How do celebrity endorsers influence the
brand image of the promoted brand?
Master thesis, MscBA, specialization Marketing Research
University of Groningen, Faculty of Economics and Business
September 27, 2012
Author: Henk Lambers
Student number: 1772473
Schaepmanlaan 83
9722 NS Groningen
tel: +31 (0)643891631
e-mail:
h.lambers@student.rug.nl
Supervisor: B. Fennis
How do celebrity endorsers influence the brand image
of the promoted brand?
Abstract:
Table of content
1. Introduction
4
1.1 General introduction
4
1.2 Structure of the thesis
5
2. Literature review
5
2.1 The influence of the celebrity status of endorsers on brand image
5
2.2 The moderating influence of product-endorser fit
6
2.3 Exploring the mechanism which causes the moderation effect
7
2.4 The mediated moderation type of relationship
9
3. Research methods
10
3.1 Pretest description
10
3.2 Participants and design of the main experiment
11
3.3 Procedure
11
3.4 Manipulation of the independent variables
11
3.5 Measurement of the dependent variables
12
4. Results and discussion
14
4.1 Test of hypotheses
14
5. Discussion and implications
17
5.1 Discussion and implications
17
5.2 Limitations
19
5.3 Directions for future research
19
5.4 Concluding remarks
20
References 21
Appendices 24
Appendix A: Copy of the pretest 24
Appendix B: Copy of the main experiment 25
Appendix C: The advertisements used in the main experiment 27
1. Introduction
1.1 General introduction
You only have to turn on the TV or get the mail from the mailbox and you cannot escape; a lot of advertisements are trying to affect your behavior in order to give to charity, to buy a certain product or to become more environmentally responsible. In a large part of these advertisements, celebrities are being used to promote brands, organizations, services and, social desirable conduct. The
celebrities are expected to make the advertisement believable and enhance brand recall (Agrawal & Kamakura, 1995). Also, the celebrities create a distinct personality of the brand (McCracken, 1989). However, the endorsements deals represent a large marketing investment and come with a number of risks. Therefore, it is important to consider possible benefits and costs of using a celebrity
endorser to promote a brand.
This study investigates the influence of celebrity endorsers on the resulting brand image of the promoted brand. The main factors of interest are the perceived persuasiveness of the endorser and, the fit between characteristics of both the endorser and the brand. The first factor is selected since it is expected that this factor mediates the relation between the celebrity status of endorsers and brand image. The second factor is expected to moderate the relationship between celebrity status of endorsers and the persuasiveness of endorsers. Combining these relations, a moderated mediation mechanism arises. This study is the first to investigate if the moderation influence of product-endorser fit is caused by the mediation mechanism.
1.2 Structure of the thesis
2. Literature review
First, it is important to define a celebrity endorser. Here, the definition of McCracken (1989) is followed: “any individual who enjoys public recognition and who uses this recognition on behalf of a consumer good by appearing with it in an advertisement”. This implies that not only celebrities are included in “celebrity endorsers”, but also other types of well-known people such as sport athletes or politicians.
2.1 The influence of the celebrity status of endorsers on brand image
It is a common practice for companies to use celebrities as endorser in order to support corporate and brand images (Erdogan, 1999). A celebrity endorser transfers meanings and associations of the endorser to the brand which is promoted (Halonen-Knight & Hurmerinta, 2010). According to McCracken (1989), the transfer of these meanings and associations is a three-stage process. In the first stage, the celebrity acquires the meanings and associations through publicity and their profession (McCracken, 1989). In the second stage, the meanings and associations of the celebrity are coupled to the promoted brand, by using the celebrity as an endorser for the product in an advertisement (McCracken, 1989). In accordance, Miller & Allen (2012) found that beliefs about the brand became more aligned to the existing beliefs related to the celebrity after the usage of the celebrity as an endorser for the promoted brand. Here, in cases when the central route of processing applies, where the target of the advertisement possesses a high degree of processing capacity, the celebrity can act as a strong argument in favor of the brand, and thereby positively change the attitude of a target towards the promoted brand (Stroebe, 2008). For example, when Usain Bolt uses a certain brand of sneakers during his races, the sneakers must be good and enable you to run fast. During the final step, the target of the advertisement adopts the meanings and associations by building a relationship with the promoted brand (Miller & Allen, 2012). In cases when the peripheral route of processing applies, where the target of the advertisement possesses a low degree of processing capacity, the meanings and associations function as heuristic cues (Petty & Cacioppo, 1986). Positive heuristic cues associated to the advertisement lead to a positive change in the attitude of the target towards the brand (Petty & Cacioppo, 1986).
brand. Therefore, in this study it is hypothesized that the celebrity status of the endorser positively influences the resulting brand image of the brand which is promoted in the advertisement.
