AMA 2012 Winter Educators' Conference

ADMIRATION: AN IMPORTANT DETERMINANT OF CELEBRITY EFFECTIVENESS Michael Basil, University of Lethbridge ABSTRACT This research examined celebrity admiration. A survey of 48 celebrities found that admiration was not related to attractiveness, but is related to similarity and identification. Next, an experiment with twelve constructed ads depicting six celebrities found that admiration was more closely related to similarity and identification than attractiveness. INTRODUCTION One of the most commonly employed strategies used to advertise consumer products is the use of a celebrity to endorse the product (Agrawal and Kamakura 1995). Celebrities are used frequently because we believe, and have evidence, both academic and proprietary, that celebrities are effective at selling products. We can even make some educated guesses as to why that is. For at least 2,500 years, people have speculated on what makes a speaker persuasive. Aristotle proposed that ethos was “the speaker’s power of evincing a personal character. . .” (1941, p. 1318). He believed that ethos is the largest determinant of the success of a persuasive attempt. Therefore, Aristotle believed that a source that is high in status, yet still similar to the audience, is the best possible source of a message. A celebrity is often seen as having a combination of status and similarity. Social scientists began to examine the importance of the spokesperson as far back as the 1950s. The Yale studies (Hovland, Janis, and Kelly 1953) tried to identify speakers who were successful and identify the traits that made them “credible.” Since the early studies, a considerable literature accrued in this area. Researchers have generally confirmed the importance of credibility as a critical factor in persuasion (e.g., Chaiken and Maheswaran 1994; Wu and Shaffer 1987). However, the definition of “credibility” often makes it difficult to develop a priori predictions of speaker attributes. With regard to the academic literature, a celebrity endorser is “any individual who enjoys public recognition and who uses this recognition on behalf of a consumer good” (McCracken, 1989, p. 310). There are a variety of familiar studies which demonstrate that celebrities are effective. Kahmen, Azhari, and Kragh (1975) found that advertisements featuring Johnny Cash led to greater awareness of the ads and a more positive corporate image. Friedman, Termini, and Washington (1977) found that American Marketing Association / Winter 2012

wine ads with Al Pacino were rated higher than those with a company president, a typical consumer, or no source. Friedman and Friedman (1979) found that ads using Mary Tyler Moore as a spokesperson led to higher ratings of the ad, attitude toward the product, and purchase intention for products involving image or taste, and led to better ad and brand name recall regardless of the product. Atkin and Block (1983) found that the effects of celebrity-based alcohol advertisements were more effective than noncelebrity-based advertisements only with younger subjects. More generally, Graham (1991) found that celebrity charisma creates “sheeplike, highly motivated followers” (p. 105). Celebrities have been shown to be so powerful that when a famous person commits suicide it can increase suicide rates (Stack 1987, 1990) and cocaine use (Kleber 1988). When applied to marketing and advertising, a variety of theories have been put forth (Erdogan 1999). For the purpose of this study three major theories of celebrity effects are most relevant – (1) credibility (2) attractiveness, (3) match-up, and (4) identification. Each of these proposes a slightly different reason for the effectiveness of celebrities. The theories themselves, the proposed mechanisms, and previous research on each is explained below. One of the main theories of celebrity effectiveness stems from Aristotle’s idea of source credibility. According to this approach, credibility is critical to how an endorser is perceived (Erdogan 1999). Typical attributes of source credibility are trustworthiness and expertise. Again, varying constructions of source credibility often makes it difficult to demonstrate test or falsify these predictions (Erdogan 1999). In 1985, Kahle and Homer predicted that celebrities would be more effective endorsers of beauty-related products because they are more attractive than non-celebrities. However, the results were not completely supportive of their prediction and instead it was found that both razor ads (a “beauty-related” product) and toothpaste ads featuring attractive celebrities resulted in more positive attitudes toward the ad than ads featuring less attractive celebrities. Thus, attractiveness of the celebrity can have an effect on attitudes toward the product and purchase intentions. In 1990, Kamins proposed that celebrities are more effective than non-celebrities because they offer special 229

