-I'I. ‘4.
, IL. . Ilhfll
In

. t... V
WWI!
,«luuiq .m

. mar :44.
«hr-vat: v.“

.:
‘._.u..x3( f. 9
.11 s...

. itfitnnhfl. v
c... asphalt). .un. 13
' . 1‘: {I

I Ir}.
thunflnlyu!

'1 |

it... 3A.“ ‘4...

A 4.0.9.11. A.
. ..

a.) .
unintun,

r ..;
.. {€13
gun: I. : Y

.. .J|
.oIrOL

.ul

. O .1
III: . J13!
. v V‘ 11.1.. n v:
a '5 if! \I 'r .
. 33!.1. .of
{SP}-..

Vt [‘1’ d‘ft u 1“ -V a‘ ‘ l i
.. .mruV v .I.vx.cl.~....r.l4lalv .
vu. Vii”:'lll?ll.‘ll|a1£l
.r? in ilv.‘ "ii a
I Nun-LIV}... V I

 

thunk": .
1.51.}... .313. . :
1 . . ‘ :--lu....t1n u .I :l ..
- : .vP .‘ifiltvvr 1.1..le

 

ww's

This is to certify that the

dissertation entitled

ATTRIBUTIONAL APPROACH TO
UNDERSTANDING CELEBRITY/PRODUCT CONGRUENCE EFFECTS:
ROLE OF PERCEIVED EXPERTISE

presented by

Sejung Marina Choi

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in AdvertiSing

 

 

Date 8-14—2002

 

MSU i: an Affirmative Action /Equal Opportunity Institution 0-12771

 

 

UBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

033.513 3 2006
JESUS

U
K!

 

~ :.
im0t280§_
11

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuo.p65—p.15

ATTRIBUTIONAL APPROACH TO
UNDERSTANDING CELEBRITY/PRODUCT CONGRUENCE EFFECTS:
ROLE OF PERCEIVED EXPERTISE

By

Sejung Marina Choi

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Advertising

2002

ABSTRACT
ATTRIBUTIONAL APPROACH TO
UNDERSTANDING CELEBRITY/PRODUCT CONGRUENCE EFFECTS:
ROLE OF PERCEIVED EXPERTISE
By

Sejung Marina Choi

Using celebrities for promoting products is a popular advertising strategy. The
selection of right celebrity endorsers is of great concern to advertisers, given the sheer
cost to secure their participation. Academic literature generally indicates many benefits
from using celebrity endorsers and has proposed several mechanisms to explain the
effectiveness of the celebrity endorsements. Despite the continued efforts to theorize the
celebrity endorsement process, little has been known about the criteria for selecting the
appropriate celebrities to endorse a particular product. Correspondingly, celebrity images
and the match between a celebrity endorser and a product have received the increased
attention both by academics and practitioners.

The purpose of this research was to (I) explore dimensions of celebrity images
that might influence celebrity credibility perceptions and (2) investigate the roles of the
relationship between a celebrity spokesperson and a product endorsed in generating
endorsement effectiveness from an attributional perspective. Two studies were
undertaken to examine the celebrity image dimensions and the celebrity/product
congruence effects. In study 1, a survey was administered to develop and refine a scale to

measure a celebrity’s images. In study 2. an experiment consisting of a 3 (level of

congruence) x 2 (product use claim) factorial design was conducted to examine the
celebrity/product congruence effects.

Results showed that celebrity images are multidimensional and are related to the
perceived celebrity credibility. Four dimensions of the celebrity images were identified
and the scale to measure the celebrity images were developed and validated. The findings
of the second study highlight the importance of the congruence between the celebrity and
the product being endorsed. Congruence influenced the celebrity endorsement
effectiveness in terms Of (l) the process of consumer inference-making in which the
reasons why a particular celebrity becomes associated with the particular brand in a
consumer’s mind are made, and (2) the subsequent effects of these attributions on the
consumer evaluations of the endorser, the ad and the brand involved in the endorsements.
Specifically, congruent associations between the celebrity and the product allow
consumers to attribute the celebrity endorsement motive more to the endorser’s liking for
product and need for image enhancement, but less to his or her motivation to receive
money or some other incentive. As a result, credibility perceptions of the celebrity
endorser, attitude toward the ad, attitude toward the brand, and purchase intent were
enhanced. The findings of this study contribute both at theoretical and practical levels.
Implications both for theory and practice, limitations of this study. and suggestions for

future research are discussed.

Copyright by
SEJUNG MARINA CHOI
2002

ACKNOWLEDGMENTS

First and foremost, I would like to thank my dissertation advisor Dr. Nora J.
Rifon. Without her support and guidance, this work would not have been possible. It
owes much of its quality to her suggestions on how to approach problems and her
profound knowledge of the literature. I greatly appreciate her inspiration, encouragement,
and friendship throughout my life at Michigan State University. I would like to thank the
other members of my dissertation committee, Dr. James W. Dearing, Dr. Bonnie B.
Reece, and Dr. Charles T. Salmon. I deeply appreciate their help in making this
dissertation as complete as possible. They have been inspirational mentors to me,
providing me wisdom and warm concern.

I am grateful to many fellow students of Mass Media Ph.D. program for their
friendship and support. They have been caring friends as well as great colleagues. I
would also like to thank many friends in East Lansing and Ann Arbor for the pleasant
memories in Michigan over the years.

I cannot thank my parents enough for their love and sacrifices. Without their
consistent support and encouragement, I would not have been able to complete this study.
My special gratitude extends to my lovely sisters Sookjung, Yunjung, and Kijung. They
have been a source of my consolation. Also, many thanks to my parents-in-law and
brother-in-law for their support and love. Finally, I would like to thank my husband
Hyung 800 for everything. He has been always there for me loving me, understanding

me, supporting me and completing me.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. viii
LIST OF FIGURES ................................................................................... x
CHAPTER I.

INTRODUCTION ..................................................................................... 1

CHAPTER 2.

CONCEPTUAL BACKGROUND .................................................................. 5
Celebrity Endorsements ..................................................................... 5
Source Credibility Model ..................................................................... 9
Source Attractiveness Model ............................................................. IO
Meaning Transfer Model .................................................................. 12
Match-Up Hypothesis ..................................................................... 14

CHAPTER 3.

THORETICAL F AMERWORK .................................................................. 17
Celebrity Credibility ....................................................................... 17
Celebrity Images and Influences ......................................................... 19
Celebrity and Product Congruence ...................................................... 23
Role of Expertise as a Match-Up Factor ................................................ 25
Celebrity and Product Congruence Based on Expertise .............................. 26
Schematic Approach to Understanding Congruence Effects ......................... 28
Congruence Effects and the Discounting Principle .................................... 30
Facilitation Cues of Celebrity and Product Congruence .............................. 35

CHAPTER 4.

METHODS .......................................................................................... 38
Study 1: Development of a Scale to Measure Celebrity Images ..................... 38

Initial Scale Item Selection ...................................................... 38
Celebrity Selection Procedure ................................................... 39
The Exploratory Phase ............................................................ 40
The Confirmatory Phase ......................................................... 41
Study 2: Testing Hypotheses ............................................................. 41
Participants ......................................................................... 41
Design .............................................................................. 42
Selection of Endorser and Products ............................................. 42
Stimulus Materials ................................................................ 43
Procedure ........................................................................... 44
Dependent Measures .............................................................. 44

vi

CHAPTER 5.

RESULTS ............................................................................................ 50
Study 1: Scale Development .............................................................. 50
Exploratory Factor Analysis for the Celebrity Image Scale ................. 50
Confirmatory Factor Analysis for the Celebrity Image Scale ............... 52

Study 2: Hypothesis Testing .............................................................. 54
Manipulation Checks ............................................................. 54
Congruence ........................................................................ 57
Attributions ........................................................................ 58
Endorser Credibility ............................................................... 61
Attitude Toward the Ad .......................................................... 63
Attitude Toward the Brand ...................................................... 64
Purchase Intent ..................................................................... 64
Gender Differences ............................................................... 65
Summary of Results ............................................................... 66

CHAPTER 6.

DISCUSSION ........................................................................................ 69
Summary of Research ..................................................................... 69
Conclusions ................................................................................. 71

Celebrity Image Dimensions ..................................................... 71
Congruence Perception: Role of Perceived Expertise ........................ 72
Role of Congruence ............................................................... 73
Role of Product Use Cue ......................................................... 74
Implications ................................................................................. 75
Limitations .................................................................................. 76
Future Research ............................................................................ 79

APPENDICES ....................................................................................... 83
Appendix A. Tables ........................................................................ 83
Appendix B. Figures ....................................................................... 95
Appendix C. Sample Questionnaire — Study 1 ........................................ 100
Appendix D. Sample Questionnaire —— Study 2 ....................................... 106
Appendix E. Stimulus Materials — Study 2 ............................................ 113

BIBLIOGRAPHY ................................................................................. 120

vii

LIST OF TABLES

Table 1. Factor Analysis Of Results of Celebrity Images ...................................... 84
Table 2. Factor Loadings Of Celebrity Image Dimensions .................................... 85
Table 3. Correlations of Latent Constructs ...................................................... 86

Table 4. Analysis of Variance: Celebrity/Product Match and Product Use Cue Effects on

Overall Congruence Perception ......................................................... 87
Table 5. Overall Congruence: Means and Std. Deviations .................................... 87
Table 6. Factor Analysis Results of Attributions of Endorser Motives ...................... 88

Table 7. Analysis of Variance: Congruence and Product Use Cue Effects on Attribution
to Money ................................................................................... 89

Table 8. Analysis of Variance: Congruence and Product Use Cue Effects on Attribution
to Product .................................................................................. 89

Table 9. Analysis of Variance: Congruence and Product Use Cue Effects on Attribution
to Image .................................................................................... 89

Table 10. Attributions to Money, Product, and Image: Means and Std. Deviations ....... 90

Table 11. Analysis of Variance: Congruence and Product Use Cue Effects on Endorser
Attractiveness ........................................................................... 91

Table 12. Analysis of Variance: Congruence and Product Use Cue Effects on Endorser
Trustworthiness ......................................................................... 91

Table 13. Analysis of Variance: Congruence and Product Use Cue Effects on Endorser
Expertise ................................................................................. 91

Table 14. Endorser Credibility Perceptions: Means and Std. Deviations .................. 92

Table 15. Analysis of Variance: Congruence and Product Use Cue Effects on Attitude
toward the Ad ........................................................................... 93

Table 16. Analysis of Variance: Congruence and Product Use Cue Effects on Attitude
toward the Brand ........................................................................ 93

viii

Table 17. Analysis of Variance: Congruence and Product Use Cue Effects on Purchase
Intent ..................................................................................... 93

Table 18. Ad Attitude, Brand Attitude and Purchase Intent: Means and Std. Deviations
............................................................................................ 94

LIST OF FIGURES

Figure 1. Effects of Congruence on Attributions ................................................ 96
Figure 2. Effects of Congruence on Ad Effectiveness .......................................... 97
Figure 3. Effects of Congruence and Product Use Cue on Product Attribution ............ 98
Figure 3. Effects of Congruence and Product Use Cue on Image Attribution .............. 99

CHAPTER 1

INTRODUCTION

Advertisers have employed various types ofendorsers including celebrities,
experts, typical consumers, CEOS, and animated-characters in order to promote products
and brands. By far, celebrity endorsements have been the most popular strategy, with
approximately 25% of United States television commercials featuring celebrities (Shimp
2000). In general, celebrity endorsers have been found to be more effective in inducing
favorable attitudes towards brands and purchase intention than non-celebrities (Ohanian
1990). Some specific advantages of celebrity endorsements are their ability to draw the
audience's attention, penetrate the increasing clutter of advertising spots, generate high
recall rates, differentiate product images, and facilitate global marketing communications
(Erdogan 1999; Kaikati 1987; Mieiak and Shanklin 1994). Celebrity endorsers have
drawn special attention due to their unique characteristics such as status, recognition and
popularity, as well as their embodiment of rich meanings not typically found in other
forms of endorsement. Although marketers invest large sums of money to associate their
brands with well-known personalities, the strategy is not risk-free. For example, when a
spokesperson's image becomes tarnished through their misdeeds or controversy (e.g.,
Mike Tyson, O. J. Simpson, and Michael Jackson) companies Often sacrifice their

reputation and risk losing both current and potential customers (Mieiak and Shanklin

1994). In addition, not every pairing ofa celebrity and brand is successful. Advertisers
find some match-ups disappointing, especially when celebrity endorsers, who have
previously proven quite effective for other products, appear to have little effect on their
brand’s performance in the marketplace (e.g., Bill Cosby for E.F. Hutton and Texas
Instruments and John Houseman for McDonald's) (Mieiak and Shanklin 1994; Till and
Busler 1998).

Conventional wisdom is that celebrity endorsement effectiveness is a function of
the fit between the celebrity and the product (Forkan 1980). Accordingly, once the
decision to use a celebrity endorser has been made, determining the right celebrity, the
one with the best fit, becomes essential to the strategy’s success. Presently, there is no
formula for creating a fit, but the effects of celebrity endorsers appear to interact with
product types (Friedman and Friedman 1979). Correspondingly, a stream of research has
examined celebrity endorser effectiveness under the label of "match-up hypothesis."

The core notion of the match-up hypothesis is that endorsements are more
effective when the images or messages the endorsers carry are congruent with the
endorsed products (Kahle and Homer 1985; Kamins 1990; Kamins and Gupta 1994;
Lynch and Schuler 1994; Misra and Beatty 1990; Till and Busler 1998, 2000). Although
research has examined the effects of the link between endorser and product, the evidence
is not as convincing as the idea. The elusiveness of the findings may be due in part to the
murkiness of the fit or match—up concept, and the varied definitions of celebrity/brand
match or congruence that appear in the literature. Primarily, the construct has been

manipulated in terms Of overall images (Kamins and Gupta 1994; Misra and Beatty

1990), physical attractiveness (Kahle and Homer 1985; Kamins 1990; Lynch and Schuler
1994), and expertise (Till and Busler 1998, 2000).