H1: The celebrity status of the endorser positively influences the brand image of the promoted brand.
2.2 The moderating influence of product-endorser fit
According to the match-up hypothesis, celebrities might not be effective for all brands in all
situations (Kahle & Homer, 1985; Koernig & Boyd, 2009). The match-up hypothesis has indicated that the fit between the characteristics of the celebrity and the characteristics of the promoted brand influences the relationship between the celebrity status of the endorser and the resulting brand image of the promoted brand (e.g. Kahle & Homer, 1985; Kamins & Gupta, 1994). The fit between the endorser and the promoted brand is defined as the congruence between the characteristics of the endorser and the attributes of the promoted brand (Seno & Lukas, 2007). For example, the fit between attractive endorsers and beauty enhancing products is present since the characteristics of the attractive endorsers (beautiful) are congruent with the attributes of the promoted brand (beauty enhancing).
Past research evidence shows that for situations where the fit between the celebrity endorser and the promoted product is present, more favorable attitudes towards the brand result (e.g. Kahle & Homer, 1985; Kamins, 1990; Kamins & Gupta, 1994; Seno & Lukas, 2007). Siemens, Smith, Fisher & Jensen (2008) suggest that the more favorable attitudes towards the brand result from a higher degree of perceived credibility of the endorser. The endorser is perceived to be more credible since the endorser is attributed with a higher degree of knowledge concerning the promoted brand (Siemens et al., 2008). These findings in literature suggest that in cases where the fit between the celebrity endorser and the promoted brand is present, the positive effect of the celebrity status of the endorser on the resulting brand image increases. Therefore, in this study it is hypothesized that the congruence between the characteristics of the endorser and the attributes of the promoted brand moderate the relationship between the celebrity status of the endorser and the resulting brand image to the extent that when the fit is present, the positive effect of the celebrity status on the resulting brand image increases.
2.3 Exploring the mechanism which causes the moderation effect
In order to establish the cause of the moderating effect of the fit between the celebrity endorser and the promoted brand, this study attempts to find out via which mediating mechanism the moderating effect of fit materializes. The segments of this mediating mechanism are discussed below.
The first segment of the mechanism via which the moderating effect of fit materializes concerns the relationship between the celebrity status of the endorser and the persuasiveness of the endorser. Persuasion is a fundamental form of social influence on human decision making (Klucharev, Smidts & Fernandez, 2008). When a respondent is more familiar with a source of a message, the greater the agreement of the respondent is to the message put forward by the source (Roskos-Ewoldsen & Fazio, 1992). Thus, when a respondent is familiar with a source, the persuasive influence of the
spokesperson increases. This suggests that a high level of familiarity of the source causes the message put forward by the source to be more persuasive. This finding is confirmed by Weisbuch, Mackie & Garcia-Marques (2003). Celebrity endorsers are expected to have high levels of public recognition and thereby having a high degree of familiarity to the people which are processing the advertisement. This high degree of familiarity implies that celebrity endorsers have more persuasive influence than non-celebrity endorsers, which do not have a high degree of familiarity to people. Therefore, in this study it is hypothesized that celebrity endorsers have more persuasive influence compared to non-celebrity endorsers.
H3: Celebrity endorsers are perceived to be more persuasive than non-celebrity endorsers.
