The Benefit of Celebrity Endorsement
Allison Nagler, David Ma, and Isabella Rosales
Since the emergence of television and radio in the early twentieth century, advertisements have filled the airwaves. Perhaps the most conspicuous were those featuring endorsements from celebrities of varying types. Daniel J. Boorstin said, “a sign of a celebrity is that his name is often worth more than his services” (Augustine, 2010). Thus, celebrities can have persuasive power by simply putting their name on the product, regardless of their talent. This trend of celebrity advertisement has been maintained throughout the decades, and it is considered a highly effective strategy today. It is difficult to say, however, whether male or female celebrities are more effective as endorsers. Depending on the gender of the consumer and the product, either one has the potential to be more effective.
The value of utilizing celebrities to endorse products has been a topic of study for several decades, and more studies are being conducted to further determine its effect. Kamins (1989) examined the effectiveness of celebrity endorsed products. The product used for this study was a home computer. The participants for this study were 77 graduate students. This study was administered with four experimental conditions where a questionnaire was handed out with scale ratings about their reactions to product-related characteristics, which includes topics such as, “global overall attitude toward the brand and the advertisement, spokesperson attractiveness, believability, familiarity, likability, and purchase intention” (Kamins, 1989, p. 38). The findings of the research suggest that brand attitude is enhanced due to the utilization of celebrities in advertising. Hence, consumers’ positive response to the advertisement was influenced by the presence of a celebrity endorser.
Advertisement agencies have discovered the value of utilizing celebrity endorsers who target young adults, as celebrities are often perceived as role models. Dix, Phau, and Pougnet (2010) did a study to investigate how sports celebrities could be perceived as role models and influence young adult consumers' purchase and behavioral intentions. The participants for this study were 207 young adults, aged 17-25, who completed a survey comprised of an “athlete role model influence scale”, a “purchases and behavioral intentions scale” as well as a section asking about the demographic of the respondent. One of the most significant results that the authors found was that a sizeable amount of young adults surveyed believed that their favorite athletes were good models for them to follow in terms of work ethic and behavior to imitate. If kids think behavior of an athlete is something that should be mimicked, then watching athletes consume and use a product is likely to be imitated as well. One might conclude that, given the hardworking and successful nature of sports celebrities, they are very useful in advertisement geared towards young adults. And in fact, there has been a spike in celebrity advertising in recent years. This trend shows that advertisements without some sort of celebrity attachment could be on their way out and may be losing effectiveness for the future.
As celebrity athlete endorsements have been steadily increasing, discovering what makes them particularly effective has been under investigation. Shuart (2007) tested the impact of celebrity endorsement on consumer interests of purchasing Ultima sports drink.. The results suggested that someone who is considered both a hero and a celebrity is the most effective spokesperson for a product. Additionally, it was found that while source credibility are important predictors of purchase intention, the perceived celebrity status of and heroism of the endorser were the strongest predictor and contributor to the persuasion of product purchase. this study suggests that marketers take advantage and differentiate between the endorsement power of being considered a celebrity and a hero.
It important for the celebrity endorser to be matched appropriately with the product they are selling. It is also important that they are perceived to like the product they are advertising in order to persuade consumers to purchase it. Silvera and Austed (2004) investigated whether consumers infer that endorsers like the products they endorse. In two experiments, 66 college students were given a celebrity endorsed advertisement and subsequently asked to fill out a questionnaire that measured inferences made about the correspondent (measured by asking the participants if they believed the celebrity liked the product, used the product, and assessed their own views of the product), attitude toward the advertisement, attitude toward the product, and attitude toward the celebrity endorser. The results of both experiments demonstrated that the average person was viewed as liking the product more than the celebrity endorser, indicating a reversal of the correspondence bias. As such, both experiments highlight the importance of positive attitudes toward the endorser in predicting attitudes toward the product.