expertise. Misra (1990) described this effect as a “match” between the product and celebrity. To the extent that this expertise relates to the product, they are more effective. This is usually referred to as the “match-up hypothesis” (Kamins 1990). In support of the importance of match-up, research has shown that a celebrity with expertise about the product was significantly more effective than a noncelebrity expert (Buhr, Simpson, and Pryor 1987; Ohanian 1991). These results generally indicate that a celebrity is more effective than a non-celebrity when there is some level of expertise offered by the celebrity. More recently, Brown and Basil (1995) proposed that celebrities are more effective than non-celebrities because people are likely to see the celebrity as a friend. Previous evidence supports the assertion that both friends and similar others are more influential than unknown others (Rogers 1995). As a result, this identification explains why celebrities are more effective than non-celebrities. This is usually referred to as the “identification hypothesis” (Basil 1996). Research by Basil and colleagues has examined the relationship between the level of identification and attitudinal and behavioral outcomes using several celebrities including Magic Johnson (Basil 1996; Basil and Brown 1997; Brown and Basil 1995), Mark McGwire (Brown, Basil, and Bocarnea 2003a), and Princess Diana (Brown, Basil, and Bocarnea 2003b). In all of these instances, the level of identification mediated behavioral outcomes. This mediation effect is not only supportive of the identification process, but also supports identification itself as the likely underlying mechanism. Overall, the results of this research support the effectiveness of celebrities, and they offer their own explanations for the effectiveness of celebrities. But have they really captured the mechanism? One attempt to identify whether we had captured the various forms of credibility was attempted by Ohanian (1990). In her study, she suggested two general factors – credibility (expertise and trustworthiness) and attractiveness (familiarity, likeability, similarity, and attractiveness). Her results support attractiveness, trustworthiness, and expertise as separate components of credibility. However, as she acknowledges, these factors were derived from previous theoretical work, and there is no way of knowing whether these dimensions are exhaustive or not. Regardless of which method has been applied, then, we have been limited to our hunches concerning what makes a celebrity effective. But have we captured all of the reasons for their effectiveness, and have we identified all of the mechanisms? An alternative approach to this question was to examine practitioners’ reasons for selecting celebrities. In one application of this approach, Erdogan, Baker, and Tagg (2001) found that practitioners reported selecting celebrities for a variety of factors, including their trustworthi-

230

ness, expertise, and image. Interestingly, the important factors for advertising agencies are similar to those posed by experimental studies. However, this may be partly because the practitioners had learned these theories in school, and it may be partly due to the fact that the questionnaire was constructed with these studies in mind. ADMIRATION One mechanism through which people can be effective in influencing others that seems to be missing from the research to date is admiration. To admire is defined in the dictionary as to “esteem, respect, or look up to.” Some research in social science has examined peoples’ admiration of heroes (Allen 1975; Arzhakova 2000; Gibbon 1999; Torbjor 1998). The idea of admiration is consistent with Aristotle’s opinion that identified status is an important determinant of credibility – to the extent that people admire or look up to another person, that person should be more successful. As evidence, research that has shown that celebrities who endorse multiple products may have their credibility and likeability harmed (Tripp, Jensen, and Carlson 1994). This decline in credibility can best be understood by understanding what is lost – some of the ability to look up to, or admire, the endorser. Because almost all of the previous research has focused on celebrities such as movie stars and television celebrities, attractiveness, and identification have emerged as important predictors. However, if we were to choose a wider variety of celebrities than movie and television stars, and included politicians with legitimate power and intellectuals, it is quite possible that additional factors would emerge as other determinants of celebrity effectiveness – in this case, admiration. Next, we will compare the admiration concept to the previous theories of celebrity effects. This will help us develop a test of the admiration concept. ATTRACTIVENESS Previous research has shown that attractiveness can have an influence on a whole range of judgments about people, often dubbed “halo” effects (Dion, Berschied, and Walster 1972; Wapnick, Mazda, Darrow 1998, 2000). One interesting halo effect that has received considerable study is that attractiveness is related to a person’s income (French 2002). Attractiveness may be one of the reasons that celebrities may be effective, too. Kahle and Homer (1985) put forth their theory that physical attractiveness would determine consumers’ attitudes and purchase intention of a consumer product. In a test of the hypothesis, they compared likeability versus attractiveness on subjects’ interest in purchasing a razor. The results supported the importance of physical attractiveness, but not likeability.