The relationship between the celebrity status of the endorser and the persuasiveness of the endorser is expected to be moderated. According to Bower & Landreth (2001), the perceived expertise and credibility of the celebrity endorser with respect to the promoted product increases when the product-endorser fit is present, since the endorser is attributed with a higher degree of knowledge concerning the promoted product. Petty & Wegener (1998) stated that high expertise sources are generally found to be more persuasive in comparison to low expertise sources. Since a good product-endorser fit increases the expertise of the product-endorser, this leads to a higher persuasiveness of the endorser. This implies that when the product-endorser fit is present, the positive effect of the celebrity status on the perceived persuasiveness is larger compared to situations in which the product-celebrity fit is not present. Therefore, it is hypothesized that if the product-endorser fit is present, the positive effect of celebrity status on the persuasiveness of the endorser and the advertisement increases.
The final segment of the mediating mechanism concerns the influence of the persuasiveness of the endorser on the resulting brand image. The perceived persuasiveness of the endorser can influence the brand image of the promoted brand via either the peripheral route or the central route of processing (Stroebe, 2008). When an advertisement is processed via the peripheral route, attitude changes result from associations made between the content of the advertisement and positive or negative cues based on various simple heuristic cues (Petty & Cacioppo, 1984). Any message variable can serve as these positive or negative heuristic cues (Brinol & Petty, 2006). This suggests that the persuasiveness of the endorser can also serve as a heuristic cue, where it is reasonable to expect that a high degree of persuasiveness leads to positive heuristic cues and vice versa. Positive heuristic cues lead to a positive change in the attitude of a target (Petty & Cacioppo, 1986), implying that a high persuasiveness of the endorser leads to a positive change in the brand image of the promoted brand. In contrasting cases where an advertisement is processed via the central route, attitude changes result from the effort of the targets to analyze the benefits and costs of the proposition which is promoted in the advertisement (Petty & Cacioppo, 1984). The quality of the arguments determines the direction of the generated thoughts and thereby the direction of the attitude changes, where strong argument create favorable thoughts and positive attitude change while weak argument create unfavorable thoughts which result in no attitude change (Stroebe, 2008). When an endorser is perceived to be more persuasive, the arguments are more likely to be perceived as being strong arguments, thereby generating positive thoughts and thus, positive attitude change towards the brand. Since via both routes of processing the perceived persuasiveness of the endorser positively influences the brand image of the promoted brand, it is hypothesized that the perceived
persuasiveness of the endorser is positively related to the resulting brand image of the promoted brand.
H5: The perceived persuasiveness of the endorser is positively related to brand image of the promoted brand.
The hypotheses above draft the mediated mechanism which is hypothesized to cause the moderating effect of fit in the relationship between the celebrity status of the endorser and the brand image. The moderating effect is hypothesized to be caused by a higher perceived
persuasiveness of the endorser, due to the present endorser fit. In cases where the product-endorser fit is present, the magnitude of the positive influence of celebrity status on the
persuasiveness of the endorser increases. This increase in the magnitude of the effect is hypothesized to lead to a better brand image, since the persuasiveness of the endorser is
mechanism. In this manner, this study can contribute to field of literature concerning the match-up hypothesis.
2.4 The mediated moderation type of relationship
The identified hypotheses indicate that a mediated moderation type of relationship is present. Based on the mediated moderation framework of Muller, Judd & Yzerbyt (2005) and the reasoning above, the following relationship is hypothesized in this study. The celebrity status of the endorser is found to positively influence the brand image of the promoted brand. This relationship is moderated by the fit between the endorser and the brand, where it was found that in cases where the fit is present, the positive effect of the celebrity status of endorser increases, while in cases where the fit is not
present, the positive effect of celebrity status decreases. The moderating effect of fit is hypothesized to be caused by the mediating mechanism where the celebrity status of the endorser positively influences the persuasiveness of the advertisement, which leads to a better brand image. The effect of celebrity status on the persuasiveness of the advertisement is moderated by the fit between the endorser and the brand, where it is hypothesized that the positive effect of celebrity status increases in magnitude when the fit is present.