The primary purpose of this research is to further explicate the concept of fit or
congruence and better understand the “match-up” effects in the area of celebrity
endorsements. Of particular interest is the role of perceived endorser expertise that has
recently been recognized as a robust and influential dimension of celebrity/product
congruence (Till and Busler 1998, 2000). Schema theory and attribution theory, which
are proven useful approaches to the understanding of celebrity endorsement effectiveness
(Folkes 1988; Kamins 1990; Lynch and Schuler 1994; Misra and Beatty 1990; Sparkman
1982; Speck, Schumann and Thompson 1988), provide theoretical bases for this study.
From these perspectives, the level ofcelebrity/product congruence influences the
celebrity endorsement effectiveness in terms of (1) the process of consumer inference-
making in which the reasons why a particular celebrity becomes associated with the
particular brand in a consumer’s mind are made, and (2) the subsequent effects of these
attributions on the consumer evaluations of the endorser, the ad and the brand involved in
the endorsements. Congruent associations between the celebrity and product allow
consumers to attribute the celebrity endorsement motive more to the endorser’s liking for
product, but less to his or her motivation to receive money or some other incentive. AS a
result, celebrity credibility and other positive responses to the endorsements will be
enhanced.

Two studies were undertaken to better understand the match-up effects of
celebrity endorsements. In study 1, a survey was administered to develop a celebrity

image scale. The factor analysis results identified four dimensions of the celebrity's

image. In study 2, an experiment was conducted to examine congruence effects in the
celebrity endorsement process. Three advertisements were generated by showing the
same celebrity (a professional tennis player) endorse three different product categories,
thereby creating three different conditions of celebrity/brand congruence. This treatment
was used to assess the effects of match-ups on attributions, endorser credibility, and
typical advertising outcome measures. Results revealed that congruence is the key to
understanding the celebrity endorsement process. Congruence between celebrity and
product enhanced attributions of endorsement motives related to the celebrity’s liking for
the product, celebrity credibility. attitude toward the ad, attitude toward the brand, and
purchase intention.

To the best Of the researcher's knowledge, this study is the first attempt to
integrate two major lines of research on celebrity endorsements: source credibility and
match-up hypothesis to offer a more comprehensive understanding of the celebrity
endorsement process. Additionally, components and effects of celebrity images were
explored with regards to the effectiveness of the celebrityendorsements. Furthermore, the
current research proposed and empirically tested attributional explanations of the
congruence effects of celebrity endorsements that have not been offered by the prior
literature. It is hoped that the present research makes theoretical contributions by
resolving the discrepancies found in the celebrity/brand match-up research and provides
practical contributions in the form of guidelines for the selection of appropriate

celebrities for a variety of products and brands.

CHAPTER 2

CONCEPTUAL BACKGROUND

The purpose of this chapter is to introduce the topic of celebrity endorsements and
provide an overview of the previous literature on the effects of celebrity endorsements. A
summary of relevant research offers a conceptual base on which the present studies have
been developed. The theoretical framework for the current research and hypotheses will

be described in detail in Chapter 3.

C elebrifitv Endorsements

By definition, celebrities refer to individuals who are well known to the public
(Speck, Schumann and Thompson 1988). Examples of celebrity types include sports
figures, actors/actresses and entertainers. These famous personalities with celebrity status
possess high profile, idiosyncratic qualities, and popular images in the eyes Of the public.
Some young consumers are often observed to idolize celebrities and imitate their styles
and behaviors. Celebrity endorsers in advertising are defined as “any individual who
enjoys public recognition and who uses this recognition on behalf of a consumer good by
appearing with it in an advertisement” (McCracken 1989, p.310). This advertising
strategy has been found to induce positive consumer responses. A survey revealed that
consumers regarded advertisements featuring celebrity endorsers as "attention-gaining,

entertaining. likeable and impactful" (O’Mahony and Meenaghan 1997/98).

Correspondingly, endorsements using famous figures have received special attention and
have been widely used by marketers. Approximately one quarter of US. commercials
were reported to feature celebrities (Shimp 2000).

This extensive use of celebrity endorsers ascribes to its potential benefits.
Celebrity endorsements are more influential relative to other types of endorsements in
several ways (Erdogan 1999). First, the use of celebrities draws more attention to the
advertisements they appear in and brands they are aligned with than those ads that do not
take advantage of an individual’s celebrity status (Stemthal, Phillips and Dholakia 1978).
Second, celebrity endorsers help advertisements break through the surrounding and
otherwise busy clutter (Erdogan, Baker and Tagg 2001). Third, with their distinctive
images, celebrities are used to create, enhance and change brand images (Debevec and
Iyer 1986; Erdogan 1999). Moreover, celebrities with worldwide popularity and
recognition can also help overcome cultural barriers in global marketing communications
(Erdogan 1999; Kaikati 1987).

More importantly, celebrity endorsements have been found to generate sales and
profits (Agrawal and Kamakura 1995, Gabor, Jeannye and Wienner 1997, Mathur,
Mathur and Rangan 1997). For example, Michael Jackson’s contract with Pepsi Co.
resulted in 8 percent increase in sales in 1984, although Michael Jackson turned out to be
a regretful choice for the company due to his alleged child molestation (Gabor, Jeannye
and Wienner 1997). Two studies using event study methodology showed that celebrity
endorsement contracts produced positive effects on stock returns and increased the
profitability of the client firms (Agrawal and Kamakura 1995, Mathur. Mathur and

Rangan 1997).

Despite these potential advantages, celebrity endorsements also involve some
risks. Although /marketers Spend a large sum of money in securing celebrities to be tied
with their brands, this investment is unrewarding when the celebrity endorsers become
involved in allegations of illicit or unethical behaviors. For example, the infamous
murder case involving OJ. Simpson concerned Hertz car rental, whose name was linked
with him through endorsement. Negative information about a celebrity endorser has been
found to influence consumer perceptions of both the celebrity and the endorsed brand
(Till and Shimp 1998). An advertiser’s reputation might suffer from the damage, thereby
tarnishing their brand image and sales due to this unfortunate selection of celebrities as
their endorserSfT‘J‘urthennore, all celebrity endorsements are not successful. Celebrities
who have proven to be highly effective as endorsers in the past can generate
disappointing results when tied to other products (Mieiak and Shanklin 1994). For
instance, when Bill Cosby, who has been a popular, successful endorser for Jello, Coke,

and Kodak, was hired to embrace E.F. Hutton and Texas Instruments, the outcomes were

not satisfactory.

’3‘ Another concern is that when some celebrities become associated with multiple <5 R I. (

products, overexposure might make the relationship between the celebrity endorser and
each of the products less distinctive. This would therefore reduce the positive influences
of the endorsements and cause the message to be less effective (Tripp, Jensen and
Carlson 1994). In addition, consumers might become more wary of the celebrity’s true
motive for endorsements. They might become skeptical about the messages the celebrity

carries in advertisements and believe that the celebrity does appear in the ads for the

compensation he or she receives, rather than sincerely believing in the merits of the
product or service.

These potential advantages and disadvantages Of employing celebrity endorsers
illustrate the imperative but difficult task of selecting the appropriate stars to promote
products. Professionals at advertising agencies and corporations have recognized the
importance of these decisions and utilize several criteria in evaluating and selecting
celebrities (Erdogan, Baker and Tagg 2001; Mieiak and Shanklin 1994).

In response to these concerns regarding celebrity selection, scholars also have
attempted to theorize the effects of the celebrity endorsements in advertising. Most
academic investigations of celebrity endorsement effects have been contextualized in the
realm of traditional source credibility effects (Ohanian 1991), which stemmed from Carl
Hovland and his associates in the early 19505 (Hovland, Janis and Kelley 1953).
Following the Source Credibility Model, the Source Attractiveness Model (McGuire
1985) also focused on source characteristics as the factors influencing the persuasiveness
of the message carried by the source. These models are categorized as source effects
models, which were designed to determine the conditions under which the message
sender or source is persuasive (Erdogan 1999). On the other hand, the Meaning Transfer
Model is unique in that it emphasized the symbolic meanings or images that celebrity
endorsers bring to the endorsement process (McCracken 1989). Another stream of
celebrity endorsement research has been labeled the “match-up hypothesis.” It examines
the fit or “match” between a celebrity and the product being endorsed (Kahle and Homer

1985; Kamins 1990). These models are discussed in detail in the following sections.

Source Credibility Model

Endorser effectiveness has been studied by mostly applying the Source Credibility
Model (Hovland, Janis and Kelley 1953, Hovland and Weiss 1951). This model identifies
expegise and trustworthiness as two dimensions of source credibility, which influences
the effectiveness of a message the source communicates. Expertise is defined as "the
extent to which a communicator is perceived to be a source of valid assertions" (Erdogan
1999, p. 298; Hovland, Janis and Kelley 1953) and the ability to provide accurate
information comes from the knowledge, experience, training or Skills a source possesses
(Erdogan 1999). Trustworthiness refers to the consumer's confidence or belief in the
communicator for carrying messages in an unbiased and honest manner (Ohanian 1991).
For instance, people often believe that their family and friends are more trustworthy than
sales representatives for offering product—related information.

The research on source credibility has shown that, in most situations, a highly
credible source is more effective than a less credible source (Stemthal, Phillips and
Dholakia 1978). Highly credible sources also have been found to generate more positive
attitudes toward the position advocated by the sources and to encourage more behavioral
changes than have less credible sources (Ohanian 1991). Celebrities in general are viewed
to be more credible and in turn exert greater effects on attitudes and purchase intent than non-
celebrities (Atkin and Block 1983; F reiden, 1984; Kamins 1989; Kamins and Gupta 1994;
Kamins, Brand, Hoeke and Moe 1989; Ohanian 1990; Petty and Cacciopo 1986). Similar to
the Source Credibility Model, the Source Attractiveness Model contends that the

effectiveness of a message depends upon the characteristics of the source (McGuire, 1968).

Source Attractiveness Model

 

Similarity, familiarity and likeability have been proposed as characteristics
affecting the message persuasiveness in the source attractiveness model (McGuire 1985).
Similarity refers to "a supposed resemblance between the source and the receiver of the
message" (Erdogan 1999, p.299). Familiarity is defined as "knowledge of the source
through exposure" and likeability as "affection for the source as a result of the source's
physical appearance and behavior" (Erdogan 1999, p. 299). Studies on similarity have
examined its effects based on a variety of source characteristics or attributes that include
age, gender, occupation, social class, and education (Feick and Higie 1992; Swartz 1984)
and have found that people expect communicators who are similar to them to have
similar attitudes and preferences (Eagly, Wood and Chaiken 1978; F eick and Higie 1992;
Stafford 1998). Because of their high profile and status, consumers may perceive
celebrities to be different from rather than similar to themselves. Their personality traits,
lifestyles, public recognition of their performances as well as physical appearances might
make big-name celebrities look more attractive to consumers regardless of their actual
appearances (Erdogan 1999).

Celebrities are Often chosen as endorsers due to their attractiveness coming from
both celebrity status and physical appearances (Erdogan 1999; Singer 1983). Research on
attractiveness has shown that attractive communicators in general have generated more
positive stereotypes. more belief changes and greater purchase intentions (Baker and
Churchill 1977; Debevec and Keman 1984; Petroshius and Crocker 1989). However,
some other studies demonstrated that attractiveness did not have much of a significant

effect on perceived credibility Of the source nor persuasion (Caballero, Lumpkin and

10

Madden 1989; Joseph 1982; Maddux and Rogers 1980) or its effects are dependent upon
message receivers‘ personal characteristics (DeBono and Hamish 1988). //
Source credibility and source attractiveness models have been extensively used to
examine celebrity endorsement effects. The premise of the source effect research is that
various characteristics of the sources influence the persuasiveness of a message and thus
the message acceptance by the audiences (Hovland, Janis and Kelley 1953, Hovland and
Weiss 1951)). Although they have proven useful in researching celebrity endorsement
effectiveness, these models have also received criticism for their limitations. It is argued
that Source Credibility and Attractiveness Models do not explain the persuasiveness of a
celebrity in relation to a particular product (McCracken 1989). Because celebrities in
general are perceived to be credible and attractive, they should be effective in generating
positive consumer responses to their recommendations for a variety of products and
brands. However, not all celebrity endorsements have been successful. Furthermore,
certain celebrities have been proven better endorsers than others for a particular product.
Thus, when celebrities become associated with particular products or brands, the
effectiveness of their endorsements may be determined by other factors as well as their
credibility and attractiveness. Celebrities who are widely recognized in the culture might
possess much more characteristics that are important in understanding the effectiveness
of their endorsements and the attractiveness and credibility might be part of the picture
that cannot explain completely the process of celebrity endorsements. “Meaning Transfer

Model” adds more insights into the celebrity endorsement effectiveness.

11

 

Meaning Transfer Model

According to the meaning transfer model, there are three stages in which certain
messages become integral to advertisements. Stage 1 indicates that a variety of meanings
become attached to a celebrity often through their appearances in mass media. Stage 2
involves the celebrity image/meanings being transferred to the product through the
endorsement. Stage 3, then demonstrates the Shift in meanings from the product to
consumers through purchase and consumption (McCracken 1989). Celebrities have been
elevated to the status because they possess distinctive qualities such as public
performances and distinct lifestyles. These famous figures are considered successful in
the sense that they have achieved the celebrity status in society and therefore enjoy the
privileges that this provides them. Their popularity motivates the public’s acceptance or
desire to pursue these eelebrities’ personalities, lifestyles, appearances, and behaviors to
consumers.