Utilizing an endorser that has appeal to the targeted consumers helps increase positive evaluations of a product. Klaus and Bailey (2008) investigated how consumers’ responses to celebrity endorsements would be affected by gender. Both the gender of the celebrity endorser and the gender of the consumer were taken into consideration. The researchers employed booklets that provided advertisements portraying two American professional soccer players, Mia Hamm and Landon Donovan, on two separate advertisements. Participants were asked to provide feedback about their ad-preferences by means of a brief survey. Overall, people of both gender responded more favorably to the ad featuring a female celebrity. This study aims to determine the extent consumers respond to celebrity advertisements, and to what extent their responses depend on both their own gender and the gender of the celebrity. In the past this issue has been explored to a degree, however, given the current state of technology and increased usage of social mediums, college-aged individuals seem to be more susceptible than ever to advertising strategies. According to Belch and Belch (2001), “From 1979 to 1997, the popularity of using celebrities in advertising in the USA increased from 15 to 25 percent and currently, more than 20 percent of all TV commercials feature celebrities.” This experiment seeks to shed light on why this might be. The media has recently been inundated with numerous celebrity endorsements, and many may be evaluated as disingenuous due to the lucrative nature of the celebrity status. There is an implicit belief in which it is assumed celebrities simply advertise for the money. Therefore, the efficacy of the celebrity endorsement may have decreased since its initial conception. Considering that the target-demographic for the study is college students who tend to receive a significant amount of information through the media, this research should serve to determine how college students, in a media ridden world, are influenced by celebrity endorsers
It was hypothesized that participants will have a more favorable attitude toward the advertisement with a celebrity endorser than the advertisement without a celebrity endorser. Female participants will have a more favorable attitude toward the female celebrity-endorsed advertisement than the male celebrity-endorsed advertisement. Male participants will have a more favorable attitude toward the advertisement with the male celebrity endorser than the advertisement with the female endorser. The female celebrity-endorsed advertisement will yield higher scores on the Source Credibility Scale and Advertisement, Product and Endorser Attitude Scalethan the male celebrity-endorsed advertisement because we believe women desire to identify with a female celebrity and men will be attracted to the female celebrity..
Method
Participants
Participants were Connecticut College students, predominantly Caucasian, with 11% consisting of other races, ranging from ages 18 to 22. Participants were recruited through classes and received course credit for their participation or they were recruited in the campus library and received candy for their participation.. Seventy-two participants were obtained, with 24 randomly assigned to each of the three experimental groups. There were 32 males and 40 females , which is representative of the ratio between males and females on the college campus. There were three experimental groups; one with a male endorsed advertisement of SmartWater, one with a female endorsed advertisement of SmartWater, and a control group an advertisement for SmartWater without an endorser. SmarWater was chosen because it is gender neutral.
Materials
Source Credibility Scale(Ohanian, 1990). The celebrity endorser’s source credibility is characterized by three dimensions: the sources attractiveness, trustworthiness, and expertise. Each dimension of source credibility is composed of five differential items scored on a 7-point scale. Thus, the measure is multidimensional, and scores on each dimension are derived by summing the responses per items within each dimension. Attractiveness refers to the perceived physical attractiveness of the endorser to the consumer and, to a lesser extent, the emotional attractiveness of the source to the consumer. This includes elements of physical beauty, sexiness and elegance (i.e. Ugly 1 2 3 4 5 6 7 Beautiful). Trustworthiness is defined as the degree of confidence in the communicator’s intent to communicate the assertions he or she considers most valid. This includes both trust and acceptance of the endorser and message (i.e. Dishonest 1 2 3 4 5 6 7 Honest). Expertise is defined as the extent to which the endorser is perceived to be a source of valid assertions about the advertisement. This includes the source’s competence, expertise, and qualifications with regard to the message (i.e. Inexperienced 1 2 3 4 5 6 7 Experienced). Construct reliability from confirmatory factor analysis showed strong internal consistency for the three subscales. In Ohanian’s study, the Cronbach’s alpha was .89 for attractiveness, .89 for trustworthiness, and .89 for expertise. In this study, the Cronbach’s alpha was .77 for attractiveness, .87 for trustworthiness and .91 for expertise.
Advertisement, Product and Endorser Attitude Scale(Silvera & Austad, 2004;). All attitude items used semantic differential scales with a value of one associated with the more negative word and a value of seven associated with the more positive word. Attitude toward the advertisement was measured using the items pleasant/unpleasant, likeable/not likeable, interesting/uninteresting, and good/bad; attitude toward the product was measured using the items desirable/not desirable, pleasant/unpleasant, likeable/not likeable, and good/bad; attitude toward the endorser was measured using the items interesting/uninteresting, pleasant/unpleasant, likeable/not likeable, and good/bad. Each of the three attitude measures showed satisfactory internal reliability. Attitude toward advertisement had a Cronbach’s alpha of 0.85. Attitude toward the product had a Cronbach’s alpha of 0.82. Attitude toward the endorser had a Cronbach’s alpha of 0.87. In this study, the Cronbach’s alpha was .87 for attitude toward the advertisement, .92 for attitude toward the product, and .89 for attitude toward the celebrity.
Procedure
Participants were assigned to one of three groups depending on the order they arrived in. The first group received a non-endorsed advertisement, the second group received the female celebrity endorsed advertisement (Jennifer Aniston), and the third group received the male celebrity endorsed advertisement (Tom Brady). A packet containing the materials for the experiment was distributed. Packets were handed out so that every third participant entering the study was in the same experimental group. Group one filled out the Advertisement, Product and Endorser Attitude Scale (Silvera & Austad, 2004). Both group two and three filled out the Source Credibility Scale (Ohanian, 1990) and the Advertisement, Product and Endorser Attitude Scale. The packet contained the advertisement on top, followed by the questionnaires corresponding to that particular advertisement, and lastly, a demographics questionnaire . The demographics questionnaire asked general information about themselves and their degree of media consumption and exposure.