American Marketing Association / Winter 2012

To the extent that admiration arises from attractiveness and that celebrities are physically attractive, attractiveness should result in higher levels of admiration. H1: Attractiveness will predict the level of admiration of celebrities. SIMILARITY Bandura’s “social learning theory” (1977) and “social cognitive theory” (1986) found that people are more likely to perform modeled behaviors when they see themselves as similar to the person modeling the behavior. These theories propose that when people see others as similar to themselves, they are more likely to imitate the modeled behavior, perhaps including purchasing a product. In studying the diffusion process of innovations, Rogers found that similarity, or “homophily,” enhanced the likelihood of a wide variety of suggested behavior changes (1995, pp. 18–19). In a study of celebrities, Williams and Qualls (1989), for example, found that black consumers identify with black celebrities, perhaps suggesting an increased likelihood of admiration. To the extent that admiration is related to similarity, we should find that: H2: People will be more likely to admire celebrities who are more similar to themselves with regard to (a) Age, (b) Gender, and (c) Race. PARASOCIAL IDENTIFICATION Another potential source of celebrity power that has been proposed is called “parasocial identification.” This phenomenon was first identified in the 1950s when two clinical psychologists, Horton and Wohl (1956), began noticing patients who felt an unrealistic sense of intimacy with television personalities. Research on the identification process has demonstrated that repeated exposure to media figures through the mass media creates a sense of friendship or intimacy in media users (Levy 1979). Audience members commonly look to media personalities as “friends” and those with whom they feel “comfortable.” Evidence of parasocial relationships has been observed between television viewers and newscasters, talk show hosts, and soap opera stars (Levy 1979; Rubin and McHugh 1987). Therefore, this line of research suggests that one reason that celebrities may be effective is through a “faux” friendship from media exposure. Parasocial relationships with media personalities can be said to exist when media audiences think or feel as if they know the media personalities to which they are regularly exposed (Rubin and McHugh 1987). Popular tabloids, magazines, television news–entertainment pro-


grams, and even the mainstream press offer information with which a person can follow his or her favorite celebrity. The mass media, then, may have a special ability to instill this perceived intimacy leading to identification and admiration (Fowles 1992). The research which has examined parasocial identification has found that people often do have a sense of identification with celebrities. Adams-Price and Greene (1990) found that adolescents identify with celebrities, and Greene and Adams-Price (1991) showed that people’s beliefs in a celebrity’s personality were biased as a result of the identification process. Alperstein (1991) found that people’s relationships with television celebrities could be described as a pseudo-social interaction in which viewers are “involved” in this relationship. The greater the identification, the more likely the person will see important attributes in the celebrity. Basil (1996) proposed that a critical determinant of celebrity effectiveness is based on the fact that the public identifies with them as friends. Research has shown that identification does appear to occur through television viewing, print, and radio use (Basil 1996; Basil and Brown 1997; Brown, Basil, and Bocarnea 2002a). Further, this research appears to demonstrate that parasocial identification is a possible mechanism by which people desire to emulate celebrities (Basil 1996; Basil and Brown 1997; Brown and Basil 1995). Thus, a third possible explanation for celebrity power is parasocial identification. If identification is responsible for influencing consumers, we should find that: H3: Media use will predict the level of admiration of celebrities. STUDY 1 Methods This study made use of two data sources to test the predictors of admiration. First, data on admiration levels, similarity, and media use were obtained from the ConsumerStyles survey. This is an annual survey mailed to 10,000 people across the country. A total of 6,065 responded to the survey for a 61 percent response rate. Second, physical attractiveness ratings were obtained by posting photos of the celebrities on picturejudge.com. This website was designed for people to evaluate the physical attractiveness of photographs of people, so this was a perfect outlet for this evaluation. Measures Admiration. Question 2 of the ConsumerStyles survey asked, “Below is a list of people who are liked and