Advertising is a way of communicating and enhancing the meanings perceived to
reside in celebrities in the consumer world. When celebrities are hired as advertising
endorsers, marketers hope that consumers accept and consume the meanings or images
celebrity endorsers represent and link these meanings with their products. Consumers are
constantly moving symbolic properties out of consumer goods into their lives to construct
aspects of their self and world. This constructed self makes some celebrities exemplary
and inspirational figures to the consumer (Peter and Olson 1996). Celebrities represent
important cultural meanings that many consumers find personally relevant. And, when
consumers see the meanings and images a celebrity conveys desirable, they want to be

like the celebrity. Accordingly, by purchasing and using the product endorsed by the

celebrity, consumers can obtain some of those meanings and use them in constructing a
satisfying self-concept (McCracken 1989).

This conceptual model of meaning transfer in the celebrity endorsement process
was empirically tested in two studies conducted by Langmeyer and Walker (1991a, b).
For example, Madonna and Christie Brinkley were perceived to communicate different
meanings and images. These images possessed by the celebrity endorsers were found to
influence the meanings believed to be reflected by the products, when the celebrities were
combined with the products (e. g., bath towels and blue jeans) (Langmeyer and Walker
1991b). Before being linked with the celebrities, the products were perceived to have
only product category images. However, when being endorsed by the celebrities, the
products were viewed to represent some of the meanings that the celebrities originally
had. These findings supported the proposition that symbolic meanings that audiences
think are associated with celebrities moved to the products being endorsed.

One interesting point McCracken (1989) made was that celebrities are rich but
distinctive in the cultural meanings they convey to consumers and thus cannot be
persuasive for every product. The mix of personalities and qualities Offers different
meanings to consumers. Hence, only celebrities who deliver meanings that are
appropriate for the product they endorse Should be effective for the product. Likewise, he
pointed out that some celebrity endorsements are more effective than others due to a fine
match between the celebrity and the product. He argued that a “celebrity who best
represents the appropriate symbolic properties” of the product should be selected
(McCracken 1989, p.316). The relationship between endorser and product has been

studied under the name of "product match-up hypothesis."

l3

Match-Up Hypothesis

Literature on the match-up hypothesis attests that a good match-up between a
celebrity image and a product image is more effective than a bad fit between them. That
is, the messages conveyed by the celebrity and the product should be congruent for. the
advertising to be effective (Kahle and Homer 1985; Kamins 1990; Kamins and Gupta
1994; Lynch and Schuler 1994; Misra and Beatty 1990; Till and Busler 1998, 2000). This
line of research began with a focus on physical attractiveness (Kahle and Homer 1985;
Kamins 1990). The contention is that attractive celebrity endorsers are more effective
than unattractive spokespersons only when the product is related to attractiveness (e.g.,
perfume, cosmetics, and hair care products). Because the products are used to maintain or
enhance one's attractiveness, the physical attractiveness of the endorsers is viewed as
relevant to the product’s attributes and serves as a cue demonstrating the performances
and the quality of the product. The use of attractive endorsers, as a result, leads to the
enhancement of brand evaluation and purchase intent. Conversely, technology-related,
fiinctional products might not yield this type of interaction due to the irrelevance of
attractiveness to the product type. Although this premise seems apparent, empirical tests
have not provided strong support for this claim.

Originally initiated by Kahle and Homer (1985), four conditions were created to
examine the celebrity attractiveness by product type interaction in the first study that
fully tested the match-up hypothesis (Kamins 1990). Specifically, an attractive celebrity
was paired with a product used to enhance one's attractiveness (Tom Selleck/luxury car),

the same attractive celebrity was paired with a product not used to enhance one's

attractiveness (Tom Selleck/home computer), a less attractive celebrity was paired with
an attractiveness-related product (Telley Savalas/luxury car), and the same unattractive
celebrity was paired with an attractiveness-unrelated product (Telley SavalaS/home
computer). The results supported the match-up hypothesis for Spokesperson credibility
and advertising attitudes, but not for advertiser believability, advertiser credibility,
spokesperson believability, brand attitude, nor purchase intention.

Results from other research regarding the match-up hypothesis are reviewed and
discussed in more detail in the following chapter. This stream of research demonstrates
that it is important to keep in mind that celebrity endorser effectiveness is not consistent
across all conditions. Rather, success of celebrity endorsements is dependent upon a good
match between a celebrity endorser and the product endorsed. 3/

Overall, prior literature suggests that celebrities are influential, but the
effectiveness of celebrity endorsements are influenced by several factors such as the
perceived credibility and attractiveness of the celebrity endorsers, celebrity/product
match, and overall meanings attached to the celebrity spokespersons. However, the
findings are inconclusive regarding the appropriate or right star. Selecting an appropriate
celebrity with regards to product characteristics is important and advertisers are interested
in the characteristics that make a certain celebrity a good fit with a particular product
(Erdogan, Baker and Tagg 2001). Despite the considerable amount of the research in this
area, little has been known about the key criteria for determining the right celebrity
endorser.

Celebrities might not be the same in the eyes ofconsumers. Through the evolution

of their individual careers. celebrities have built distinctive images from various fields.

These images are important when celebrities appear to endorse products in advertising
because their images become associated with the products they promote. Because these
endorsements involve an affiliation between a celebrity and a product, the celebrity and
product relationship becomes crucial. Correspondingly, questions regarding consumer
responses to various pairings of celebrities and products for endorsements arise.
Therefore, the purpose of the present study is to expand the current literature on celebrity
endorsements by exploring celebrity images and assessing consumer reactions to
different celebrity endorsements from an attributional perspective. Chapter 3 will review
relevant literature in further detail to develop a theoretical framework for the current

research to address these issues.

16

CIIAPTER 3

THEORETICAL FRAMEWORK

The purpose of this chapter is to integrate prior literature to develop a theoretical
framework for guiding current research. As discussed earlier, the effectiveness of
celebrity endorsers has received special academic and practical attention due to the
characteristics that distinguish them from other types of sources or endorsers. However,
research on the celebrity endorsements has yet provided a comprehensive picture of the
mechanism. Questions as to the influences of famous individuals within the category Of
celebrities also arise. Because differential impacts of various types of celebrities as
endorsers in relation to products deserve more empirical investigations, this study
explores what image components celebrities project to consumers and how these images
influence consumer responses to celebrity endorsements. Furthermore, this research seeks
to examine how perceived congruence between celebrity endorsers and products
influences credibility of the celebrity endorser, consumer attitudes towards the ad and the

brand, and intention to purchase the brand from an attributional perspective.

Celebrity Credibility
Ohanian (1990) synthesized previous literature on source effects and proposed a
scale for celebrity credibility by integrating the source credibility and source

attractiveness models. Celebrities are one type of source among many, but their effects

might be unique because by definition, they are already well known by consumers.
Through history, certain individuals have been elevated to the level of the celebrity status
due to their recognition and distinctive qualities. As public figures, this popularity makes
celebrities look attractive to consumers, and thereby translating to credibility and power.
Although extensive explanations for the inclusion of attractiveness were not provided,
Ohanian (1990) regarded attractiveness as “an important factor” in understanding
celebrity credibility and proposed attractiveness as one of the three dimensions of
perceived credibility of celebrities (expertise, trustworthiness, and attractiveness).
Although some research found a source’s attractiveness to enhance credibility perception
of the source (Joseph 1982), the inclusion of attractiveness as part of the celebrity
credibility model might illustrate the idea that celebrities are different from general types
Of sources whose credibility and attractiveness have been studied in two separate models:
source credibility model and source attractiveness model.

The effectiveness of the messages that celebrities carry has been examined mainly
through the credibility that celebrities bring to the message. However, the effects Of
celebrities as endorsers in advertising might go beyond credibility. The images or
meanings that celebrities communicate to consumers are important in understanding a
more complete picture of the celebrity endorsement effectiveness (McCracken 1989).
Because celebrities possess and convey idiosyncratic combinations of images or
meanings, these differences might result in different consumer reactions to their
endorsements in advertising. Although little is known about the effects of celebrity
images in the endorsement process. some research has explored this aspect of celebrity

influences.

Celebrity Images and Influences

Celebrities are individuals who are symbolic icons popular in the culture and
transfer their symbolic meaning to the products they endorse in advertising (McCracken
1989). Consumers know what the images or meanings of individual celebrities reflect and
thus accept the associations between celebrities and products through appropriate
endorsements. In turn, consumers take in these images associated with celebrities and
products by purchasing and using the products. Although individual celebrities have been
known to possess and communicate distinctive bundles of various meanings or images,
the kinds of images or meanings that celebrities carry are unknown. In addition, although
celebrity images were found to influence product images when the celebrity were
associated with the products (Langmeyer and Walker 1991a, b), the effects of these
images on consumer responses to the endorsements are yet unknown.

Relevant to celebrity images and influences, marketing and communications
literature has proposed several ideas. One concept of a consumer’s relationship to
celebrities is a reference group. which is defined to be an actual or imaginary individual
or group conceived of having significant relevance upon an individual’s evaluations,
aspirations, or behavior (Park and Lessi g 1977). Cocanougher and Bruce (1971) suggest
that the use of “distant others” (e. g., movie stars, athletes, and the like) in all forms of
advertising is a clear illustration Of the basic assumption in marketing that there is a
relationship between an individual’s attraction to a socially distant reference group or
individual and the amount of influence the group or individual exerts. Indeed, any
marketing communication that urges consumers to act in terms of the behavior of a

referent individual or group with whom the consumer has no regular personal interaction

would appear to be based on the assumption that such distant referents can influence the
consumer’s product aspirations. Yet, despite the prevailing acceptance of such an
influence, little empirical support has been generated, because most previous reference
group research has concentrated on small face-to-face groups where interaction is on a
regular, routine basis, or where the situation has been experimentally induced
(Cocanougher and Bruce 1971).

In addition, Kelman's (1961) social influence theory has implications useful for
understanding the celebrity endorsement process (Erdogan 1999; Friedman and Friedman
1979). Among the three proposed processes of social influence (compliance,
identification, and internalization) that result in an individual adopting the attitude
advocated by the communicator, identification has been most employed to understand the
source effects of celebrity endorsers in advertising. This identification occurs when
individuals conform to the attitude or behavior advocated by another person because
these individuals derive satisfaction from their belief that they are like that person. This
applies to a situation wherein individuals emulate the attitude or behavior of another
person or group simply because they aspire to be like that person or group.

Another promising, but largely ignored, perspective on the relationship between
consumers and celebrities is the nature of their “parasocial” interaction. Audiences often
create an imaginary relationship with famous figures due to their repetitive appearances
in mass media (Alperstein 1991; Basil 1996). Defined as “intimacy at a distance” by
Horton and Wohl (1956), individuals tend to have the illusion of an interpersonal
relationship with celebrities. This is especially true of television figures because

television provides the illusion that these figures are talking directly to them — even

20

though there is no direct contact between them. This parasocial relationship appeared to
connect consumers to celebrities who endorse products in television commercials
(Caughey 1984). Since celebrities are well known to consumers due to their frequent
appearances in the media, consumers already have prior knowledge and attitudes toward
the celebrity spokespersons even before they appear in advertising. The images
consumers have had in their mind can therefore influence their relationship with the
celebrity endorsers and responses to the endorsements and the products being endorsed
(Alperstein 1991).

Consumers also have a self-esteem motivation to act in a way that establishes and
enhances their ideal self (Sirgy 1982). Celebrities might act as inspirational figures who
activate some consumers’ ideal self. Consumers may even develop an imaginary intimacy
with these celebrities. Accordingly, in an attempt to achieve their ideal images or
continue their relationship with celebrities, consumers tend to conform to the attitudes
and behaviors endorsed by celebrities. However, this acceptance of celebrities’ product
endorsement depends on consumers’ perceived images of the celebrities. Favorable
responses to celebrity endorsements might occur only when consumers perceive the
images or meanings the celebrities represent and convey as desirable.

Despite the increased importance of images in understanding celebrity
endorsement effectiveness, research on celebrity images is sparse. This might be due in
part to the lack of an established definition of “image” and the dimensions of its creation.
Although the term has been used widely in various fields, its meaning is vague. Image is
defined as “the concept of someone or something that is held or projected by the public”

by The American Heritage Dictionary of the English Language (1980, p. 352). This

21

seems to pertain to broad aspects of people and objects. However, as this paper is
concerned with, when limited to celebrities in particular, several characteristics might
emerge as consistently relevant to their images. Credibility, the most studied construct
with regard to celebrity endorsements, might be a component of celebrity images because
it is a more specific aspect of celebrity images. Or, these images can also influence
consumer perceptions of celebrity credibility. Hence, the first part of the present research
examines image components or dimensions of celebrities and their relationship to
celebrity credibility perceptions. Due to the exploratory nature of this study, a specific

hypothesis is not formed. Instead, research questions are provided.

RQ 1. What personal characteristics or dimensions comprise celebrity images?
RQ 2. How do the celebrity image dimensions relate to celebrity credibility

perceptions?

Although celebrity images are proposed to be an important factor in
understanding the celebrity endorsement process, the traditional approach to assessing
celebrity endorsement effectiveness still needs further investigations. Recent research on
celebrity endorsements has focused on the differential impacts of celebrity endorsements
with regards to products, but provided inconsistent findings. To provide a more complete
and accurate picture of the celebrity endorsement process, the relationship between

celebrity endorsers and products is fiirther examined.