Results
In order to test the hypothesis that participants will have a more favorable attitude toward the advertisement with a celebrity endorser than the advertisement without a celebrity endorser, two 3(group) x 2(gender) Factorial ANOVAs were conducted on both the advertisement attitude and product attitude. A significant main effect for group was found for the advertisement attitude score, F (2,65) = 6.12, p .01, as the female endorser (M = 5.26, SD = .98) and male endorser (M = 5.27, SD = 1.29) received significantly higher scores on advertisement attitude than the advertisement without a celebrity endorser (M=4.29, SD=.90). Similarly, there was a significant main effect for group for the product attitude score, F (2,66) = 4.85, p = .01, as the female endorser (M=5.33, SD=.96) and male endorser (M=5.45, SD=1.06) received significantly higher scores on product attitude than the advertisement without an endorser (M=4.47, SD=1.37). Thus, the participants had a significantly more favorable attitude toward the ad with the celebrity than the ad without the celebrity.
In order to test the two hypotheses that female participants will have a more favorable attitude toward the female celebrity-endorsed advertisement than the male celebrity-endorsed advertisement and that male participants will have a more favorable attitude toward the advertisement with the male celebrity endorser than the advertisement with the female celebrity, the same two 3(group) x 2(gender) Factorial ANOVA were performed. While there was a significant main effect for group, as mentioned above, there was no main effect for gender for either the advertisement attitude score, F (1.65) = .04, p = .85, or product attitude score, F = (1,66)= .14, p = .71. There was also no interaction effect for attitude toward the advertisement, F (2,65) = .57, p = .57 or for attitude toward the product, F (2,66)= .56, p = .57. Therefore, neither hypothesis regarding gender was supported.
Three independent samples t-tests were conducted to test the hypothesis that the female celebrity-endorsed advertisement will yield higher scores than the male celebrity-endorsed advertisement. The tests revealed no differences on the three dimensions of source credibility, attractiveness, t (46) = .52, p = .61, trustworthiness, t (45) = .48, p= .64, or expertise, t (45)=.82, p = 42. There were also no differences on celebrity attitude, t (46) = .71, p = .49.The hypothesis was not supported.
Discussion
It was hypothesized that participants would have a more favorable attitude toward the advertisement with a celebrity endorser than the advertisement without a celebrity endorser. This hypothesis was supported, as participants reported significantly higher scores for the celebrity-endorsed advertisements than the advertisement that only featured the product. It was also hypothesized that female participants would have a more favorable attitude toward the female celebrity-endorsed advertisement than the male celebrity-endorsed advertisement. In addition, it was hypothesized that male participants would have a more favorable attitude toward the advertisement with the male celebrity endorser than the advertisement with the female endorser. Neither of these hypotheses was supported. No conclusions can therefore be drawn about the gender of the participants and their preference toward a male or female celebrity based on this chosen product, as there was no significant difference between the way in which males and females responded. However, conclusions may be able to be drawn based on the use of a product that is more gender related. Lastly, it was hypothesized that the female celebrity-endorsed advertisement would yield higher scores than the male celebrity-endorsed advertisement. This hypothesis was also not supported, demonstrating that there were no significant differences between the scores of the male celebrity endorsed advertisement and the female celebrity endorsed advertisement. Participants perceived the credibility of Jennifer Aniston and Tom Brady equally. Product advertisement with celebrity endorsements was more effective than product advertisement without celebrity endorsements.
While Klaus and Bailey (2008) found that participants of both genders responded more favorably to an advertisement featuring a female celebrity, the present study hypothesized that participants will favor the advertisement with the celebrity of the same gender based on the findings of Shuart (2007). In line with Shuart's findings that that an individual who is considered both a hero and celebrity is the most effective spokesperson for a product, it was believed that participants would see a hero in someone of the same gender, as it is easier to relate and aspire to be like an individual similar to oneself than an individual different from oneself (Citation). However, the hypotheses regarding gender of the participants were not supported in the present study, meaning that there were no differences in advertisement preference based on the participants’ gender. These results suggest that the gender of the audience may not be as important of a factor in considering what type of celebrity to utilize as previously believed. In today’s modern society, women and men may be socialized in a way in which they can identify just as easily with the opposite gender as with their own gender. Thus, gender defined group differences and identities are not as exaggerated as they once used to be.