231

admired by many Americans. Which ones, if any, do you admire?” Respondents were asked to “X” any of the 48 celebrities who followed. Similarity. Three questions on the ConsumerStyles survey asked about respondents’ age, gender, and race. These were used to examine the level of respondents’ similarity with celebrities. Their age, gender and race were also recorded (Tiger Woods was the only case of multiple races, and for the purpose of increasing ethnic representation, was coded as black.) A dummy variable was coded so that when the respondent’s age, gender, or race matched that of the celebrity, the appropriate dummy variable (age match, gender match, or race match) was coded as “1" (otherwise it was a “0"). In the case of age, this occurred if the respondent was born within ten years of the celebrity. Physical Attractiveness. Photographs were evaluated by a large number of respondents on a website designed to evaluate physical attractiveness. Two photos of each celebrity were posted. The photos were gathered from the Web by searching for images on google.com. Several criteria were used for the selections. First, the photo had to be without copyright. Second, the photo had to be a color head-and-shoulder shot. Typically, this consisted of publicity stills. If a photo was not head-and-shoulders, it was trimmed to meet this criterion with sufficient resolution for the head-and-shoulders image. Third, the photo was trimmed to an 8 × 5 cm (3 × 2 inch) computer image. The first set of photos was posted in July 2004. The second set of photos was posted in August 2004. This insured that users would not see multiple photos of the same celebrity in close proximity to one another. In fact, in many cases, they were likely to be rated by different users. All images were available on the www.picturejudge.com website and users rated each photo on a 1-to-10 scale (the anchors were an image of ice or fire). The pictures were posted until the “celebrity” category was removed in 2005. The physical attractiveness measures resulted from 66 to 303 responses per celebrity (M = 196 for females, 84 for males). The attractiveness ratings for the two photos were correlated very highly (r = .89). Considering that these were two different photos, this suggests that the measure of attractiveness was consistent across the two photos. For the purpose of this study, we were interested in the overall physical attractiveness of the celebrity, and not of a particular photo, so the separate attractiveness ratings for the two photos were averaged to a single “attractiveness” score. Results Across respondents, the highest level of admiration was for Bill Cosby (53%) and the lowest was for Matt Drudge (1.7%). To test Hypothesis 1, whether attractive232

ness predicted admiration, the unit of analysis was the celebrity (n = 48). As a result, both measures were continuous variables (from 0 to 10 in attractiveness, 0 to 100 percent admiration). For this analysis, a correlation between the mean attractiveness rankings and overall level of admiration was conducted. The correlation between attractiveness and admiration was not significant (r = .007, p = .490). This does not support H1. Splitting the celebrities by gender revealed no relationship between attractiveness and admiration for males (r = .04, n = 11, p = .43) or for females (r = -0.15, n = 19, p = .32). For Hypotheses 2 and 3, the unit of analysis was the respondent. We examined whether or not the respondent identified with each particular celebrity. Because the dependent variable was dichotomous, 48 separate logistic regressions were conducted. Hypothesis 2 predicted that similarity would lead to celebrity admiration. For H2a, the results show that age match was generally not significant (average B = .05, 26 of 48 in the predicted direction). For H2b, gender match was important (average B = .41, 42 of 48 in the predicted direction). For H2c, racial match was also an important similarity factor (average B = .49, 37 of 48 in the predicted direction). Hypothesis 3 predicted that media use would predict celebrity admiration. The results of these regressions showed the largest effects for television (average B = .19), followed by magazines and newspapers (average Bs = .13) and a small effect for radio (average B = .09), thus mostly supporting H3. Discussion First, the lack of a relationship between attractiveness and admiration suggests that admiration appears to be distinct from attractiveness. Attractive people are no more likely to be admired. Second, both racial and gender similarity were related to admiration (with respondents more likely to admire celebrities of their own race and gender); age, however, was not important. This suggests that similarity does appear to relate to admiration. Third, consistent with the identification hypothesis, respondents’ media use was also predictive of celebrity identification. These results support the importance of the mass media in shaping admiration of the celebrities, in line with the parasocial identification hypothesis. The lack of support for Hypothesis 1 (attractiveness), combined with support for Hypotheses 2 and 3 (similarity and identification) suggest admiration is probably a distinct factor from attractiveness (Ohanian 1990). However, the finding that a person is more likely to admire a celebrity who is similar to him or herself than one who is dissimilar suggests that similarity is important in determining admiration. Because our findings also found a American Marketing Association / Winter 2012