22

Celebrity and Product Congruence

Consistency (or congruence) between celebrity endorsers and the products that
they promote has been proposed as an important factor influencing the effectiveness of
celebrity endorsements. Early studies found that the effects of celebrities as endorsers
varied with the product types being endorsed (Friedman and Friedman 1979). Certainly.
this inconsistent impact of celebrity endorsements is often witnessed in practice. Some
endorsements are more successful when an appropriate match between celebrity and
product is made. In fact, consumers were found to expect congruence between celebrity
endorsers and the product categories that they endorse (O’Mahony and Meenaghan
1997/98).

Furthermore, advertising practitioners have identified the match between celebrity
and product as an important factor to consider when selecting appropriate celebrity
endorsers (Erdogan, Baker and Tagg 2001; Mieiak and Shanklin 1994). With the
importance of the relationship between endorser and brand at hand, McCracken (1989)
suggests that a “celebrity who best represents the appropriate symbolic properties” of the
product be selected (p.316). This idea of fit between endorser and product has been
studied under the name of "match-up hypothesis."

Misra and Beatty (1990) and Kamins and Gupta (1994) emphasized the
importance of the perceived endorser/brand congruence at an overall level in testing
empirically the congruence effects. Misra and Beatty (1990) highlighted the "holistic"
view of the celebrity endorser endorser/brand match and suggested that for the

spokesperson/brand match to be congruent, "the highly relevant characteristics of the

23

Spokesperson are consistent with the highly relevant attributes of the brand" (Misra and
Beatty 1990, p. 161). Operationally, congruence or incongruence was determined by
similarity or dissimilarity between a characteristic of the celebrity and an attribute of the
endorsed brand. Results showed that the congruent match generated higher brand recall,
facilitated affect transfer from the endorser to the brand. and enhanced positive brand
affect.

Kamins and Gupta (1994) examined congruence between the overall image of the
endorser and that of the brand. They selected the congruent or incongruent pairings Of
endorser and brand based on the consumers’ assessment of how well the endorser and the
brand matched. Specifically, (1) Leonard Nimoy, a celebrity endorser and (2) an
accounting company president, a non-celebrity endorser were paired with either (1) a
higher congruence product (personal computer) or (2) a lower congruence product
(running shoes), which resulted in four spokesperson/product combinations. Findings
showed that the celebrity spokesperson was perceived to be more believable when
endorsing the more congruent product mersonal computer) than when advertising the
incongruent product (running shoes). These effects were not observed in the non-
celebrity spokesperson condition. In addition, the congruence effects were not found for
other dependent measures (i.e., ad attitudes. brand attitudes and purchase intention).

In an attempt to resolve these inconsistent findings, Till and Busler (1998, 2000)
have conducted two consecutive studies of the match-up hypothesis by comparing two
relevant factors: physical attractiveness and expertise. Consistent with their predictions,
expertise appeared to be a stronger match-up factor than attractiveness, suggesting that

expertise deserves more attention as a match-up basis.

24

Role of Expertise as a Match-Up Factor

The important role of expertise in matching a celebrity and a product is also
evidenced in findings from a study on consumer responses to celebrity endorsements. In
answers to Open-ended questions of product types that were deemed suitable for each of
the pre-determined four celebrities to promote particular products, and of the reasons
regarding compatibility, consumers listed the product categories in which the celebrities
were viewed to possess expertise, and images that were congruent with their career and
lifestyles portrayed in media (O'Mahony and Meenaghan 1997/98). For example, Jack
Charlton was considered to be an appropriate endorser for products associated with his
career in sports and his widely known fishing hobby because respondents felt that he
possessed expertise in these topics.

Expertise is defined as "the extent to which a communicator is perceived to be a
source of valid assertions" (Erdogan 1999, p. 298; Hovland, Janis and Kelley 1953) and
the expertise comes from the knowledge, experience, training or skills possessed by a
person (Erdogan 1999). That is, expert endorsers should have the ability to assess and
discuss the product qualities and thus provide accurate product information based on the

knowledge or experiences they have acquired in the past. Of particular interest is that

expertise is not evaluated by using objective standards but is instead determined by a {/c -. , r? ‘

consumer’s subjective perception. Erdogan (1999) emphasizes this subjective nature of

expertise and states, "all that matters is how the target audience perceives the endorser"

k.
(p. 298). __J

25

In early studies on celebrity endorsements. a celebrity was defined as "an
individual who is known to the public (i.e., actor, Sports figure, entertainer, etc.) for his or
her achievements in areas other than that of the product class endorsed" (Friedman and
Friedman 1979, p. 63). In this definition, celebrities are not considered to possess
expertise. However, celebrities Often promote products that are related to their
professional areas in practice. Some popular examples are Tiger Woods for Nike Golf
and Michael Jordan for Nike Basketball. These are among the most persuasive
endorsements because these well-known sports celebrities bring instant credibility to the
brands through the close associations between their career in the sports and the products.
Consistently, expertise has been identified as a strong factor that influences source
credibility and purchase intent (Ohanian 1990, 1991).

Taken together, the perceived expertise is deemed important in determining the
overall perception of congruence between celebrity and product (Till and Busler 2000).
Although Kamins and Gupta (1994) contended that physical attractiveness or perceived
knowledge of an endorser is only a component of the "more general notion of congruence
between spokesperson image and product" (p. 572), perceived expertise in the endorsed
product is still a strong, obvious basis on which the congruence perception is formed

(Speck, Schumann and Thompson 1988).

C elebritLand Product Congruence Based on Expertise
Given the promising role in celebrity/product congruence, perceived expertise of
celebrities deserves closer scrutiny. In the Till and Busler (2000) study. a fictitious

endorser, Ted Franklin, described either as a "US. Olympic Track and Field athlete" or

as a "stage and screen actor" was paired with either energy bars (expertise-related
product) or candy bars (expertise-unrelated product). Consistent with their expectations,
the results showed endorser by product type interaction for brand evaluations (although it
was not found for purchase intent). That is, the athlete induced significantly higher brand
attitudes than the actor when endorsing energy bars. However, this interaction and
relationship was not found for candy bars. The operational definition of the expertise-
based congruence raises a question as to whether different degrees or types of expertise-
based congruence exist. Specifically, it is questioned if an athlete is really an "expert" of
energy bars compared to, for example, a nutritionist who has the knowledge to
scientifically analyze and prove the product quality. Instead, an athlete might make a
better endorser for tennis racquets or shoes due to his/her expertise in the sport.

Real-world examples of celebrity endorsements reflect this idea of different
degrees of relatedness between celebrity expertise and products. They range from close
ties between celebrity professional achievements and brands (e.g., Kobe Bryant for
Adidas Shoes, and Davis Love III for Titleist golf clubs) to no obvious relationship
between celebrity and product (e.g., Mike Piazza and John McEnroe for 10-10-220,
Shacquille O’Neal for Burger King, Michael Jordan for Hanes underwear and Rayovac
battery, Charles Barkley for Coors beer, and Pete Sampras for Dannon yogurt).

Some other examples might reflect moderate congruence somewhere in between
the close match and mismatch in terms of the perceived celebrity expertise. Consider
Halle Berry for Revlon and Michael Jordan for Gatorade. Halle Berry, the academy
award winning actress, might be perceived to have a lot of experience with beauty

products and the product might be understood to be a part of her beauty formula. The

27

basketball athlete, Michael Jordan might be viewed to know how to quench thirst in order
to physically perform well in his sport. Because the products they endorse are related to
the areas in which they established their celebrity status, they might be regarded as being
competent to discuss those products based on their accumulated experiences in the fields
— even though they do not have the professional Skills or scientific knowledge to evaluate
those products.

These observations might call for more in-depth investigations of consumer
perception of expertise and congruence. Past research has assessed the difference
between extreme cases of congruence (i.e., perfect match and mismatch) and ignored the
existence of moderate degrees of congruence evidenced in practice. Hence, the current
research will examine how different degrees of expertise or experience-related
congruence are perceived and subsequently influence relevant endorsement outcomes. A
schematic perspective on the endorser and brand congruence effects also has provided

useful insights into the process.

Schematic Approach to Understanding Congruence Effects

Schema theory is a useful framework for understanding why the presence or
absence of congruence between celebrity endorser and brand results in differences in
consumer responses to the spokespersons, ads, and brands (Lynch and Schuler 1994). A
schema is a cognitive structure that represents an Object or domain including a person,
event, place, and so on (Taylor and Crocker 1981). This organization of knowledge

regarding the target is developed through experiences over time and influences the

28

information-processing process including encoding. comprehension. retention, and
retrieval of information.

From this schematic perspective, consumers have well-developed schemas for
well-known celebrities. The schemas may have a set Of attributes or cognitive
representations that describe the celebrities’ personalities, abilities, lifestyles, and other
relevant characteristics. When a celebrity endorses a brand, a schema for the celebrity
and another for the product category or the brand would be activated in the consumer’s
mind. These characteristics and image of that celebrity may then be compared to the
attributes of the brand for congruence or fit assessment (Lynch and Schuler 1994; Misra
and Beatty 1990). Match—up or congruence between endorser and brand implies
consistency between “the highly relevant characteristics of the Spokesperson and the
highly relevant attributes of the brand” (Misra and Beatty 1990). This suggests the
importance of salience or relevance of the characteristics or dimensions on which the
overall congruence perception is based. Depending on the product types, different
characteristics of the endorsers may be deemed more salient and relevant as part of the
schema.

These characteristics of celebrity endorsers related to product attributes might be
interpreted as expertise. In the previous example of an athlete endorsing an energy bar,
the images of athletes and the product category might belong together (Lynch and
Schuler 1994; Misra and Beatty 1990; Till and Busler 2000). Sports figures often convey
energetic, physically fit, and dynamic images, and in this case these images are closely
related to or are consistent with attributes Of the product category, namely energy bars.

The strong and physically active image of sports stars might prove the qualities Of the

29

product. Furthermore. consumers might perceive that the athletes are competent to
discuss the product. In fact, a masculine image of an endorser was found to positively
influence his perceived expertise for a fitness product by adding a belief "Spokesperson S
knows about Brand X" to the preexisting schema of the endorser (Lynch and Schuler
1994)

Taken together, congruence consistently has been found to have positive effects
on Spokesperson believability and ad and brand evaluations (Kamins 1990; Kamins and
Gupta 1994; Misra and Beatty 1990; Till and Busler 2000). Yet, the mechanism of how
the celebrity and product congruence works is unknown. Therefore, the current study is

to investigate why these congruence effects take place from an attributional perspective.

Congruence Effects and the Discounting Principle

Mowen (1979) suggests that endorser credibility perceptions "may be a function
of the consistency or congruence of the image or knowledge base of the Spokesperson in
relation to the product image" (Kamins and Gupta 1994). However, little has been known
about why congruence influences the spokesperson credibility perceptions. Attribution
theory is a useful theoretical approach in understanding this effect Of match-up or
congruence on credibility and subsequent endorsement outcomes.

Attribution theory refers to a set of theories that attempt to explain how
individuals make casual inferences. Consumers infer causes of events, actions of others,
and even their own behaviors (Bem 1972). Consumers’ perceptions of causal
relationships regarding product performance and their own consumption have been

central to consumer behavior research, however one of the other areas attribution

30

research has explored is a spokesperson’s endorsement ofa product (F olkes 1988; cf.
Mowen and Brown 1981, Moore, Mowen and Reardon 1994). From the attributional
perspective, consumers make inferences about reasons for an endorser’s recommendation
9 for a particular brand; why did the celebrity agree to associate him- or herself with the

product? The discounting principle in particular has been recognized among several
attribution theories as being an appropriate consideration in analysis of this topic

\l (Sparkman 1982). According to the discounting principle, “the role of a given cause in
producing a given effect is discounted if other plausible causes are also present” (Kelley
1972, p. 8; Sparkman 1982, p. 277). Additionally, people tend to minimize the effects of
an attribution for a behavior if an alternative reason that accounts for the same action
exists (Folks 1988).

For the celebrity’s endorsements in advertising, two rivalry causes have been c
proposed in previous literature: the product and the money that the endorser is paid M, I
(Mowen and Brown 1981; Sparkman 1982). Consumers may attribute the celebrity’s
willingness to be associated with the product either to the inherent product qualities or to
the money he or she obtains in return for the endorsement. In addition, although not
examined in prior studies, another cause for celebrity’s endorsements is often witnessed
in practice. Celebrities are conscious about their images to the public and strive to
maintain (and even enhance) their images. This effort involves their association to
products with images compatible with theirs. As a result, appearances in television
commercials might help to perpetuate their popularity. Thus, some might believe that bi g—

name stars are motivated to endorse a product because of their true affinity, or for their

personal image management, not by the endorsement fee (Atkin and Block 1983; Kamen,

31

Azhari and Kragh 1975). However, large sums of money involved in contracts between
celebrities and companies for endorsements have also been widely recognized.

When consumers infer that monetary incentives are the motivating factor that
accounts for a celebrity’s favorable comments towards the product, they subsequently
discount the effects of the endorsement of the celebrity (Mowen 1979; F olkes 1988).
Accordingly, the attributions to monetary gains were found to cause a decrease in the
perceived credibility of the celebrity endorser (Sparkman 1982). These findings suggest
that the use of uncompensated spokespeople (if this is well publicized) Should be more
effective. Reducing monetary attributions, however, might be more realistic because
uncompensated celebrity endorsements are not common practice. As a result, the
question that arises is: What causes the creation of inferences regarding monetary gains
of celebrity endorsements?