Rossiter and Smidts (2010) conducted a similar study that analyzed celebrity advertisements. Similar to the present study, a control group with a product-only advertisement was shown to participants. Additionally, various pairings of celebrities and products were shown to participants. The results of the present study are complement the results of Rossiter and Smidts' research in that, celebrity-endorsed advertisements can be more appealing to consumers than product-only advertisements. There were however, additional findings showing that celebrities who were "matched poorly" with the products they were endorsing, were actually detrimental to the efficacy of the advertisement. Perceived trustworthiness and expertise in using the product were two of the most important factors to participants determining their opinions of the advertisement. Ultimately, certain celebrity-product pairings were highly effective, while others actually dissuaded participants from consuming the products. Rossiter and Smidts' participants were shown more varied celebrity-product pairings than the participants in the present study which may have contributed to their findings. It would be useful to implement similar degrees of variety in future research in the effect of celebrity-product pairings.
There were several limitations and confounding variables in this study. First, the packet containing the advertisement and surveys did not have a cover page. It was possible that some participants were able to see what advertisements their peers in different groups were viewing, in addition to their own. Knowledge about the nature of the other advertisements being shown could have affected participants’ responses. Additionally, the celebrities used in this study are fairly well known, but they are celebrities from different industries. Tom Brady is a professional football player and Jennifer Aniston is a famous actress. This difference between the type of celebrity could have skewed participants’ recognition and opinions of them, and was thus a confounding variable. Also, participants may have had preconceived notions about the celebrities, such as football fans disliking or liking Tom Brady depending on if they are New England Patriots fans. On the source credibility scale, participants may have been biased to what they already know of Tom Brady and may have chosen their answers accordingly, completely independent of his relation to the advertisement and product. The same situation is applicable to Jennifer Aniston. For example, if the participant enjoys the hit TV show Friends, in which she stars, he or she will have positive attitudes toward her and give her higher scores. This could have contributed to why there were no significant gender differences found.
Furthermore, many participants took part in this study due to the research participation requirements of certain classes. Because many participants were anxious to finish the study, they may have misinterpreted questions on the demographics survey. For example, many participants had asked for clarification on certain questions such as “What do you find most appealing about an advertisement you are viewing?” While their confusion may have been a function of their desire to complete the survey, it could also be contributed to awkward or confusing wording on the part of the researchers. Finally, the participants for the study were initially only volunteers from the Connecticut College Psychology 101 and 102 subject pool. However, there were not enough of these volunteers to conduct the appropriate statistical tests, so the second half of the participants were recruited while entering or leaving Shain Library.
The results of the study indicate that marketers should invest the necessary funds to utilize celebrity endorsement, as it was proven more effective than advertisement featuring only the product. Additionally, because there were no significant differences between the scores of the female endorser and the male endorser, gender may not play as large a role in persuasion as was initially expected. Both celebrity endorsers scored significantly higher on attitudes toward the advertisement and attitudes toward the product than the advertisement without the endorser, so gender may not be an important factor to the consumer. Rather, credibility is a more crucial factor for marketers and advertisers to consider and focus on, as the perceived appeal of the person endorsing the product may significantly affect how positively viewed the advertisement is. Due to the fact that actual advertisements were utilized in this study, the results have external validity. Because these advertisements are distributed to the general population in numerous magazines, billboards, newspapers, and other print sources, everyone has access to their consumption. Thus, the results are generalizable, which implies that not only can Connecticut College students be influenced by celebrity-endorsed advertisement, but anyone that encounters it in an everyday situation.
In conducting further research, a study could be carried out using celebrities from comparable industries. Because Tom Brady is an athlete and Jennifer Aniston is an actress, they can attract different audiences and thus yield different attitudes toward the product and the advertisement. Another future study that could be performed could be to test the effectiveness of commercials rather than print-advertisements. While still popular, print-ads are a form of media that is being replaced by television and film. Commercials could therefore target a wider audience and perhaps yield results that could provide even greater external validity. In addition, only one product was examined in this study, so future studies could incorporate additional products to see if participants will favor one advertisement over the other and see how it correlates to the celebrity endorser. While it was beneficial to use a water advertisement because of its gender-neutral value, it is not necessarily a product that everyone buys, so other products such as electronics could be tested. An additional consideration of this study is the culture in which the research was performed. The United States runs on consumerism and is vastly interested in popular culture, so the population most often idolizes certain celebrities and product brands. As such, a cross-cultural study could be performed to see if a different population would still be interested in the use of celebrities in marketing. Perhaps other cultures would find celebrities to be an unappealing way to sell a product, and therefore would have unfavorable attitudes toward celebrity-endorsed advertisements.
References
The value of utilizing celebrities to endorse products has been a topic of study for several decades, and more studies are being conducted to further determine its effect. Kamins (1989) examined the effectiveness of celebrity endorsed products. The product used for this study was a home computer. The participants for this study were 77 graduate students. This study was administered with four experimental conditions where a questionnaire was handed out with scale ratings about their reactions to product-related characteristics, which includes topics such as, “global overall attitude toward the brand and the advertisement, spokesperson attractiveness, believability, familiarity, likability, and purchase intention” (Kamins, 1989, p. 38). The findings of the research suggest that brand attitude is enhanced due to the utilization of celebrities in advertising. Hence, consumers’ positive response to the advertisement was influenced by the presence of a celebrity endorser.
