004). Measuring attitude toward the brand and purchase

intentions. Journal of Current Issues & Research in Advertising, 26, 53-66.

Underwood, A. (1990). Experiments in ecology and management: their logics, functions and interpretations. Australian journal of ecology, 15, 365-389.

MacInnis, D. J., Moorman, C., & Jaworski, B. J. (1991). Enhancing and measuring

consumers' motivation, opportunity, and ability to process brand information from ads.

The Journal of Marketing, 5, 32-53.

Wang, H. (2009). System And Method For Obtaining Advertising Exposure Time. Google: California.

West, D. C., & Sargeant, A. (2004). Taking risks with advertising: The case of the not-for-profit sector. Journal of marketing management, 20, 1027-1045.

(39)

38

APPENDICES

Appendix A: Stimuli Material

(40)

39 Moral offense appeal

(41)

40 Appendix B: Experiment questions*

Voluntary exposure**

Please look at the following advertisement as if you were flipping through a Magazine.

* Several ordinals Likert-scale measures were reversed afterward in order to obtain logical coded results in SPSS ** Participants were displayed one of the three shown conditions; the timing element was not visible for the participants.

(42)

41 Elaboration

(43)

42 Brand attitude

Ad attitude & Ad quality

(44)

43 Manipulation

Willingness to pay

(45)

44 Demographics