3. Research methods
3.1 Pretest description
In order to manipulate the celebrity status of the endorser and the product-endorser fit, a pretest was performed. This pretest was based on a pretest performed by Kamins & Gupta (1994), which resulted in significant differences for the variables concerning the celebrity status of the endorser and concerning the product-endorser fit. A group of 43 participants were presented with three endorsers which were expected to have a celebrity status, being Maarten Stekelenburg, Robert Gesink and Inge de Bruijn, and of two endorsers which were expected to have no celebrity status, being a laptop salesman and a car salesman. The participants were asked to answer a set of three questions (very unfamiliar/very familiar, bad/good reputation, not famous/very famous) on a 7 point Likert scale to indicate to what extent the endorser was a celebrity in their opinion. The mean of the scoring of each participant was calculated and used for further analysis. The three above mentioned expected celebrity endorsers scored significantly higher than the two mentioned endorsers which were not expected to be a celebrity (t = 17.12; df = 35; p = <0.01). Based on the mean scores, a T-test was performed where it was found that Inge de Bruijn had a significantly higher perceived celebrity status compared to Maarten Stekelenburg and Robert Gesink (t = 6.98; df = 42; p = <0.01). Therefore, Inge de Bruijn was selected as being the celebrity endorser. Both the car salesman and the laptop salesman were found to have the lowest perceived celebrity status. However, these two endorsers did not significantly differ from each other (t = 1.667; df = 39; p = 0.103). Therefore, both endorsers were selected for the next phase of the pretest. The selected celebrity endorser, Inge de Bruijn, scored significantly higher than both the two endorsers with the lowest perceived celebrity status (t = 22.13; df = 37; p < 0.01).
lower on sports clothing compared to the car salesman (t = 2.421; df = 41; p = 0.02). Therefore the laptop salesman was selected as being the non-celebrity endorser, because of the largest non-fit. Since the laptop salesman was found to have a significant good fit with computers, computers were selected as the second product category.
3.2 Participants and design of the main experiment
125 people participated in the main experiment. Only Dutch people participated in the study. The range of the ages of the participants varied from 17 to 64, with an average of 31.5 years. The group of participants consisted of 74 males and 51 females. The participants were randomly assigned to 2 (celebrity endorser or non-celebrity endorser) x 2 (product-endorser fit; present vs. not present) between-subjects factorial design. The 2x2 between-subjects factorial design was converted to 4 different advertisements, where both factors were randomly coupled. The participants were
randomly assigned to one of these 4 advertisements. The 4 advertisements are depicted in Appendix C.
3.3 Procedure
The participants were approached via an e-mail. The experiment took place in an online environment. The participants were asked to read the instructions carefully and begin the
experiment. First, questions concerning age, gender and highest level of education were posed to the participants. Then, the participants were told to take a thorough look at the advertisement with which they were presented. Next, the participants had to answer a number of questions concerning their attitude towards the advertisement and their towards the promoted brand. These questions included a manipulation check and questions concerning the attitudes towards the brand and finally. The answers to these questions were collected and saved and form the data on which the analyses were performed.
3.4 Manipulation of the independent variables
In order to manipulate the celebrity status of the endorser, either the celebrity endorser or the non-celebrity endorser was depicted in the advertisement. At the end of the experiment, the participants had to indicate on three 9-point Likert scales (very unfamiliar/very familiar, bad reputation/good reputation, not famous/very famous), to what extent they perceived the endorser as being a celebrity. In this manner, the success of the manipulation was checked. Since the responses were internally consistent (Cronbach’s Alpha = 0.94), the responses were summed together. The
The product-endorser fit was also manipulated. The two product types were randomly coupled with either the celebrity or the non-celebrity endorser. In case of the pairing of Inge de Bruijn and sports clothing and in the case of the pairing of the laptop salesman and laptops, it was expected that participants perceived these combinations as a situations in which the product-endorser fit is present. In contrast, in cases where Inge de Bruijn was paired with laptops and cases where the laptop salesman was paired to sports clothing, no perceived product-endorser fit was expected. The success of this manipulation was also checked at the end of the experiment, by posing two 9-point Likert scales to the participants where the participants had to indicate to what extent they perceive that a product-endorser fit was present in the advertisement and, to what extent the endorser has knowledge concerning the product. Since the responses were internally consistent (Cronbach’s Alpha = 0.90), the responses were summed together. The manipulation was successful since the cases were a fit was expected scored significantly higher than cases where no fit was expected (t = 8.64; df = 123; p = <0.01).
Unfortunately, the manipulation of the independent variables suffered from a number confounds. For example, the position of the picture of the endorser differs in the advertisements. These
confounds have consequences for the interpretability of the results in this study. The implications of these confounds are discussed in the segment where the limitations of this study are discussed.