Advertisement agencies have discovered the value of utilizing celebrity endorsers who target young adults, as celebrities are often perceived as role models. Dix, Phau, and Pougnet (2010) did a study to investigate how sports celebrities could be perceived as role models and influence young adult consumers' purchase and behavioral intentions. The participants for this study were 207 young adults, aged 17-25, who completed a survey comprised of an “athlete role model influence scale”, a “purchases and behavioral intentions scale” as well as a section asking about the demographic of the respondent. One of the most significant results that the authors found was that a sizeable amount of young adults surveyed believed that their favorite athletes were good models for them to follow in terms of work ethic and behavior to imitate. If kids think behavior of an athlete is something that should be mimicked, then watching athletes consume and use a product is likely to be imitated as well. One might conclude that, given the hardworking and successful nature of sports celebrities, they are very useful in advertisement geared towards young adults. And in fact, there has been a spike in celebrity advertising in recent years. This trend shows that advertisements without some sort of celebrity attachment could be on their way out and may be losing effectiveness for the future.
As celebrity athlete endorsements have been steadily increasing, discovering what makes them particularly effective has been under investigation. Shuart (2007) tested the impact of celebrity endorsement on consumer interests of purchasing Ultima sports drink.. The results suggested that someone who is considered both a hero and a celebrity is the most effective spokesperson for a product. Additionally, it was found that while source credibility are important predictors of purchase intention, the perceived celebrity status of and heroism of the endorser were the strongest predictor and contributor to the persuasion of product purchase. this study suggests that marketers take advantage and differentiate between the endorsement power of being considered a celebrity and a hero.
It important for the celebrity endorser to be matched appropriately with the product they are selling. It is also important that they are perceived to like the product they are advertising in order to persuade consumers to purchase it. Silvera and Austed (2004) investigated whether consumers infer that endorsers like the products they endorse. In two experiments, 66 college students were given a celebrity endorsed advertisement and subsequently asked to fill out a questionnaire that measured inferences made about the correspondent (measured by asking the participants if they believed the celebrity liked the product, used the product, and assessed their own views of the product), attitude toward the advertisement, attitude toward the product, and attitude toward the celebrity endorser. The results of both experiments demonstrated that the average person was viewed as liking the product more than the celebrity endorser, indicating a reversal of the correspondence bias. As such, both experiments highlight the importance of positive attitudes toward the endorser in predicting attitudes toward the product.
Utilizing an endorser that has appeal to the targeted consumers helps increase positive evaluations of a product. Klaus and Bailey (2008) investigated how consumers’ responses to celebrity endorsements would be affected by gender. Both the gender of the celebrity endorser and the gender of the consumer were taken into consideration. The researchers employed booklets that provided advertisements portraying two American professional soccer players, Mia Hamm and Landon Donovan, on two separate advertisements. Participants were asked to provide feedback about their ad-preferences by means of a brief survey. Overall, people of both gender responded more favorably to the ad featuring a female celebrity. This study aims to determine the extent consumers respond to celebrity advertisements, and to what extent their responses depend on both their own gender and the gender of the celebrity. In the past this issue has been explored to a degree, however, given the current state of technology and increased usage of social mediums, college-aged individuals seem to be more susceptible than ever to advertising strategies. According to Belch and Belch (2001), “From 1979 to 1997, the popularity of using celebrities in advertising in the USA increased from 15 to 25 percent and currently, more than 20 percent of all TV commercials feature celebrities.” This experiment seeks to shed light on why this might be. The media has recently been inundated with numerous celebrity endorsements, and many may be evaluated as disingenuous due to the lucrative nature of the celebrity status. There is an implicit belief in which it is assumed celebrities simply advertise for the money. Therefore, the efficacy of the celebrity endorsement may have decreased since its initial conception. Considering that the target-demographic for the study is college students who tend to receive a significant amount of information through the media, this research should serve to determine how college students, in a media ridden world, are influenced by celebrity endorsers
It was hypothesized that participants will have a more favorable attitude toward the advertisement with a celebrity endorser than the advertisement without a celebrity endorser. Female participants will have a more favorable attitude toward the female celebrity-endorsed advertisement than the male celebrity-endorsed advertisement. Male participants will have a more favorable attitude toward the advertisement with the male celebrity endorser than the advertisement with the female endorser. The female celebrity-endorsed advertisement will yield higher scores on the Source Credibility Scale and Advertisement, Product and Endorser Attitude Scalethan the male celebrity-endorsed advertisement because we believe women desire to identify with a female celebrity and men will be attracted to the female celebrity..