The perceived degree of congruence between endorser and brand might serve as a
cue that determines the inferences about why the celebrity promotes the brand. From the
schematic perspective. incongruity has been found to lead to more elaborate processing
whereby attitudes are formed based on more detailed analyses of the relevant attributes
(Fiske and Pavelchak 1986; Meyers-1-,evy and T ybout 1989). For celebrity endorsements,
the inconsistency between endorser and brand might trigger or stimulate more careful
thoughts about the endorser’s motive for the endorsement.

Because a logical connection between the celebrity and the endorser is absent
(e.g., the celebrity is perceived not to have the expertise to evaluate the product) in the
incongruity condition, people would likely attribute the cause of the endorsement to

monetary gains instead of product characteristics. When a natural fit between endorser

and brand is Observed, consumers might not put more thoughts in making inferences
about the endorser’s motive. Although it is common knowledge that celebrities are
compensated with significant sums of money for their endorsements, it might not become
activated in the consumer’s mind as an alternative cause for the endorsement under the
congruent condition. As a result, more attributions of product performance will be made
than attributions towards monetary incentives. Likewise, consumers will make more
attributions to images under congruence condition than in incongruity condition.
Perceived consistency between celebrity endorser and product might drive more image-
oriented attributions between them.

The attributions Of compensation for the endorsement appeared to decline the
perceived credibility of the spokesperson (Sparkman 1982). In contrast, product
attributions will enhance the credible image of the endorser. Although the effects of
image attributions on credibility has not been suggested in the previous literature, the
aforementioned effects of endorser/brand congruence on the attributions and the
subsequent credibility perception of the endorser are summarized in the following

hypotheses.

Hla. Celebrity and brand congruence will result inlessattribution to the monetary
gains for the endorsement.

Hlb. Celebrity and brand congruence will result in more attribution to the product
for the endorsement.

ch. Celebrity and brand congruence will result in more attribution to the
celebrity image enhancement for the endorsement.

Hld. Celebrity and brand congruence will lead to a more credible perception of

the celebrity endorser.

33

Misra and Beatty (1990) tested the effects of celebrity Spokesperson and brand
congruence on affect by creating three conditions: (1) congruent, (2) incongruent, and (3)
irrelevant matches. The results showed that transfer of affect from the endorser to the
brand occurred only when the celebrity spokesperson was congruent with the brand and
the brand affect was significantly more positive in the congruent condition than in the
incongruent and irrelevant conditions. Consistent with the findings, Fiske (1982)
proposed that schematic match—ups determine affect transfer. That is, if an object is
congruent with an existing, relevant schema, it will receive the affect associated with that
schema. If the match does not exist, the transfer of affect may not take place.

Brand associations are often influenced when the brand becomes linked with a
celebrity through the endorsement or an event through sponsorship activities (Dean 1999;
Gwinner and Eaton 1999; Keller 1993). In the hope that the positive meanings and
images that celebrities possess are transferred to their brands through association
(McCracken 1989), advertisers affiliate their brands to celebrities. Because celebrities, in
general, are attached with meanings and images that are popular in current culture,
celebrity/product congruence would produce more favorable attitudes toward the ad and
the brand.

Furthermore, credible communicators have been found to be effective in
generating positive ad/brand evaluations and purchase intent (Ohanian 1991). Positive
credibility perceptions of the endorser by the consumer are expected to induce favorable
responses toward the ad/brand, which is derived from fewer attributions of monetary
incentives. The foregoing discussion leads to the following hypotheses to test the

subsequent effects Of the congruence on the ad/brand evaluations.

34

H2a. Celebrity and brand congruence will lead to more positive attitudes toward
the ad.

H2b. Celebrity and brand congruence will lead to more positive attitudes toward
the endorsed brand.

H2c. Celebrity and brand congruence will lead to higher purchase intent for the

endorsed brand.

Facilitation Cues of Celebrity and Product Congruence

For some product types, it might be difficult to find a celebrity who possesses
strong expertise (i.e., food, beverages and high-tech products). Other types of endorsers
(e. g., typical consumers, anonymous models) can be used, but these endorsers cannot
impart the benefits of using celebrities. In the cases in which endorsers do not have
Special knowledge or professional relevance to the product, the advertisements might
need to "interpret" or explain the relationship between the endorser and the product
(Crimmins and Horn 1996). If this linkage makes sense to consumers, they might view
the endorser and brand relationship to be appropriate, congruent, and acceptable.

This might be an additional burden on advertisers because, although they should
create and communicate the meanings Of the connection between endorser and brand that
are not readily available from preexisting celebrity and brand schemas, this goal is not
necessarily executed. Furthermore, this communication should be performed effectively
with the time and/or space limits that advertisements deal with. If this additional
information in the advertisements does not easily enter in the consumer encoding process,
the effect of explanation of the endorser/product relationship might be minimal.

However, little research has examined the potential effects of this sort of strategies.

35

One possible way of associating a celebrity with a product might be to insert a
statement of the celebrity’s role or relationship to the product within the advertisement. A
study examined the effects of verbal cues in identifying the endorser’s role (expert,
neutral or typical user) in advertisements along with preexisting, professional
celebrity/product connection and processing time (Speck, Schumann and Thompson
1988). Contrary to their prediction, the results indicated no significant effects for these
relational cues on recall of product information. However, the statements that define the
celebrity endorser’s role in relationship to the product might influence other advertising
outcomes such as brand attitudes and purchase intent. If consumers View the facilitation
cues of an endorser’s role to be convincing and thus perceive the endorsement as
acceptable, their response to the advertisement and the brand might be enhanced. This
might be true especially for consumers who initially see incongruity between the
celebrity endorser and the product. A related question asks what kind of cue is considered
to be effective to perform this faciliatory effect.

The link between celebrity and product is sometimes weak or absent, however it
becomes more appropriate when the celebrity claims that he or she uses the product.
Because consumers often view testimonials of typical consumers, they may accept the
product use claim as reasonable. This statement might lend more weight to product
attributions by making a genuine affection for the product more compelling as the
celebrity's motive for the endorsement. Furthermore, it might subsequently enhance

ad/brand evaluations. The following hypotheses therefore address these concerns:

36

H3a. A cue that implies product use will result in less attribution to the monetary
gains for the endorsement.

H3b. The product use cue will result in more attribution to the product for the
endorsement.

H3c. The product use cue will result in more attribution to the image for the
endorsement.

H3d. AS a result, the product use cue will lead to a more credible perception of the

celebrity endorser.

H4a. The product use cue will lead to more positive attitudes toward the ad.

H4b. The product use cue will lead to more positive attitudes toward the endorsed
brand.

H4c. The product use cue will lead to higher purchase intent for the endorsed

brand.

Although they are discussed in the results section of this research, limited
literature on the relevant issue prevents the formation of particular hypotheses regarding
congruence by product use cue interactions. A general prediction is that celebrity and
product combinations with less obvious links might gain more with the cue (versus strong
preexisting celebrity/product relationship) by helping consumers make sense of and
hence accept the associations.

Chapter 4 will describe the methods that are used to develop a scale to measure
celebrity images (Study 1) and to test the proposed hypotheses regarding effects of

celebrity and brand congruence (Study 2).

37

CHAPTER 4

METHODS

The purpose of Chapter 4 is to explain the methods that are used in the present
research. AS described previously, there are two major Objectives of this study: (1) to
develop an instrument to measure celebrity images, and (2) to assess the effects of
relationship between celebrity endorser and product endorsed. Given these Objectives,

two studies were performed.

Study 1: Development of a Scale to Measure Celebrity Images
To develop a scale to measure celebrity images, two phases of the research were
undertaken. In the exploratory phase, the initial items were generated and the structure of
the scale was examined. In the confirmatory phase, the list of the items in the scale were

further refined. validated, and finalized.

Initial Scale Item Selection

 

In the initial phase of study 1, the relevant literature in the areas of psychology,
marketing, and advertising was reviewed to identify words or adjectives that can be used
to describe celebrity images. Because few words were found to be directly relevant for
measuring perceptions of celebrities, a large pool of adjectives describing personality

traits developed in clinical psychology and items measuring brand personality and self-

38

concept from marketing were used as the main basis for the initial development of the
celebrity’s image scale. In particular, the self-concept scale (Sirgy 1982), the scale for
measuring self-concepts, person concepts, and product concepts (Malhotra 1981), the
brand personality scale (Aaker 1997), and the celebrity appeal scale (Stever 1991) were
the main sources for the item generation. This process resulted in a pool of 70 candidate
items.

A panel of three expert judges provided feedback to reduce the universe of items
into a set that would be reasonable for consumer rating. Each judge, either a professor or
doctoral student in advertising, was asked to independently evaluate the 70 items and
eliminate items that are less applicable for measuring celebrity’s images. This stage left

58 semantic differential word pairs for the study.

Celebrity Selection Procedure

A pretest was used to identify professional athletes who had to date been widely
used as endorsers for various products. Thirty-five students rated the familiarity and
likeability of professional athletes from various sports fields, such as tennis, baseball,
basketball, golfing, and skating, on seven-point semantic differential scales. Only well-
known sports figures with familiarity scores greater than 6 were considered for use in this
study because participants in the study Should have prior knowledge of the celebrities to
accurately assess their images. As a result, Andre Agassi (mean = 6.61), Anna
Koumikova (mean = 6.49), Mike Tyson (mean = 6.77), and Tonya Harding (mean =
6.14) were selected to reflect a set of familiar, but diversely viewed celebrities (l =

unfamiliar / 7 = familiar).

39

To determine the appropriateness of the items to measure celebrity images, the
scale should be applicable to celebrities varying in their images. The four celebrities
represented both well-liked and disliked, male and female sports stars based upon the
likeability scores: Andre Agassi (mean = 6.23), Anna Koumikova (mean = 5.97), Mike
Tyson (mean = 2.77), and Tonya Harding (mean = 1.63) (l = not likeable / 7 = likeable).
This selection Of professional athletes was expected to embrace both positively and

negatively perceived celebrities.

The ExploratorLPhase

In the exploratory phase of this research, four different questionnaires were
developed, each of which contained two parts for one male and one female Sports figures.
The respondents were asked to evaluate respectively each of the two celebrities, given the
58 semantic differential items generated in the Item Selection section. Although four (two
males and two females) professional athletes were selected in the pretest, only two were
presented in each questionnaire to avoid response set bias, boredom, and hopefully obtain
good quality data. (A list of 58 items was believed to be quite long, and a likely cause of
respondents’ fatigue, so it was decided not to repeat the scale for all four celebrities.)
Furthermore, to control for order effects, the order in which the two celebrities were
presented in the questionnaires were counterbalanced, resulting in a total of eight
versions of the questionnaires.

A group Of 258 students who were enrolled in communication and advertising

classes completed the survey. The sample consisted of 84 males (32.6%) and 174 females

40

(67.4%). Their ages ranged from 18 to 31 years with an average age of22 years. Over

96% of the respondents were Single and 80% of them were Caucasian.

The Confirmatory Phase

In the confirmatory phase of the study, the scale consisting of 22 items developed
in the exploratory phase was tested using a second independent sample of subjects. Given
the second research question as to the relationship of celebrity images to celebrity
credibility, measures for celebrity credibility dimensions: attractiveness, trustworthiness,
and expertise were included in the questionnaire. 207 students initially competed the
survey. After eliminating incomplete surveys, a total of 198 male (101) and female (97)
undergraduates comprised the final sample. For this phase of the study, a different
celebrity, Pete Sampras, was used to support the generalizability of the scale. He was
selected as a professional athlete (tennis player), who is comparable to the celebrities
previously used for the first survey, with relatively high familiarity (mean = 6.1) and

likeability (mean = 6.0).

Study 2: Testing Hypotheses
Participants
A total of203 male (104) and female (99) undergraduates who were enrolled in
advertising, communications, and telecommunication classes at a major Midwestern
university participated in the main study. Their ages ranged from 18 to 32 years with an
average of 22 years. Over 96 percent of the participants were single and approximately

68 percent were Caucasian.

41

MD

The study employed a 3 (congruence) x 2 (product use cue) between-subjects
factorial design. The first independent variable, celebrity/product congruence presented
three levels (high, moderate, and low) with one sports figure paired with three different
products. The other independent variable, which was a verbal cue indicating that the
celebrity endorser uses the product being endorsed, had two conditions (presence or

absence of the cue).

Selection of Endorser and Products

Prior to the main experiment, two pretests were performed to select an endorser
and products. In the first pretest, 35 undergraduates rated familiarity, popularity,
attractiveness, and likeability of 30 male and female professional athletes. Among those
who appeared to be familiar to the subjects with familiarity scores greater than 6, Pete
Sampras was deemed appropriate. In addition, Sampras has not been involved in many
endorsements and does not promote any products currently unlike other well-known
sports figures on the list (e.g., Tiger Woods, Michael Jordan. Andre Agassi, and Anna
Koumikova).

In the next pretest, 22 students evaluated congruence of pairings of the
professional tennis player, Pete Sampras and a variety Of products including products
used in previous studies such as a personal computer, an energy bar, and a candy bar
(Kamins 1990; Till and Busler 2000). As a result, a tennis racquet, shampoo, and a coffee
maker appeared to be the appropriate products representing the intended levels of

congruence. A within-subjects ANOVA showed that the subjects’ perception of these

42

three pairings were significantly different (F912) = 30.03. p < .001). Three paired t-tests
also revealed that a tennis racquet (mean = 6.65), a shampoo (mean = 4.50). and a coffee
maker (mean = 3.19) resulted in significant differences in the participants’ congruence
perception Of Pete Sampras’ endorsements of these products (p < .005). Correspondingly,
Pete Sampras was featured to endorse a tennis racquet in the high congruence condition.
a shampoo in the moderate congruence condition, and a coffee maker in the low

congruence condition.