relationship with media use, this suggests that the media may play an important role in developing admiration. This may occur through a process by which people can follow their favorite celebrity, and suggests that admiration is related to parasocial identification. An assumption of this paper is that admiration would be important in determining the effects of celebrities in advertising. There are many good reasons to believe that people should be more likely to listen to and heed the advice of someone they admire than someone they don’t. An empirical test of the subsequent effects of admiration, especially its comparison to existing concepts such as attractiveness, expertise, and trustworthiness, on attitude formation and purchase intentions is the next study. Further, this is an opportunity to examine whether admiration is distinct from two typical persuasion measures – credibility and expertise. STUDY 2 An experiment was used to test the effects of admiration, attractiveness, similarity, and identification on adver-tising effectiveness. Pretests were used to select celebrities who varied in attractiveness, credibility, and identification. Eight celebrities were selected. The final study consisted of a 2 (celebrity credibility) * 2 (celebrity attractiveness) * 2 (celebrity identification) * 2 (match-up) mixed experimental design. For the eight celebrities, sixteen advertisements were prepared, half of which were designed to match the expertise of the eight celebrity spokespersons, and half of which had no match. Each subject then viewed eight ads – four of them matched and four mismatched. The between-subjects factor was which set of ads was seen. The eight products were all intended to be low-involvement products. The products or services included cologne, movies, nightclubs, political organizations, and TV programs. The original advertisements were modified by the addition of celebrities and their photos. For most of these ads, the celebrity and content were added to advertisements originally designed without a person. A sample ad is shown in Appendix 1. Subjects. The final sample consisted of 50 people recruited from a university in western North America. Twenty-seven of the subjects (54%) were female, and twenty-three of them (46%) were male. Respondents’ ages ranged from 18 to 53, with an average of 26 years. The majority of respondents were citizens (58%), but a number of other countries were represented – China (30%), England (2%), European countries (2%), Japan (2%), and other (6%). Procedure. Respondents were given an informed consent before they participated in the experiment, telling


them the purpose of the study and asking them to sign “I understand” in the informed consent, but not to sign their name. Respondents who agreed to participate in the experiment were then given the questionnaire booklet. They viewed each advertisement and filled in the corresponding questions. After all eight advertisements, they then filled in the demographic questions. Measures. Scales were used to assess the celebrity endorsers’ credibility, attractiveness, and expertise were adapted from Ohanian (1990). Items were chosen according to the item reliability of the scale. To measure credibility, “dependable” and “trustworthiness” were selected (Cronbach’s α = .87). To measure attractiveness, “attractive” and “handsome/beautiful” were selected (α = .93). To measure expertise, “expert” and “knowledgeable” were selected (α = .91). These questions were given using a seven-point Likert scale. Two questions regarding celebrity endorsers’ identification were selected from Basil’s (1996) research on identification study of celebrity effects (α = .73). As in the first study, a single item was used to measure admiration. In this case, the item was worded “I admire [celebrity’s name]” on the same seven-point Likert scale. Ad outcomes were evaluated using (1) Aad and (2) Interest in learning more about the product or service. Subjects evaluated the ads using a seven-point semantic differential scale with endpoints of bad/good, negative/ positive, and unfavorable/favorable (Davis 1995; α = .96). Subjects also evaluated their interest in learning more using a seven-point Likert scale with two questions asking about their future interest in learning more about the products (α = .95). Results Correlation, regression, and a structural equation analysis were used to examine the relationships between these variables. Hypothesis 1 predicted that attractiveness would predict the level of admiration of celebrities. There was a significant relationship between attractiveness and admiration (r = .45, β = .44, p < .001). However, this was mediated through identification such that attractiveness becomes insignificant when identification is included (β = .04, p = .32). Hypothesis 2 predicted that similarity would predict admiration. For Hypothesis 2a, there was a significant relationship between age similarity and admiration (r = .11, p = .01). Age similarity was not related to admiration when it was the only variable in the regression (β = .01, p = .96) but became larger when identification was included (β = .04, p = .09). For Hypothesis 2b, gender similarity showed the predicted effect (M = 3.6 versus 3.3, p [onetailed] = .04); however, this effect was also eliminated when identification was controlled (β = .03, p = .44).