Method
Participants
Participants were Connecticut College students, predominantly Caucasian, with 11% consisting of other races, ranging from ages 18 to 22. Participants were recruited through classes and received course credit for their participation or they were recruited in the campus library and received candy for their participation.. Seventy-two participants were obtained, with 24 randomly assigned to each of the three experimental groups. There were 32 males and 40 females , which is representative of the ratio between males and females on the college campus. There were three experimental groups; one with a male endorsed advertisement of SmartWater, one with a female endorsed advertisement of SmartWater, and a control group an advertisement for SmartWater without an endorser. SmarWater was chosen because it is gender neutral.
Materials
Source Credibility Scale(Ohanian, 1990). The celebrity endorser’s source credibility is characterized by three dimensions: the sources attractiveness, trustworthiness, and expertise. Each dimension of source credibility is composed of five differential items scored on a 7-point scale. Thus, the measure is multidimensional, and scores on each dimension are derived by summing the responses per items within each dimension. Attractiveness refers to the perceived physical attractiveness of the endorser to the consumer and, to a lesser extent, the emotional attractiveness of the source to the consumer. This includes elements of physical beauty, sexiness and elegance (i.e. Ugly 1 2 3 4 5 6 7 Beautiful). Trustworthiness is defined as the degree of confidence in the communicator’s intent to communicate the assertions he or she considers most valid. This includes both trust and acceptance of the endorser and message (i.e. Dishonest 1 2 3 4 5 6 7 Honest). Expertise is defined as the extent to which the endorser is perceived to be a source of valid assertions about the advertisement. This includes the source’s competence, expertise, and qualifications with regard to the message (i.e. Inexperienced 1 2 3 4 5 6 7 Experienced). Construct reliability from confirmatory factor analysis showed strong internal consistency for the three subscales. In Ohanian’s study, the Cronbach’s alpha was .89 for attractiveness, .89 for trustworthiness, and .89 for expertise. In this study, the Cronbach’s alpha was .77 for attractiveness, .87 for trustworthiness and .91 for expertise.
Advertisement, Product and Endorser Attitude Scale(Silvera & Austad, 2004;). All attitude items used semantic differential scales with a value of one associated with the more negative word and a value of seven associated with the more positive word. Attitude toward the advertisement was measured using the items pleasant/unpleasant, likeable/not likeable, interesting/uninteresting, and good/bad; attitude toward the product was measured using the items desirable/not desirable, pleasant/unpleasant, likeable/not likeable, and good/bad; attitude toward the endorser was measured using the items interesting/uninteresting, pleasant/unpleasant, likeable/not likeable, and good/bad. Each of the three attitude measures showed satisfactory internal reliability. Attitude toward advertisement had a Cronbach’s alpha of 0.85. Attitude toward the product had a Cronbach’s alpha of 0.82. Attitude toward the endorser had a Cronbach’s alpha of 0.87. In this study, the Cronbach’s alpha was .87 for attitude toward the advertisement, .92 for attitude toward the product, and .89 for attitude toward the celebrity.
Procedure
Participants were assigned to one of three groups depending on the order they arrived in. The first group received a non-endorsed advertisement, the second group received the female celebrity endorsed advertisement (Jennifer Aniston), and the third group received the male celebrity endorsed advertisement (Tom Brady). A packet containing the materials for the experiment was distributed. Packets were handed out so that every third participant entering the study was in the same experimental group. Group one filled out the Advertisement, Product and Endorser Attitude Scale (Silvera & Austad, 2004). Both group two and three filled out the Source Credibility Scale (Ohanian, 1990) and the Advertisement, Product and Endorser Attitude Scale. The packet contained the advertisement on top, followed by the questionnaires corresponding to that particular advertisement, and lastly, a demographics questionnaire . The demographics questionnaire asked general information about themselves and their degree of media consumption and exposure.
Results
In order to test the hypothesis that participants will have a more favorable attitude toward the advertisement with a celebrity endorser than the advertisement without a celebrity endorser, two 3(group) x 2(gender) Factorial ANOVAs were conducted on both the advertisement attitude and product attitude. A significant main effect for group was found for the advertisement attitude score, F (2,65) = 6.12, p .01, as the female endorser (M = 5.26, SD = .98) and male endorser (M = 5.27, SD = 1.29) received significantly higher scores on advertisement attitude than the advertisement without a celebrity endorser (M=4.29, SD=.90). Similarly, there was a significant main effect for group for the product attitude score, F (2,66) = 4.85, p = .01, as the female endorser (M=5.33, SD=.96) and male endorser (M=5.45, SD=1.06) received significantly higher scores on product attitude than the advertisement without an endorser (M=4.47, SD=1.37). Thus, the participants had a significantly more favorable attitude toward the ad with the celebrity than the ad without the celebrity.