Stimulus Materials

Full-page, black and white. magazine-type advertisements were created for use as
the stimulus materials in this study (See Appendix E). These ads featured a headline read
“Pete Sampras, thirteen grand slam trophies. nothing but the best,” an image of Pete
Sampras as the celebrity endorser. and a small product photo of a tennis racquet, a
Shampoo, or a coffee maker. that corresponded to the experimental conditions. A
fictitious brand name, “Aerius” was used for all three of the products to prevent any
possible confounding effects of pre-existing perceptions Of well-known brands (Till and
Busler 2000).

Statements Of product use were also developed: “When I use Aerius, I make the
best shots. A winner every time.” for the tennis racquet; “When I use Aerius, I’m the best
I can be. A winner every time.” for the Shampoo; and “When I use Aerius, I make the
best coffee. A winner every time.” for the coffee maker. In addition, Pete Sampras’
Signature followed the personal statement. Two versions of the ad for each product were

developed: one with the product use statement and the other without it. This resulted in a

43

total of Six advertisements that were professionally created to represent the experimental
conditions. The executional formats of these ads remained identical across conditions

except for the manipulations.

Procedure

The experimental sessions were administered in large groups of twenty to sixty
students. Subjects were randomly given one of the Six booklets containing a front page,
the experimental ad, and a questionnaire. Participants were told that the purpose of the
study was to develop an advertising campaign for a new product that would be introduced
to the market in the near future. After viewing the ad, participants filled out the
questionnaire. Upon completing the experimental sessions, subjects were debriefed and

then dismissed.

Dependent Measures

 

Attributions of Celebrity Endorsers ’ Motives

Literature on celebrity endorsements suggested two plausible causes for a
celebrity’s endorsement of a product: (1) the celebrity’s belief in the product qualities
and/or genuine affection for the product and (2) the compensation the celebrity receives
in return for the endorsement. However, as discussed in Chapter 3, there might be another
attribution that consumers would make when they are exposed to celebrity endorsements.
Big-name stars are also concerned about their images to the public and strive to maintain
and enhance positive images. Associations with products that do not fit with their images

might hamper celebrities’ image management. Correspondingly, consumers might

44

attribute a celebrity's endorsement of a certain product to their image maintenance or
development. To further explore this issue. 50 undergraduates were asked for reasons
why celebrities appear in advertising to endorse products. Open-ended questions asked
the subjects to provide as many reasons as they could think of. Consistent with what was
expected, the results Showed that money, product qualities, and image management were
the most likely attributions. The next most frequently mentioned reasons included
receiving free products in return for the endorsements and exposure to the audiences
through appearances in mass media. Hence, measures for these attributions were
developed and included in the final questionnaires.

Product, money, and image attributions were measured on a seven-point, bipolar,
likelihood scale, from “extremely likely” to “extremely unlikely.” Subjects were asked to
indicate their agreement with each of the statements. Three statements were inteneded to
measure product attributions: (l) “The endorser agreed to do the ad because of the good
characteristics of the brand” (Sparkman and Locander 1980). (2) “The endorser appears
in the ad because he likes the brand,” and (3) “The endorser does the advertisement
because he really believes in the brand” (Mowen and Brown 1981). To measure
attribution to money, three items were developed: (1) “The endorser does the ad solely

,

for the money, ’ (2) “The endorser does the ad only because he is paid to do it,” and (3)
“The endorser appears in the ad only to receive the compensation” (Mowen and Brown
1981).

Although the product and money attribution measures were adapted from

previous research, items for measuring attribution to celebrity’s image management were

not Offered in prior literature. The following items were created for assessing attributions

45

to image: ( l) “The endorser becomes associated with the brand to improve his image,”
(2) “The endorser promotes the brand because it is consistent with his image,” and (3)
“The endorser appears in the ad to enhance his public image.” Additionally, three more
items were included based on the pretest: (l) “The endorser does the ad to get free
products,” (2) “The endorser appears in the ad to get exposure,” and (3) “The endorser
promotes the brand to receive attention.”

Because some of the items for measuring several attributions found in the pretest
were created for this study, the structure of the scale has not been explored. Thus, a factor
analysis was initially performed and the results guided subsequent analyses. Please refer
to the Results section for the factor analysis results and final sets of the items measuring

each attribution with their reliabilities (See Table 6).

Endorser Credibility
Celebrity endorser expertise, trustworthiness, and attractiveness were measured

on a seven-point, semantic differential scale (Ohanian 1990). Specifically, five items with

99 66

anchors of “expert/not an expert, experienced/inexperienced,”

99 6‘

“knowledgeable/unknowledgeable, qualified/unqualified,” and “skilled/unskilled” were

used to measure expertise (alpha = .81); five items with endpoints of “Sincere/insincere,”

39 66

“dependable/undependab1e,” “honest/dishonest, reliable/unreliable,” and

“trustworthy/untrustworthy” assessed trustworthiness (alpha = .81); and five items

53 56

anchored by “attractive/unattractive,” “classy/not classy, beautiful/ugly,”

“elegant/plain,” and “sexy/not sexy” measured attractiveness (alpha = .92). In addition,

46

their familiarity and likeability were measured on Single—item, seven-point. semantic

differential scales (“familiar/unfamiliar” and “likeable/dis]ikeable”).

Attitude Toward the Ad and Attitude Toward the Brand

Attitude toward the ad were measured on a seven-point, five-item, semantic
differential scale anchored by “good/bad,” “favorable/unfavorable,” “like/dislike,”
“interesting/boring,” and “pleasant/unpleasant” (MacKenzie and Lutz 1989). These items
were found to be internally consistent (alpha = .96). Subjects indicated their attitude
toward the brand on a seven-point, four-item, semantic differential scale anchored with

99 6‘

“favorable/favorable, good/bad,” “like/dislike,” and “pleasant/unpleasant” (MacKenzie

and Lutz 1989). The scale appeared to be reliable (alpha = .95)

Purchase Intent

Subjects were asked “How likely is it that you would consider Aerius, the next
time you purchase a tennis racquet (a shampoo or a coffee maker)?” and rated their
agreement with the statements and “The next time I consider purchasing a tennis racquet
(a shampoo or a coffee maker), I will inquire about Aerius” and “The next time I
purchase a tennis racquet (a Shampoo or a coffee maker), I will buy Aerius.” These items
are seven-point bipolar scales anchored by “extremely unlikely/extremely likely” and

“strongly disagree/strongly agree.” These items were found to be reliable (alpha = .92).

47

Congruence and Product use

Expertise-related or professional connection-based congruence was assessed on
the four-item, seven-point “extremely likely/extremely unlikely” scale. The items were:
(1) “The endorser is able to assess the qualities of products (tennis racquets, shampoo, or
coffee maker) in general,” (2) “The endorser knows a lot about products in general,” (3)
“Products are related to endorser’s performance or activities,” and (4) "The endorser is
competent to discuss products in general based on his professional experiences” (alpha =
.90). In addition, because it is also important to assess consumers’ beliefs in the celebrity
endorser’s use of the products in general, they were measured on a two—item, seven-point
“extremely likely/extremely unlikely” scale that consisted of (1) “The endorser actually
uses a product (tennis racquet, Shampoo, or coffee maker),” and (2) “The endorser has
used products” (alpha = .69).

Overall congruence was indicated on seven-point, semantic differential scales

9’ ‘6

using the following items: “congruent/incongruent, compatible/not compatible,”

“consistem/inconsistent,” “good fit/bad fit,” “good match/ bad match,”

’9 ‘6

"appropriate/inappropriate, effective/ineffective,” and “relevant/irrelevant” (Kamins
and Gupta 1994; Till and Busler 2000). The scale was found to be reliable (alpha = .97).
For manipulation check on the product use cue, two questions asked participants to
indicate the likelihood of the celebrity endorser’s actual use of the brand he endorses on a
seven-point “extremely likely/extremely unlikely” scale. The items were (1) “The

endorser actually uses the brand” and (2) “The endorser has used the brand" (alpha =

.68). It is expected to influence consumer judgment of the celebrity endorser’s actual use

48

of the brand whether consumers are exposed to a statement that the celebrity actually uses
the brand he endorses.
Additionally, subjects’ involvement with product categories was measured on a

seven-point, five-item, semantic differential scale with endpoints of

99 ‘6 ,9 C"

“unimportant/important, of no concern to me/of concern to me, irrelevant/relevant,”

“means nothing to me/means a lot to me,” “doesn’t matter/matters to me” (alpha = .96)

(Zaichkowsky 1985).

49

CHAPTER 5

RESULTS

The purpose of Chapter 4 is to present the results of this study. As outlined
previously, two studies have been conducted. This chapter is structured as follows: (1)
the results of Study 1 in which the scale to measure celebrity images is developed, (2) the

results of Study 2 in which the hypotheses were tested, and (3) the summary of the results.

Study 1: Scale Development
The purpose of the first study was to develop a scale for celebrity images. This
task was accomplished in two stages. In the exploratory phase, the initially generated
items were further reduced to a smaller number and the structure of the scale was

assessed. In the confirmatory phase, the items were finalized and the scale was refined.

Exploratory Factor Analysis for the Celebrity Image Scale

To assess the structure of the scale, all of the items in the questionnaire were
factor analyzed, using principle components analysis with varimax rotation. The initial
factor solution resulted in twelve factors with eigen-values greater than one. The twelve-
factor solution accounted for 66.6 percent of the total variance. In order to purify the list,

items with loadings of 0.3 or greater on more than one of the factors were eliminated

50

(Ohanian 1990). Items with low loadings below 0.6 were further removed. The reduced
list was factor analyzed again. This resulted in six factors with eigen-values greater than
one, while accounting for 65.9 percent of the variance. To further refine the list, the
procedure used at the earlier stage was repeated. Factor analysis with the reduced list was
rerun and this resulted in a rotated factor pattern consisting of four factors with eigen-
values of greater than one. This four-factor solution of 22 items explained 65.6 percent of
the total variance. All of the items hadhigh loadings over 0.6. Table 1 presents the results
of the final factor analysis.

The names determined to represent the types of concepts subsumed in each of the
four dimensions were Genuineness, Competence, Excitement, and Sociability. Factor 1,
labeled “Genuineness,” reflected human traits that are generally viewed to be favorable in
our culture. The items that loaded on this factor represented a human character who is
humble, modest, pleasant, and responsible. The second factor, “Competence” contained
items related to confidence, power, and enthusiasm. Factor III, “Excitement” reflected
masculine qualities such as ruggedness and dominance, and the fourth factor,
“Sociability” displayed an outgoing, extrovert personality. Due to its large number, the
set of items in the first factor was factor analyzed individually to identify facets that
perhaps reside in the factor. However, only one factor emerged, thereby reflecting one
dimension of celebrity images.

Reliabilities were calculated for each of the four dimensions using the 22-trait
scale. The resulting values were high except for the fourth factor: Genuineness = .94,
Competence = .82, Excitement = .83, and Sociability = .52. Of note is that the fourth

dimension has only two items. Although it was found to reflect a distinct dimension of

51

the celebrity images, it might need caution to use and interpret the factor to measure a

celebrity’s image. More items that describe the dimension might be added to the scale.

Confirmatory Factor Analysis for the Celebrity Image Scale

A confirmatory factor analysis was run on a second data set to evaluate whether
the new data confirmed the proposed structure of the celebrity image scale as determined
in the exploratory stage of the research. To verify the relationship between the proposed
scale items and the three celebrity image dimensions, a different data set was collected
and analyzed.

AMOS4 was utilized to perform a confirmatory factor analysis on the second data
set. Structural Equation Modeling methodology using AMOS4 or LISREL was used as
more rigorous statistical techniques to validate the proposed factor model indicated form
the results of the exploratory factor analysis of the first data set (Joreskog 1979). When
the four-factor model of 22 items was initially tested, the fit statistics indicated a
relatively poor fit with the data: the goodness-of-fit index (GFI) = .831, adjusted
goodness-of-fit index (AGFI) = .790, confirmatory fit index (CFI) = .781, root mean
square error of approximation (RMSEA) = .076, and chi-square = 435.82 with 203
degrees of freedom; p < .001. Regression coefficients of individual items were examined
and seven of them appeared to have loadings of 0.5 or smaller. This might be due to the
large number of items that loaded on the first factor because all of them, except for one
that loaded on the second factor, were items describing the first dimension.

Correspondingly, in an attempt to obtain a more parsimonious measurement

model, these items that loaded weakly on the factors were removed and the ensuing

52

factor model was reanalyzed. This resulted in a reasonable model fit: the goodness-Of-fit
index (GFI) = .91 I, adjusted goodness-Of-fit index (AGFI) = .872, confirmatory fit index
(CFI) = .919, root mean square error of approximation (RMSEA) = .060, and chi-square
= 142.28 with 83 degrees of freedom; p < .001. For a list Ofthe final set ofitems that
measure the four dimensions of celebrity images, see Table 2.

To test the relationship of the celebrity image scale to the celebrity credibility,
another measurement model including celebrity credibility constructs as well as the
celebrity image dimensions was constructed and analyzed. The results showed that all the
celebrity image factors except for Factor IV were significantly related to the celebrity
credibility dimensions: attractiveness, trustworthiness. and expertise (p < .05). Table 3
presents correlations between the latent constructs.

To further examine the relationships between these constructs, a structural
equation model with linkages from the image dimensions to the credibility components
was created and tested. Due to a lack of prior literature on this issue, specific
relationships were not proposed, however the celebrity image dimensions were expected
to influence the credibility constructs because the images reside in individual celebrities
and thus might not vary easily with situations. On the other hand, credibility perceptions
of a celebrity might change with the message he or she communicates. Also, consumers
might already have a celebrity’s images in mind that they bring to the situations wherein
they should make judgments about the celebrity’s trustworthiness and expertise.