3.5 Measurement of the dependent variables
Persuasiveness of endorsers
In order to measure the perceived persuasiveness of the endorsers depicted in the processed advertisement, a measure was developed based on the studies of Naylor, Lamberton & Norton (2011) and Karmarkar & Tormala (2009). After processing the advertisement, participants were confronted with four 9 point Likert scaled questions. The participants had to indicate if they would consider buying the product, to what extent the message was convincing, to what extent the message was trustworthy and, whether they had favorable or unfavorable affection with the advertisement. The responses of the participants were highly consistent on these four items (Cronbach’s alpha = 0.87) and therefore, the scores were averaged and used as a single, composite item. A higher value on this composite item indicates a higher level of persuasiveness of the relevant endorser.
Brand image
4. Results
4.1 Test of hypotheses
A 2 (celebrity status of endorser; present vs. not present) x 2 (endorser-product fit; present vs. not present) ANOVA, using the brand image measure as the dependent variable was conducted.
Consistent with the hypothesis 1, a significant main effect of celebrity status was present (F(1,123) = 19.311, p = <0.001). Participants evaluated the brand significantly better when a celebrity endorser (M = 17.00) was used to promote the product compared to cases where a non-celebrity endorser (M = 13.67) was used. The results also showed a significant main effect of product-endorser fit (F(1,123) = 4.63, p = 0.03). The main effect of product-endorser fit showed that the resulting brand image is better when the product-endorser fit is present (M = 16.15, SD = 0.53) than in cases when the product-endorser fit is not present (M = 14.52, SD = 0.55). The main effect of celebrity status on brand image is also found to be moderated by the product-endorser fit (F(1,123) = 4.63, p = 0.03). The positive effect of celebrity status was larger when the product-endorser fit was present (M = 18.63 vs. M = 13.67), compared to situations where the product-endorser fit was not present (M = 15.37 vs. M = 13.67). This confirms hypothesis 2.
A 2 (celebrity status of endorser; present vs. not present) x 2 (endorser-product fit; present vs. not present) ANOVA, using the persuasiveness measure as the dependent variable was conducted. Consistent with hypothesis 3, the ANOVA showed a significant main effect of celebrity status (F(1,123) = 9.50, p = 0.03). Participants rated the persuasiveness of the celebrity endorser
Table 1: Regression results for brand image
Standard errors are reported in parentheses * indicates significance at the 99% level
In order to analyze whether the mediated moderation type of relationship is present, the standard procedure developed by Muller et al. (2005) is executed. Based on three estimated regression models, the type of relationship present in the data can be identified. The resulting brand image was the dependent variable for the first and third regression model, while the persuasiveness of the endorser was the dependent variable for the second regression model. The results of the regression models are depicted in Table 2.
Table 2, resulting parameters of the three regression models based on Muller et. al (2005)
In order for the hypothesized mediated moderation relationship to be present, the main effect of celebrity status and the moderation effect of celebrity status must significantly differ from zero in the first regression model. In addition, the main effect of celebrity status and the moderation effect of celebrity status and product-endorser fit must be significantly different from zero in the second regression model. Finally, the main effect of the persuasiveness of the advertisement must
significantly differ from zero in the third regression model. As can be seen in Table 2, these factors all are significantly different from zero, which implies that the hypothesized mediated moderation relationship is present and is confirmed. In addition, it was found that the moderating effect of celebrity status and product-endorser fit does not significantly differ from zero in the third regression. This implies that the mediating mechanism is fully causing the moderating effect of product-endorser fit in the relationship between celebrity status and brand image.
In conclusion of the results section, it was found that all the hypotheses were confirmed. In addition the regression models based on Muller et al. (2005) also confirm the hypotheses. Finally, the
hypothesized mediated moderation relationship is found to be present. This implies that the
5. Discussion and conclusions
5.1 Discussion and implications
The results of this study are found to confirm the hypothesis that the celebrity status of the endorser is positively related to the resulting brand image of the promoted brand. The results indicate that in general, the usage of celebrity endorsers is more effective than the usage of non-celebrity endorsers to promote a brand, since a better brand image results. This finding is in line with literature (e.g. McCracken, 1989) and implies that the positive meanings and associations related to the celebrity endorser transfer to the promoted brand.