In order to test the two hypotheses that female participants will have a more favorable attitude toward the female celebrity-endorsed advertisement than the male celebrity-endorsed advertisement and that male participants will have a more favorable attitude toward the advertisement with the male celebrity endorser than the advertisement with the female celebrity, the same two 3(group) x 2(gender) Factorial ANOVA were performed. While there was a significant main effect for group, as mentioned above, there was no main effect for gender for either the advertisement attitude score, F (1.65) = .04, p = .85, or product attitude score, F = (1,66)= .14, p = .71. There was also no interaction effect for attitude toward the advertisement, F (2,65) = .57, p = .57 or for attitude toward the product, F (2,66)= .56, p = .57. Therefore, neither hypothesis regarding gender was supported.
Three independent samples t-tests were conducted to test the hypothesis that the female celebrity-endorsed advertisement will yield higher scores than the male celebrity-endorsed advertisement. The tests revealed no differences on the three dimensions of source credibility, attractiveness, t (46) = .52, p = .61, trustworthiness, t (45) = .48, p= .64, or expertise, t (45)=.82, p = 42. There were also no differences on celebrity attitude, t (46) = .71, p = .49.The hypothesis was not supported.
Discussion
It was hypothesized that participants would have a more favorable attitude toward the advertisement with a celebrity endorser than the advertisement without a celebrity endorser. This hypothesis was supported, as participants reported significantly higher scores for the celebrity-endorsed advertisements than the advertisement that only featured the product. It was also hypothesized that female participants would have a more favorable attitude toward the female celebrity-endorsed advertisement than the male celebrity-endorsed advertisement. In addition, it was hypothesized that male participants would have a more favorable attitude toward the advertisement with the male celebrity endorser than the advertisement with the female endorser. Neither of these hypotheses was supported. No conclusions can therefore be drawn about the gender of the participants and their preference toward a male or female celebrity based on this chosen product, as there was no significant difference between the way in which males and females responded. However, conclusions may be able to be drawn based on the use of a product that is more gender related. Lastly, it was hypothesized that the female celebrity-endorsed advertisement would yield higher scores than the male celebrity-endorsed advertisement. This hypothesis was also not supported, demonstrating that there were no significant differences between the scores of the male celebrity endorsed advertisement and the female celebrity endorsed advertisement. Participants perceived the credibility of Jennifer Aniston and Tom Brady equally. Product advertisement with celebrity endorsements was more effective than product advertisement without celebrity endorsements.
While Klaus and Bailey (2008) found that participants of both genders responded more favorably to an advertisement featuring a female celebrity, the present study hypothesized that participants will favor the advertisement with the celebrity of the same gender based on the findings of Shuart (2007). In line with Shuart's findings that that an individual who is considered both a hero and celebrity is the most effective spokesperson for a product, it was believed that participants would see a hero in someone of the same gender, as it is easier to relate and aspire to be like an individual similar to oneself than an individual different from oneself (Citation). However, the hypotheses regarding gender of the participants were not supported in the present study, meaning that there were no differences in advertisement preference based on the participants’ gender. These results suggest that the gender of the audience may not be as important of a factor in considering what type of celebrity to utilize as previously believed. In today’s modern society, women and men may be socialized in a way in which they can identify just as easily with the opposite gender as with their own gender. Thus, gender defined group differences and identities are not as exaggerated as they once used to be.
Rossiter and Smidts (2010) conducted a similar study that analyzed celebrity advertisements. Similar to the present study, a control group with a product-only advertisement was shown to participants. Additionally, various pairings of celebrities and products were shown to participants. The results of the present study are complement the results of Rossiter and Smidts' research in that, celebrity-endorsed advertisements can be more appealing to consumers than product-only advertisements. There were however, additional findings showing that celebrities who were "matched poorly" with the products they were endorsing, were actually detrimental to the efficacy of the advertisement. Perceived trustworthiness and expertise in using the product were two of the most important factors to participants determining their opinions of the advertisement. Ultimately, certain celebrity-product pairings were highly effective, while others actually dissuaded participants from consuming the products. Rossiter and Smidts' participants were shown more varied celebrity-product pairings than the participants in the present study which may have contributed to their findings. It would be useful to implement similar degrees of variety in future research in the effect of celebrity-product pairings.
There were several limitations and confounding variables in this study. First, the packet containing the advertisement and surveys did not have a cover page. It was possible that some participants were able to see what advertisements their peers in different groups were viewing, in addition to their own. Knowledge about the nature of the other advertisements being shown could have affected participants’ responses. Additionally, the celebrities used in this study are fairly well known, but they are celebrities from different industries. Tom Brady is a professional football player and Jennifer Aniston is a famous actress. This difference between the type of celebrity could have skewed participants’ recognition and opinions of them, and was thus a confounding variable. Also, participants may have had preconceived notions about the celebrities, such as football fans disliking or liking Tom Brady depending on if they are New England Patriots fans. On the source credibility scale, participants may have been biased to what they already know of Tom Brady and may have chosen their answers accordingly, completely independent of his relation to the advertisement and product. The same situation is applicable to Jennifer Aniston. For example, if the participant enjoys the hit TV show Friends, in which she stars, he or she will have positive attitudes toward her and give her higher scores. This could have contributed to why there were no significant gender differences found.