The results revealed only one significant relationship from Factor I,
“Genuineness” to the trustworthiness dimension (standardized coefficient = 1.43, p

= .003). Many non-significant linkages between the latent variables led to a relatively

53

poor fit Ofthe model with the data: GFI = .839, AGFI = .806, CFI = .922, RMSEA = .050.
Chi-square = 575.40 with 386 degrees of freedom; p < .001. However, further attempts

were not made to improve the model because the focus Of this analysis was to assess the
relationships between the celebrity image and credibility dimensions, not to propose and

test a theory-based model.

Study 2: Hypothesis Testing
In Study 2, the proposed hypotheses regarding celebrity/brand congruence and
product use cue were tested. Prior to the experimental analyses, manipulation checks
were conducted to determine if the expertise-related congruence and product use cue

manipulations were successful.

Marimlation Checks

 

To ensure the selection of the celebrity endorser, subjects were asked if they
recognized the endorser. 82 % of the participants recognized Pete Sampras and provided
his correct name and profession. The mean Of the celebrity familiarity was 5.2 and that of
his likeability was 5.3 on the seven-point scale.

A manipulation check was performed between presence and absence of the
endorser’s statement of product use. Subjects were more likely to believe that the tennis
player actually used the brand he endorsed when they saw the personal statement of the
product use (mean = 4.82) than when they did not (mean = 4.40; I(l,201) = 2.10, p = .037).

Although the two conditions created statistically different perceptions of the celebrity’s

54

actual use Of the brand, the difference was small and the means were near the mid-point
of the seven-point scale.

To further examine the manipulation of the product use cue, its simple effects
were assessed at every level of the pairs. The endorser’s statement of the brand use
produced a significant difference in the belief that the endorser actually used the brand
only for the shampoo (H1497) = 8.10, p = .004), but not in the tennis racquet (F (1,197)
= .010, p > .1) and coffee maker conditions (Fmgn = 1.44, p > .1), although the means
are in the expected directions. These results imply that the statement of brand use is
effective only when consumers are not certain about the celebrity endorser’s actual use of
the brand. Perhaps when the celebrity is believed to possess strong or weak professional
ties with products, consumers might easily judge the possibility of the celebrity’s use of
the brand without relying on cues present in advertising. In contrast, in cases where the
relationship between a celebrity and a product endorsed is not obvious, consumers cannot
easily make judgment about the celebrity’s actual use of the product. Thus, their
assessment might be more influenced by the celebrity’s personal statement, which
implies that the celebrity actually uses the product.

Another manipulation check was undertaken to assess the effectiveness of the
different pairs of the celebrity and the products. A one-way ANOVA revealed that the
celebrity/product congruence manipulation in general worked (F (2,200, = 155.85, p < .001).
As expected, Pete Sampras was perceived to have strong expertise on tennis racquets
(mean = 6.33), a moderate degree of professional connection with Shampoos (mean =
3.40), and a weak level of professional knowledge on coffee makers (mean = 3.25).

Because the difference between the shampoo and coffee maker conditions seemed small,

55

a planned contrast analysis was performed to further examine it. NO significant difference
in the belief in the celebrity’s possession of expertise on or professional connection with
the product was found between when a shampoo was endorsed (mean = 3.40) and when a
coffee maker was endorsed (mean = 3.25; H1300) = .58, p > .1).

In addition, consumer perception of the celebrity’s product use was examined by
performing a univariate ANOVA. Results Showed a Significant difference in the
consumer evaluation of the likelihood that the celebrity used the product in general
(F (2,200, = 21.92, p < .001 ). Pete Sampras was most strongly viewed to use tennis racquets
(mean = 6.30), next most likely to use Shampoos (mean = 5.53), and least likely to use
coffee makers (mean = 4.69). Although the means were in the expected direction, they
were all above the mid-point of the scale. That is, subjects believed that the celebrity has
used all the three products although the belief was relatively different in strength
depending on the products. Tennis racquets are necessary for the celebrity’s profession
and both shampoos and coffee makers are product types that are likely to be used by any
person. Thus, consumers might assess the possibility of the celebrity’s use of all these
products as relatively high.

Differences in the belief in the celebrity’s product use experiences were firrther
examined with planned contrast analyses. The results showed that the pairing of the
celebrity and a tennis racquet produced a significantly stronger belief in his product use
(mean = 6.30) than the pairing with a shampoo (mean = 5.53; F0200) = 10.18, p = .002),
which in turn resulted in a stronger belief than the association between the celebrity and a
coffee maker (mean = 4.69; Fugoo) = 1 1.83, p = .001). Relevant to this manipulation,

overall congruence between celebrity and product was examined in the following section.

56

Congruence

A two-way ANOVA was performed on consumers’ perceived overall congruence
between the celebrity and the product. A significant main effect for product categories
was found (Fmgn = 87.74, p < .001) (See Tables 4 and 5). Tennis racquets were
perceived to be most congruent with the celebrity endorser (mean = 6.0), followed by
shampoos (mean = 3.74) and coffee makers (mean = 3.16). Because the results from the
omnibus ANOVA indicated the overall effects of the different pairs between the celebrity
and the products, a comparison was made to assess specifically whether the shampoo and
coffee maker induced different congruence perceptions of the celebrity/product match.
The results revealed a significant difference in the overall congruence between when the
celebrity endorsed a shampoo (mean = 3.74) and when he promoted a coffee maker
(mean = 3.16; F0300, = 6.40, p = 012). This means the condition of a shampoo induced a
significantly higher degree of congruence perception than the coffee maker condition.
Thus, although the perceived congruence between the celebrity and a coffee maker was
near the mid-point of the seven-point scale and not low enough, the levels of the
congruence were in the right order as planned.

The effect of the product use cue on the overall congruence perception appeared
to be marginally significant (F (1,197) = 3.46, p = .064), although the presence of the
product use statement produced a more congruence perception (mean = 4.47) than its
absence (mean = 4.13). This means that consumers viewed the pairing of the celebrity
and the product to be more congruent when they read a statement of personal product use

present within the ad than when they did not. Of note is that the product use cue resulted

57

in larger differences under the moderate and low congruent conditions. See Table 5 for
the means and standard deviations for the perceived congruence by condition. NO

significant interaction was found, however (p = .344).

Attributions

Items for measuring consumer attributions of a celebrity endorser’s motive(s)
were adapted from prior research or developed for this study. To assess the structure of
the scale, twelve items of attributions were factor analyzed, using principle components
analysis with varimax rotation. The analysis resulted in a rotated factor pattern that
consisted of three factors with eigen-values greater than one. The item of getting free
products did not load significantly on any factor and thus was dropped. The explained
variance for the three-factor solution of the remaining eleven items was 71.6%. As
expected, three factors represented attributions to money, product, and image. Items that
load on the three factors appear in Table 6. Factor 1, “Money Attribution” contained items
pertaining to a celebrity’s motive for the endorsement as the compensation he or she
receives. Factor 11, “Product Attribution” reflected a celebrity’s genuine affection for the
product endorsed or beliefs in the product qualities. Factor 111, “Image Attribution,”
which has not been offered in prior literature, displayed a celebrity’s motivation to
enhance his or her public image through his or her association with a product. Of note is
that one of the image items proposed, “The endorser promotes the brand because it is
consistent with his image” unexpectedly loaded on Factor 11, instead of Factor III.

Consistency of the brand image was perhaps regarded as a feature of the product.

58

Subjects might believe that a celebrity appears in the ad because he or she cares for the
product of which image is consistent with him or her.

Reliabilities were assessed for the proposed dimensions of celebrity endorser
motive attributions. The reliabilities for the three factors were strong. The money
attribution had an alpha coefficient of .878. Attribution to product was also reliable with
an alpha of .813. The image attribution also appeared to be reliable with an alpha of .844.
Each attribution index was determined by averaging scores on the corresponding items.

Univariate ANOVAs were run to examine the effects of the independent variables
on each type of attribution. As for the attribution to money, subjects most strongly
believed that the reason for Pete Sampras’ appearance in the ad was the compensation he
received when endorsing shampoos (mean = 5.67), followed by coffee makers (mean =
5.27) and tennis racquets (mean = 4.78). This provided partial support for Hypothesis 1a
(H1197, = 9.66, p < .001) in that the moderate level Of congruence resulted in the highest
score on the money attribution (See Table 7). To further examine small differences in the
attribution to the monetary gains among the conditions, planned contrast analyses were
performed. Two pair-wise comparisons showed that the three levels of congruence led to
significant differences in the money attribution. That is, the moderate level of congruence
induced higher attributions to money than the low congruence condition (F(1,200) = 3.92, p
= .049), which is followed by the high level of celebrity/product congruence (H1200) =
5.78, p = .017). However, main effects of the product use cue were, not significant,
disconfirming Hypothesis 3a (p > .1). NO significant interaction was found (p >.1).

Product categories exerted a similar influence on the attribution to the product.

Partially supporting Hypothesis 1b, the subjects attributed the celebrity’s endorsement to

59

his genuine affection for the product most likely when he was associated with a tennis
racquet (mean = 4.54), followed by a coffee maker (mean = 3.36) and a Shampoo (mean
= 3.00) (130,197): 35.45, p < .001) (See Table 8). A pair-wise comparison further
examined the difference between the low and moderate levels of congruence, which
appeared to be marginally significant (Fugoo) = 3.26, p = .072). Main effects for the
product use statement (1:11.197) = 2.91 _. p = .089) and interactions were also found to be
marginally significant (me) = 2.95. p = .055). Thus, Hypothesis 3b was weakly
supported.

AS for the image attributions, Significant main effects for both product categories
and statement of product use were found. Specifically, the subjects evaluated the image
enhancement as the reason for Pete Sampras’ appearance in the ad to be most likely when
endorsing a tennis racquet (mean = 4.66), followed by a shampoo (mean = 3.48) and a
coffee maker (mean = 3.22) (me, = 9.66, p < .001), providing support for Hypothesis
1c (See Table 9). The difference between the moderate and low levels of congruence was
further examined by a pair-wise comparison and results revealed that the difference was
not Significant (Pagan) = 1.71, p = .193). Hypothesis 3c was also confirmed; the subjects
attributed the celebrity’s endorsement more to his desire for image enhancement when
they were exposed to a cue Of product use present within the ad (mean = 3.94) than when
they were not (mean = 3.64) (Fmrm = 4.48, p = .036). The interaction was marginally
significant (F(2_|97) = 2.79. p = .064). Table 10 presents means and standard deviations for
the attributions.

Overall, the results showed a pattern ofconsumer assessments of the celebrity’s

motive for endorsing a product. Consumers’ attribution of the endorsement reason

60

differed with what products the celebrity was associated with. The subjects viewed the
celebrity’s reason for appearing in the ad more unfavorably (product and money
attributions) when the celebrity endorsed a moderately congruent product compared to
the products that are highly or lowly congruent with the celebrity endorser. The high
level of congruence led the subjects to attribute more to the positive reasons: product and
image, but less to the unfavorable reason: monetary gains. Although the order in which
the attribution scores appeared did not exactly coincide with the predications provided in
the hypotheses, the overall pattern was consistent with the expectations: the high
congruence resulted in more favorable consumer assessments of the celebrity endorser

motive compared to the moderate and low congruence conditions.

Endorser Credibility

Univariate ANOVAS assessed the effects on endorser credibility perceptions
(attractiveness, trustworthiness, and expertise). Different levels of congruence between
celebrity and product appeared to have significant effects on the perceived expertise and
trustworthiness of the celebrity endorser, but not on the attractiveness perception (p > .1;
See Table 11). In support of Hypothesis 1d, higher congruence led to more credible
perceptions of the celebrity endorser. Specifically, the subjects viewed Pete Sampras to
be most trustworthy when endorsing a tennis racquet (mean = 5.02), followed by a
shampoo (mean = 4.72) and a coffee maker (mean = 4.56) (Form = 4.21, p = .016) (See
Table 12). Although the ANOVA results confirmed the general prediction regarding the
influence of congruence on endorser credibility, two pair-wise comparisons showed that

the high and moderate levels of congruence were marginally Significant in inducing

61

credibility perception of the celebrity endorser (1:11.200) = 3.60. p = .059) and the
difference between the moderate and low levels of congruence was not significant (Foam)
= .93, p > .1 ).

Similar to trustworthiness, the celebrity was perceived to be more of an expert
when endorsing a tennis racquet (mean = 6.26), compared to when promoting a shampoo
(mean = 4.87) or a coffee maker (mean = 4.74) (F(2,197)= 31.36, p < .001; See Table 13).
The small difference between the shampoo and coffee maker conditions was further
examined by performing a pair-wise comparison and it was observed to be non-
significant (F (1,200, = .42, p > .01). Disconfirrning Hypothesis 3d, neither significant
effects for the statement of the product use nor interactions were observed (p > .1). Table
14 presents means and standard deviations for the endorser credibility perceptions.

Attractiveness, one dimension of celebrity credibility (Ohanian 1990) did not
appear to be influenced by the levels of celebrity and product congruence. This might not
be surprising since attractiveness was not included as a component Of source credibility
in the traditional source credibility model in which trustworthiness and expertise
consistently have been identified as the two core dimensions of source credibility. In
addition, attractiveness in general might be a characteristic of a person that might be
relatively constant across conditions, whereas trustworthiness and expertise is dependent
upon situations. What messages a person communicates might determine the perceived
trustworthiness and expertise, but not the attractiveness of the source. This is consistent
with prior research that found attractiveness scores remained stable whereas
trustworthiness and expertise ratings changed after the message or endorser manipulation

(Stemthal, Dholakia and Leavitt 1978; Till and Busler 2000).