The results also confirm the hypothesis with respect to the moderating influence of the fit between the celebrity endorser and the promoted brand. Here, it was expected that in cases where the fit was present, the magnitude of the positive effect of the celebrity status of the endorser increases. This finding confirms literature with respect to the match-up hypothesis (e.g. Kamins & Gupta, 1994; Lynch & Schuler, 1994). The practical implication of this finding is the notion that marketers must carefully select which celebrity endorser to use in order to promote their brand. In order to attain the maximum efficiency, it is important for marketers to assure that the characteristics of the celebrity are congruent with the attributes of the promoted brand.
The results also show that the celebrity status of the endorser is positively related to the persuasiveness of the endorser. This finding confirms the results of Weisbuch et al. (2003) and Roskos-Ewoldsen et al. (1992). This indicates that celebrity endorsers have a higher degree of familiarity to the people processing the advertisement which causes the persuasiveness of celebrity endorsers to be higher compared to non-celebrity endorsers. This implies that a celebrity endorser is better able to communicate the message put forward in the advertisement. From a practical
standpoint, that the usage of a celebrity endorser enables marketers to be able to better communicate the attributes of the brand.
endorsers are better capable of convincing the targets of advertisements when they have similar characteristics as the promoted brand. This implies that marketers must assure that the congruence between the characteristics of the celebrity endorser and the brand is present. In that way, celebrity endorsers are better capable of convincing the targets of the advertisement that the message of the advertisement is well-founded.
The results show that when endorsers are found to be more persuasive, a better brand image results. Hence, more persuasive endorsers are better capable of delivering the message of the advertisement to the targets of the advertisement. This finding was also hypothesized in this study. Since the messages are intended to positively alter the attitude towards the brand, more persuasive endorsers are more capable to improve the image of the promoted brand, which leads to a better resulting brand image. This suggests that it is important for marketers to make sure that the endorser is perceived to be very persuasive, since a more persuasive endorser is found to lead to a better resulting brand image.
The most important finding of this study concerns the mediated moderation relationship. This study is the first to identify a mediated moderation relationship between the celebrity status of the endorser and the resulting brand image. Based on the results and the framework of Muller et al. (2005), it was identified that the main effect of celebrity status of the endorser on the resulting brand image is moderated by the fit between the endorser and the promoted brand. The magnitude of the positive effect of celebrity status depends on this fit whereby the magnitude of the positive effect is larger when this fit is present. This moderating effect is fully caused by the intervening mechanism where celebrity status, again moderated by the fit between the endorser and the
5.2 Limitations of the study
This study has a number of limitations. The most severe limitation concerns the confounds in the manipulation of the independent variables. The developed advertisements, which are depicted in Appendix C, suffer from a number of confounds which reduces the quality of the research. For example, the pictures of the endorsers are positioned differently in the advertisements. The confounds cause inaudibility’s, which results in impurities in the measurement of the effects of the manipulations. These impurities decrease the contribution of this study since the confidence of the results is declined due to the existence of the confounds. However, the manipulation checks indicate that there was a significant difference between the endorsers with respect to their celebrity status and their fit with the respective products. These manipulation checks give some support to the results of this study. However, despite the support from the manipulation checks, the confounds in the manipulation of the independent variables are a major limitation of this study.
An additional limitation concerns the number of product categories included in this study. Only two product categories are included in this study. In order to be able to generalize the findings of this study across product categories, more research is necessary in which other product categories are examined. Adding to this, the impact of the endorsement of multiple brand by the same celebrity is not included in this study. Tripp, Jensen & Carlson (1994) found that the number of products
promoted by the same endorser, negatively influences the credibility and likability of the endorser. It is therefore reasonable to expect the number of promoted products by the same endorser influences the identified relationship in this study.
5.3 Directions for future research
Much of the past research with respect to the product-endorser fit has focused on the physical characteristics of the endorser (e.g. Kahle et al., 1985; Kamins, 1990). In contrast, Till & Busler (2000) found that the more favorable attitudes towards the brand did not result from an attractiveness-related fit with the promoted brand, but found that the attractiveness of the endorser caused the more favorable attitudes. This suggests that a gap in the literature is present and future research on the product-endorser fit is required, where the influence of the product-endorser fit is examined for a product category which is not related to the physical characteristics of the endorser.
researcher to determine whether this effect influences the found mediated moderating type of relationship.