Furthermore, many participants took part in this study due to the research participation requirements of certain classes. Because many participants were anxious to finish the study, they may have misinterpreted questions on the demographics survey. For example, many participants had asked for clarification on certain questions such as “What do you find most appealing about an advertisement you are viewing?” While their confusion may have been a function of their desire to complete the survey, it could also be contributed to awkward or confusing wording on the part of the researchers. Finally, the participants for the study were initially only volunteers from the Connecticut College Psychology 101 and 102 subject pool. However, there were not enough of these volunteers to conduct the appropriate statistical tests, so the second half of the participants were recruited while entering or leaving Shain Library.
The results of the study indicate that marketers should invest the necessary funds to utilize celebrity endorsement, as it was proven more effective than advertisement featuring only the product. Additionally, because there were no significant differences between the scores of the female endorser and the male endorser, gender may not play as large a role in persuasion as was initially expected. Both celebrity endorsers scored significantly higher on attitudes toward the advertisement and attitudes toward the product than the advertisement without the endorser, so gender may not be an important factor to the consumer. Rather, credibility is a more crucial factor for marketers and advertisers to consider and focus on, as the perceived appeal of the person endorsing the product may significantly affect how positively viewed the advertisement is. Due to the fact that actual advertisements were utilized in this study, the results have external validity. Because these advertisements are distributed to the general population in numerous magazines, billboards, newspapers, and other print sources, everyone has access to their consumption. Thus, the results are generalizable, which implies that not only can Connecticut College students be influenced by celebrity-endorsed advertisement, but anyone that encounters it in an everyday situation.
In conducting further research, a study could be carried out using celebrities from comparable industries. Because Tom Brady is an athlete and Jennifer Aniston is an actress, they can attract different audiences and thus yield different attitudes toward the product and the advertisement. Another future study that could be performed could be to test the effectiveness of commercials rather than print-advertisements. While still popular, print-ads are a form of media that is being replaced by television and film. Commercials could therefore target a wider audience and perhaps yield results that could provide even greater external validity. In addition, only one product was examined in this study, so future studies could incorporate additional products to see if participants will favor one advertisement over the other and see how it correlates to the celebrity endorser. While it was beneficial to use a water advertisement because of its gender-neutral value, it is not necessarily a product that everyone buys, so other products such as electronics could be tested. An additional consideration of this study is the culture in which the research was performed. The United States runs on consumerism and is vastly interested in popular culture, so the population most often idolizes certain celebrities and product brands. As such, a cross-cultural study could be performed to see if a different population would still be interested in the use of celebrities in marketing. Perhaps other cultures would find celebrities to be an unappealing way to sell a product, and therefore would have unfavorable attitudes toward celebrity-endorsed advertisements.
References
- Augustine, N. (2010, August 10). Quotes of the Day/qotd.org. Retrieved May 2, 2011 from http://www.qotd.org/search/search.html?aid=1380
- Belch, G.E., & Belch, M.A. (2001). Advertising and Promotion: An integrated marketing communications perspective. 5th ed. New York: McGraw-Hill.
- Biswas, S., Hussain, M., & O'Donnell, K. (2009). Celebrity endorsements in advertisements and consumer perceptions: A cross-cultural study. Journal of Global Marketing, 22(2), 121-137.
- Dix, S., Phau, I., & Pougnet, S. (2010). “Bend it like Beckham”: The influence of sports celebrities on young adult consumers. Young Consumers. 11(1), 36-46.
- Kamins, M. A. (1989). Celebrity and noncelebrity advertising in a two-sided context. Journal of Advertising Research, 29(3), 34-42.
- Klaus, N., & Bailey, A. (2008). Celebrity endorsements: An examination of gender and consumers' attitudes. American Journal of Business: Applying Research to Practice, 23(2), 53-61.
- Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers’ perceived expertise, trustworthiness, and attraction. Journal of Advertising, 19(3),39-52.
- Rossiter, J.R., & Smidts, A. (2010). Print advertising: Celebrity presenters. Journal of Business Research, 64(7), 657-784.
- Shuart, J. (2007). Heroes in sport: Assessing celebrity endorser effectiveness. International Journal of Sports Marketing & Sponsorship, 8(2), 126-140.
- Silvera, D., & Austad, B. (2004). Factors predicting the effectiveness of celebrity endorsement advertisements. European Journal of Marketing, 38(11-12), 1509-1526.