233

FIGURE 1 Structural equation model of experimental results

Expertise .35 Credibility .35 Attractiveness

-.34

Sex Match

.38

.24 Admiration .30

.17 .59

.65

Aad

.07 Identification Age Similarity

.12

To examine the mediated relationships further, a structural equation model was conducted using AMOS 5.0. The results are shown in Figure 1. Overall, this model shows a good fit (CFI = 767, RMSEA = .206). With regard to the specific relationships being examined, it also suggests that admiration is closely related to identification (path = .65) and attractiveness (path = .59). Admiration is also related to credibility (path = .24). The structural equation model suggests that admiration has independent effects on the intention to gather more information about the product or service in the future (path = .30). Overall, then, the path model suggests that admiration mediates the relationship between credibility and future intention, while identification has a direct effect on Aad. These results support the importance of admiration. Admiration is related to credibility, but more strongly related to identification. Discussion The results of the experiment support the importance of admiration in determining celebrity effectiveness. In this case, admiration is related to most of the same predictors as in the last study – age and gender match and identification. However, this study assessed not only the correlations between these variables themselves, but also their effects on typical outcome measures of attitude 234

.37

.33

Future Intention toward the ad (Aad) and behavioral intention (to find out more). An important outcome was the effect that admiration of the celebrity had on the intention to find out more about the product or service mentioned in the advertisement. This finding seems consistent with the general idea of admiration as a means of establishing credibility for a product or service. When someone admired talks about a product, some of that admiration may transfer to positive thought about the product or service. GENERAL DISCUSSION Overall, the results of these two studies, conducted in very different ways, support the importance of admiration to advertising and marketing. Specifically, it appears that people admire celebrities for a variety of reasons. Admiration is related to age and gender similarity as well as credibility and especially to the level of identification with the celebrity. Therefore, admiration is likely to be an important attribute of celebrities, and should probably be used in their selection and evaluation. More theoretically, the results are consistent with the importance of celebrity identification (Basil 1996). Here, we found that identification is a critical factor determining the effectiveness of a celebrity. Given the relationship between credibility and admiration shown in the second American Marketing Association / Winter 2012

study, there is some evidence that admiration may be a separate but closely related aspect of credibility. In the case of the celebrities chosen here, attractiveness had no relationship with admiration for the variety of celebrities

REFERENCES Allen, B.P. (1975), “Social Distance and Admiration Reactions of ‘Unprejudiced’ Whites,” Journal of Personality, 43, 709–26 Alperstein, N.M. (1991), “Imaginary Social Relationships with Celebrities Appearing in Television Commercials,” Journal of Broadcasting and Electronic Media, 35, 43–58. Agrawal, J. and W.A. Kamakura (1995), “The Economic Worth of Celebrity Endorsers: An Event Study Analysis.” Journal of Marketing, 59 (3) 56–62. Arzhakova, I.S. “The Objects of Upper-Grade Students’ Admiration and Identification,” Russian Social Science Review, 41, 23–28. Atkin, C. and M. Block (1983), “Effectiveness of Celebrity Endorsers,” Journal of Advertising Research, 23, 57–61. Bandura, A. (1977), Social Learning Theory. Englewood Cliffs, NJ: Prentice-Hall. ____________ (1986), Social Foundations of Thought and Action: A Social-Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall. Basil, M.D. (1996), “Identification as a Mediator of Celebrity Effects,” Journal of Broadcasting and Electronic Media, 40, 478–95. ____________ and W.J. Brown (1997), “Marketing Aids Prevention: The Differential Impact Hypothesis Versus Identification Effects,” Journal of Consumer Psychology, 6, 389–411. Brown, W.J. and M.D. Basil (1995), “Media Celebrities and Public Health: Responses to ‘Magic’ Johnson’s HIV Disclosure and its Impact on Aids Risk and High-Risk Behaviors,” Health Communication, 7, 345–70. ____________, ____________, and M. Bocarnea (2003a), “The Influence of Famous Athletes on Health Beliefs and Practices: Mark McGwire, Child Abuse Prevention, and Androstenedione,” Journal of Health Communication, 8, 41–57. ____________, ____________, and ____________ (2003b), “Social Influence of an International Celebrity: Responses to the Death of Princess Diana,” Journal of Communication, 53, 587–605. Buhr, T.A., T.L. Simpson, and B. Pryor (1987), “Celebrity Endorsers’ Expertise and Perceptions of Attractiveness, Likeability, and Familiarity,” Psychological


chosen in Study 1. Attractiveness had some relationship in the case of the celebrities selected in Study 2, but this disappeared when identification was controlled. In both of these cases, however, the attractiveness that is experienced appears to depend on the level of identification with the celebrity.