5.4 Concluding remarks
This study is the first to find a mediated moderation relationship. The main contribution of this is that it was found that the moderating effect of the product-endorser fit in the relationship between the celebrity status of the endorser and the brand image of the promoted brand is fully caused by the fact that the celebrity endorser is perceived to be more persuasive compared to celebrity endorser which do not fit with the product. Unfortunately, the confounds in the manipulation of the
References
Agrawal, J., & Kamakura, W.A. (1995). The economic worth of celebrity endorsers: An event study analysis. Journal of Marketing, 59: 56-62
Bower, A.B., & Landreth, S. (2001). Is beauty best? Highly versus normally attractive models in advertising. Journal of Advertising, 30: 1-12
Brinol, P., & Petty, R.E. (2006). Fundamental processes leading to attitude change: Implications for cancer prevention communications. Journal of Communications, 56: 81-104
Erdogan, B.Z. (1999). Celebrity endorsement: a literature review. Journal of Marketing Management, 15: 291-314
Halonen-Knight, E., & Hurmerinta, L. (2010). .Who endorses whom? Meanings transfer in celebrity endorsement. Journal of Product & Brand Management, 19: 452-260
Jung, J.M., Min, K.S., & Kellaris, J.J. (2011). The games people play: How the entertainment value of online ads helps or harms persuasion. Psychology & Marketing, 28: 661-681
Kahle, L.R. & Homer, P.M. (1985). Physical attractiveness of the celebrity endorser: A social adaptation perspective. Journal of Consumer Research, 11: 954-961.
Kamins, M.A. (1990). An investigation into the “match-up” hypothesis in celebrity advertising: when beauty may only be skin deep. Journal of Advertising, 19: 4-13
Kamins, M.A., & Gupta, K. (1994). Congruence between spokesperson and product type: A matchup hypothesis perspective. Psychology & Marketing, 11: 569 – 586
Karmarkar, U.R., & Tormala, Z.L. (2009). Believe me, I have no idea what I’m talking about: the effect of source certainty on consumer involvement and persuasion. Journal of Consumer Research, 36: 1033 – 1049
Klucharev, V., Smidts, A., & Fernandez, G. (2008). Brain mechanisms of persuasion: how expert power modulates memory and attitudes. Oxford: University Press
MacKenzie, S.B., Lutz, R.J., & Belch, G.E. (1986). The role of attitude towards the ad as a mediator advertising effectiveness: A test of competing explanations. Journal of Marketing Research, 23: 130-143
McCracken, G. (1989). Who is the celebrity endorser? Cultural foundations of the endorsement process. Journal of Consumer Research, 26: 310-321.
Miller, F.M. & Allen, C.T. (2012). How does celebrity meaning transfer? Investigating the process of meaning transfer with celebrity affiliates and mature brands. Journal of Consumer Psychology , 22: 443-452.
Muller, D., Judd, C.M., & Yzerbyt, V.Y. (2005). When moderation is mediated and mediation is moderated. Journal of Personality and Social Psychology, 89: 852-863
Naylor, R. W., Lamberton C. P., & Norton D.A. (2011). Seeing ourselves in others: Reviewer ambiguity, egocentric anchoring, and persuasion. Journal of Marketing Research, 18: 617-631
Petty, R.E., & Cacioppo, J.T. (1984). The elaboration likelihood model of persuasion. Advances in Consumer Research, 11: 673-675
Petty, R.E., & Cacioppo, J.T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer
Petty, R.E., & Wegener, D.T. (1998). Attitude change: multiple roles for persuasion variables. in Gilbert, D.T., Fiske, S.T., & Lindzey G. (Eds.), Handbook of Social Psychology, 4th edition (Vol. 1, pp. 323-390), Boston, MA: McGraw-Hill
Roskos-Ewoldsen, D.R., & Fazio, R.H. (1992). The accessibility of source likeability as a determinant of persuasion. Personality and Social Psychology Bulletin, 18: 19-25
Rossiter, J.R. & Smidts, A. (2012). Print advertising: Celebrity presenters. Journal of Business Research, 65: 874-879
Siemens, J.C.; Smith, S.; Fisher, D. & Jensen, T.D. (2008). Product expertise versus professional expertise: Congruency between an endorser’s chosen profession and the endorsed product. Journal of Targeting, Measurement and analysis for Marketing, 16: 159-168
Stroebe, W. (2008). “Strategies of attitude and behavior change”, in M. Hewstone, W. Stroebe & K. Jonas (Eds.), Introduction to social psychology: A European perspective. Oxford, UK: Blackwell
Till, B.D., & Busler, M. (2000). The Match-up hypothesis: Physical attractiveness expertise, and the role of fit on brand attitude, purchase intent and brand beliefs. Journal of Advertising, 39
Tripp, C.; Jensen, T.D. & Carlson, L. (1994). The effects of multiple product endorsements by celebrities on consumers’ attitudes and intentions. Journal of Consumer Research, 20: 535-547.
Appendices
Appendix A: Copy of the pretest
Construct Item Scale
Demographics What is your gender? Nominal, male/female
What is your age Ratio
What is your highest level of education? Nominal, education levels
Picture of an endorser *
Here, an endorser is shown to the
participant
Endorser recall
Did you recognize the person on the
picture? Nominal, yes/no
If so, what is the name of the person?
Endorser's celebrity status
To what extend is this person familiar to
you? Likert, 1-7, very unfamiliar / very familiar
This person has a good reputation
Likert, 1-7, strongly disagree / strongly agree
To what extend do you think this person is
a celebrity? Likert, 1-7, not famous at all / very famous
Product-endorser fit
Football shoes
To what extend do you think this person
fit with the following product? Likert, 1-7, not at all / very good Computers
To what extend do you think this person
fit with the following product? Likert, 1-7, not at all / very good
Sports clothing
To what extend do you think this person
fit with the following product? Likert, 1-7, not at all / very good Car
To what extend do you think this person
fit with the following product? Likert, 1-7, not at all / very good Furniture
To what extend do you think this person
fit with the following product? Likert, 1-7, not at all / very good
Appendix B: Copy of the main experiment
Construct Item Scale
Demographics What is your gender? Nominal, male/female
What is your age Ratio
What is your highest level of education? Nominal, education levels
Advertisement * Here, the advertisement is shown
Brand recall
Do you know the name of the brand
depicted in the advertisement? Nominal, yes/no
If so, what is the name of the brand?
Brand image
Indicate on the following scales, what your opinion is concerning the brand
Likert, 1-7, unpleasant / pleasant
Likert, 1-7, unattractive / attractive
Likert, 1-7, boring / fascinating
Likert, 1-7, positive / negative
Likert, 1-7, unreliable / reliable
Likert, 1-7, unfriendly / friendly
Likert, 1-7, bad / good
Persuasiveness
Indicate to what extend you agree to the
following statements
The advertisement was convincing Likert, 1-9, disagree / agree
The advertisement is reliable Likert, 1-9, disagree / agree
I got a positive feeling associated with the
advertisement Likert, 1-9, disagree / agree
If i would need this type of product, I
would consider this brand Likert, 1-9, disagree / agree The advertisement had good arguments Likert, 1-9, disagree / agree
Involvement
Indicate to what extend you agree to the
following statements
I felt involved with the product Likert, 1-9, disagree / agree I am interested in this type of product Likert, 1-9, disagree / agree
I took a thorough look at the
advertisement Likert, 1-9, disagree / agree
Manipulation check
Do you know the person in the
advertisment? Nominal, yes / no
If so, what is the name of this person?
Indicate to what extend you agree to the
following statements
This person has a good reputation Likert, 1-9, disagree / agree
This person is a celebrity Likert, 1-9, disagree / agree
This person fits well to the product Likert, 1-9, disagree / agree
This person has expertise concerning the
product Likert, 1-9, disagree / agree
This person is reliable Likert, 1-9, disagree / agree
I would like to try this product Likert, 1-9, disagree / agree I would like to buy this product Likert, 1-9, disagree / agree