Reports, 60, 1307–09. Dion, K.K., E. Berschied, and E. Walster (1972), “What is Beautiful is Good,” Journal of Personality and Social Psychology, 24, 285–90. Erdogan, B.Z., M.J. Baker, and S. Tagg (2001), “Selecting Celebrity Endorsers: The Practioner’s Perspective,” Journal of Advertising Research, 41 (3), 39– 48. Fowles, J. (1992), Star Struck: Celebrity Performers and the American Public. Washington, DC: Smithsonian Institution Press. French, M.T. (2002), “Physical Attractiveness and Earnings: Further Evidence,” Applied Economics, 34, 569–72. Friedman, H., S. Termini, and R. Washington (1977), “The Effectiveness of Advertisements Using Four Types of Endorsers,” Journal of Advertising, 6, 22– 24. ____________ and L. Friedman (1979), “Endorser Effectiveness by Product Type,” Journal of Advertising Research, 19, 63–71. Gibbon, P. H. (1999, June 15), “The End of Admiration: the Media and the Loss of Heroes,” Vital Speeches of the Day, 65 (17), 523–25. Horton, D. and R.R. Wohl (1956), “Mass Communication and Parasocial Interaction: Observations on Intimacy at a Distance,” Psychiatry, 19, 215–29. Hovland, C.I., L.L. Janis, and H.H. Kelley (1953), Communication and Persuasion. New Haven, CT: Yale University Press. Kahle, L.R. and P.M. Homer (1985), “Physical Attractiveness of the Celebrity Endorser: A Social Adaptation Perspective,” Journal of Consumer Research, 11, 954–961. Kahmen, J., A. Azhari, and J. Kragh (1975), “What a Spokesman Does for a Sponsor,” Journal of Advertising Research, 16, 17–24. Kamins, M.A. (1990), “An Investigation into the ‘Matchup’ Hypothesis in Celebrity Advertising: When Beauty May Be Only Skin Deep,” Journal of Advertising, 19, 4–13. Misra, S. (1990), “Celebrity Spokesperson and Brand Congruence: An Assessment of Recall and Affect,” Journal of Business Research, 21, 159–73. Ohanian, R. (1990), “Construction of a Scale to Measure Celebrity Endorser’s Expertise, Trustworthiness, and Attractiveness,” Journal of Advertising, 19, 39–52.

235

____________ (1991), “The Impact of Celebrity Spokespersons’ Perceived Image on Consumers’ Intention to Purchase,” Journal of Advertising Research, 31, 46–54. Rogers, E.M. (1995), Diffusion of Innovations, 4th ed. New York: Free Press. Rubin, A.M., E.M. Perse, and R.A. Powell (1985), “Loneliness, Parasocial Interaction, and Local Television News Viewing,” Human Communication Research, 12, 155–80. ____________ and M.P. McHugh (1987), “Development of Parasocial Interaction Relationships,” Journal of Broadcasting and Electronic Media, 31, 279-92. Stack, S. (1987), “Celebrities and Suicide: a Taxonomy and Analysis, 1948–1983,” American Sociological Review, 52, 401–12. ____________ (1990), “Audience Receptiveness, the Media, and Aged Suicide, 1968–1980,” Journal of Aging Studies, 4, 195–209. Tripp, C., T.D. Jensen, and L. Carlson (1994), “The Effects of Multiple Product Endorsements by

236

Celebrities on Consumers’ Attitudes and Intentions,” Journal of Consumer Research, 20, 535–47. Wapnick, J., J. Mazza, and A.A. Darrow (1998), “Effects of Performer Attractiveness, Stage Behavior, and Dress on Violin Performance Evaluation,” Journal of Research in Music Education, 45, 470–79. ____________, ____________, and ____________ (2000), “Effects of Performer Attractiveness, Stage Behavior, and Dress of Children’s Piano Performances,” Journal of Research in Music Education, 46, 510–21. Williams, J.D. and W.J. Qualls (1989), “Middle-Class Black Consumers and the Intensity of Ethnic Identification,” Special Issue: Psychology, Marketing, and the Black Community. Psychology and Marketing, 6, 263–86. Wu, C. and D.R. Shaffer (1987), “Susceptibility to Persuasive Appeals as a Function of Source Credibility and Prior Experience with the Attitude Object,” Journal of Personality and Social Psychology, 52, 677–88.


APPENDIX 1 Sample Ad

For further information contact: Michael Basil University of Lethbridge 4401 University Drive W. Lethbridge, AB T1K 3M4 Canada Phone: 403.329.2075 Fax: 403.329.2038 E-Mail: [email protected]


237