To further examine the supposed relationship between the endorser motive
attributions and the endorser credibility, regression analyses assessed the relative
contribution Of money attribution. product attribution, and image attribution on endorser
credibility dimensions. Only attribution to image enhancement was found to be
significantly predictive of the celebrity endorser attractiveness (standard regression
coefficient = .149, p = .05; R = .217), whereas attribution to the celebrity’s liking for the
product endorsed appeared to be the only significant indicator of perceived
trustworthiness of the celebrity endorser (standardized regression coefficient = .244, p =
.005; R = .315). For the celebrity expertise perception, the standardized regression
coefficients indicated that both product attribution (.248, p = .003) and image attribution
were the significant indicators (.223, p = .002), with R = .404. Attribution to monetary
gains did not appear to be predictive of any dimension of the celebrity endorser

credibility (p > .1).

Attitude Toward the Ad

An ANOVA Showed a significant main effect for the product categories.
Hypothesis 2a was supported; subjects evaluated the ad more positively when the
celebrity endorsed a tennis racquet (mean = 4.61) compared to the cases in which the
celebrity promoted a shampoo (mean = 3.80) or a coffee maker (mean = 3.95) (F(2.l96) =
8.08, p < .001). A pair-wise comparison indicated no Significant difference in attitude
toward the ad between the moderate and low levels of congruence (Fuqu, = .45, p > .1).

NO significant effects for the product use cue were found (p > .1). Thus, no support was

63

provided for Hypothesis 4a. There was no significant interaction (p > .1). See Table 15

for the ANOVA results.

Attitude Toward the Brand

ANOVA results revealed significant main effects for both product categories and
product use cue. No significant interaction emerged, however (p > .1). Congruence
between celebrity and product enhanced attitudes toward the brand endorsed (Fmrn, =
8.11, p < .001). The attitude toward the brand was most favorable when the celebrity
endorsed a tennis racquet (mean = 5.00), followed by a coffee maker (mean = 4.50) and a
Shampoo (mean = 4.31). Thus, Hypothesis 2b was confirmed. The difference in the
attitude toward the brand between the moderate and low levels of congruence was not
significant when a pair-wise comparison was made (F (1,200) = 1.09, p > .1). On the other
hand, the statement of product use improved attitudes toward the brand, supporting
Hypothesis 4b. Its presence led to more positive attitudes toward the brand (mean = 4.75)
than its absence (mean = 4.04) (Fmon = 8.11, p = .046). See Table 16 for the ANOVA

results.

Purchase Intent

 

An ANOVA was performed on purchase intent and the results indicated a
Significant main effect only for the product categories. NO significant interaction was
found (p > .1). Subjects showed highest purchase intent when the celebrity endorsed a
tennis racquet (mean = 4.12) followed by a coffee maker (mean = 3.21) and a shampoo

(mean = 2.23) (Fmgn = 31.25, p < .001). Thus. Hypothesis 2c was partially supported

64

due to the reverse direction of the moderate and low levels of congruence. Two pair-wise
comparisons indicated significant differences among the three conditions (Fugue) = 62.88,
p < .001; H1200) = 16.86, p < .001). There was no significant effect for the product use
statement, disconfirming Hypothesis 4c. See Table 17 for the ANOVA results. Means
and standard deviations for ad attitudes, brand attitudes, and purchase intention appear in
Table 18.

Given a concern that consumers might present different levels of involvement
with the products used in the study, an ANCOVA controlling for product involvement
was performed on the purchase intention. The main effects of congruence between
celebrity and product were observed to be still Significant on the subjects’ intent to

purchase the product (Fmoé) = 31.08, p < .001).

Gender Differences

A series of t-tests determined whether they were gender differences although the
differences between males and females were not formally hypothesized. The results
showed that females Significantly viewed the celebrity to be more attractive (mean =
4.85) and more of an expert (mean = 5.56) than did males (mean = 4.33, p < .001; mean =
5.04, p = .009). Females also believed the celebrity endorsed the product because of his
need for image enhancement more strongly (mean = 4.56) and showed more favorable
attitude toward the brand (mean = 4.75) than males (mean = 3.87, p < .001; mean = 4.45,
p = .048). Given these differences between males and females, ANCOVAS controlling
for gender were performed on the perceived trustworthiness and expertise of the celebrity

endorser, the image attribution, and the attitude toward the brand. However, the

65

Significance of the main effects and interactions was unchanged. confirming the results

from the ANOVAS discussed in earlier sections.

Summary of Results

The objective of Study 1 was to develop a framework of celebrity image
dimensions and a scale to measure the dimensions. To identify the celebrity image
dimensions, a total of 258 subjects rated a subset of four professional athletes on 58 items
in the first phase. The results of an exploratory principal components factor analysis
suggest that there are four distinct dimensions that describe a celebrity’s image:
Genuineness, Competence, Excitement, and Sociability.

In the second phase, the results of a confirmatory factor analysis relying on 198
subjects, another celebrity, and 22 items provided additional information for the scale.
The initial confirmatory factor model was refined by excluding several items with low
loading, and the fit statistics indicated an acceptable fit of the modified model with the
data. Relationships of the celebrity image dimensions to the celebrity credibility
components were also investigated. The results showed that the image dimensions,
except for Factor IV, were significantly related to the credibility perceptions, and that
Factor I influenced the perceived trustworthiness of the celebrity.

In Study 2, the congruence between a celebrity and a product appeared to be the
most important characteristic of the endorsement message. As expected, high congruence
led the subjects to more attribution to product and image and less attribution to monetary
gains (See Figure 1). Subsequently, high congruence produced more credible perceptions

of the celebrity endorser, more favorable attitudes toward the ad and the brand, and

66

greater intent to purchase the product (See Figure 2). Of note is that celebrity and brand
congruence was not prOportionally effective with an increase in the degree in enhancing
the endorsement effectiveness. In particular, the moderately congruent pairing of the
celebrity and the product was not more effective than the least congruent pairing. Further
examinations of the differences between the two conditions revealed that the two levels
of congruence were not significantly different in creating different advertising outcomes,
except for the purchase intent in which the low congruence condition generated a higher
score than the moderate congruence condition. However, the high congruence condition
consistently appeared to have the most positive influences on all of the dependent
measures.

On the other hand, the celebrity’s statement of product use appeared to have
minimal effects. Although the scores were in the expected direction (the presence of the
product use statement led to more favorable endorsement outcomes than its absence), the
differences between the two conditions were often Observed to be non-significant.
However, the subjects attributed more to the image enhancement for the celebrity’s
motive for the endorsement, and evaluated the endorsed brand to be more favorable when
they saw the statement than when they did not.

The interactions between the celebrity/product congruence and the product use
statement in general were not found to be significant although they appeared to be
marginally significant only for the product and image attributions scores (See Figures 3
and 4). Of interest is that the effects of the presence of the product use cues were stronger
under the moderate congruence condition than the high and low congruence conditions.

The celebrity Spokesperson’s motive for endorsement was believed to be his liking for

67

the product and desire for the image enhancement more strongly when the product use
statement was present than when it was not at all levels of congruence. However. the
difference was larger at the moderate level of congruence than the high and low levels of

congruence.

68

CHAPTER 5

DISCUSSION

The purpose of this chapter is to discuss the findings Of this research with the
implications and limitations, as well as provide directions Of future research. This chapter
is structured as follows: (1) summary of research, (2) conclusions, and (3) limitations of

this research and directions for future work.

Summary of Research

A considerable number of celebrities featured in ads (Shimp 1997) and the
increased costs to secure celebrity spokespersons (Agrawal and Kamakura 1995)
certainly suggest that celebrity endorsements are an omnipresent and important feature of
modern marketing communication activities. Academic literature generally indicates
many benefits from using celebrities in advertising (Erdogan 1999; Kaikati 1987), and
thus mechanisms such as source characteristics models (attractiveness, expertise, and
trustworthiness) (Ohanian 1990), attribution theory (Mowen and Brown 1981; Tripp,
Jensen and Carlson 1994), schema theory (Lynch and Schuler 1994; Speck, Schumann
and Thompson 1988), meaning transfer (McCracken 1989), and the match-up hypothesis
(Kahle and Homer 1985; Kamins 1994; Till and Busler 2000), Elaboration Likelihood
Model (Petty, Cacioppo and Schumann 1983) have been proposed to explain the

effectiveness of celebrity endorsers. Despite the continued efforts over the years, little has

69

been known about what dctemrines the appropriateness ofcertain celebrities to endorse a
particular product.

The purpose of this research was to (I) explore dimensions of celebrity images
that might influence celebrity credibility perceptions and (2) investigate the roles of the
relationship between a celebrity spokesperson and a product endorsed in generating
endorsement effectiveness from an attributional perspective. Although celebrity
endorsers have been regarded as special types of sources who embody a variety of
cultural meanings (McCracken 1989; Ohanian 1990), research on the celebrity images iS
sparse. Therefore, this research has identified dimensions of the celebrity images and
developed a scale to measure the celebrity image dimensions. In addition, the
relationships of the celebrity image dimensions to the credibility perceptions of a
celebrity were investigated.

Research on the “match-up hypothesis” began with a focus on endorser
attractiveness, and subsequent research has proposed attractiveness, expertise, or overall
congruence as bases for a fit between an endorser and a product endorsed. The focus of
the study was a professional connection between a celebrity endorser and a product. The
relationship between the celebrity/product congruence and credibility perceptions of the
celebrity has been largely ignored. Thus, the present study has integrated two major
streams of the celebrity endorsement research in that it examined the role of
celebrity/product congruence in creating the subsequent credibility perception of the
celebrity endorser and advertising outcomes, by way of consumer assessments of the

celebrity endorser motive.

70

Conclusions

Celebrity Image Dimensions

 

Four distinct dimensions of celebrity images emerged: Genuineness, Competence,
Excitement, and Sociability. Each factor or dimension reflected certain traits or
characteristics that describe the celebrity images. Although the items representing the
dimensions were mainly derived from the brand personality scale (Aaker 1997) and the
self-concept scale (Malhotra 1981), there are differences as well as commonalities
between the celebrity image scale, the general human personality scale, and the brand
personality scale.

Of particular interest is that the celebrity image scale resembles the brand
personality scale in that the four dimensions capture the ideas of sincerity, competence,
and excitement reflected in the brand personality scale although the specific items
representing each dimension of the two scales do not exactly coincide. Celebrities are
humans; however their appearances and qualities often times embody popular meanings
in culture similar to those products or brands signify. Also, celebrities are successful in
that they enjoy the status and the privileges that come with public recognition. The scale
echoes this idea by emphasizing characteristics pertaining to success. As a result,
celebrity images might be reflected and measured by a unique combination of traits that
mirror the special nature of celebrities.

As for the relationship between celebrity image and credibility, celebrity image
dimensions, except for “Sociability,” and celebrity credibility perceptions appeared to be

significantly related to each other. Credibility components (attractiveness, trustworthiness,

71

and expertise) are also characteristics ofa celebrity and thus these may be interconnected
with other traits of the celebrity. In addition, the influences of the celebrity image
dimensions on the celebrity credibility perceptions were assessed and only one causal
linkage from “Genuineness” to trustworthiness was found to be significant. Perhaps the
image dimension of the celebrity that reflects qualities of a good person (e.g., pleasant,
sophisticated, wise, and responsible) induced sincere and honest perceptions of the

celebrity.

Congruence Perception: Role Of Perceived Expertise

This study focused on the role of perceived expertise as a basis for a fit between a
celebrity endorser and a product endorsed. Although the three products (i.e., tennis
racquet, shampoo, and coffee maker) were selected in the pretest as appropriate products
for this study, the shampoo and the coffee maker did not create a significant difference in
the perceived expertise of the celebrity who was a professional tennis player although the
ratings were in the expected direction (i.e., the celebrity was perceived to have more
knowledge of a shampoo than a coffee maker). However, the two products generated
significantly different beliefs in the celebrity’s experiences with the products. As
expected, the professional athlete was believed to have used a shampoo more strongly
than a coffee maker. Subsequently, the three pairings of the celebrity and the product
induced significant differences in the overall congruence perceptions. Consistent with the
prediction, the tennis racquet was viewed to be most congruent with the tennis player,

followed by the shampoo and the coffee maker. Perhaps both the perceived product

72

expertise and experiences of a celebrity endorser play a role in creating the perception of

overall congruence between the celebrity and the product endorsed.

Role of Congruence

 

The results of this study highlighted the role of congruence that a celebrity
endorser and a product have in driving celebrity endorsement effectiveness. As predicted,
high celebrity/product congruence in general led to more credible perceptions of the
celebrity endorser, more favorable attitudes toward the ad and the brand, and higher
purchase intention. Of importance is that these effects of congruence were exerted
through attributions consumers make about the celebrity’s motive for endorsing a
particular product.

Three attributions of a celebrity’s reason for appearing in ads to promote a
product were suggested and examined in this study: (1) the compensation a celebrity
endorser receives in return for the endorsement, (2) the celebrity’s belief in the product
qualities, and (3) the celebrity’s desire for his or her image enhancement. As predicted,
different pairings of a celebrity and a product were found to determine the celebrity
endorser motives. That is, when a celebrity endorses a product that strongly relates to his
or her career and thus is perceived to be congruent with the celebrity, consumers
evaluated the celebrity’s motive for the endorsement more positively. Specifically,
consumers were more likely to attribute the endorsement to the celebrity’s genuine
affection for the product and desire for the image improvement rather than the monetary
gains. In contrast, when consumers were exposed to a less congruent match between a

celebrity and a product, they were more likely to carefully assess the celebrity’s true

73

purpose of promoting the product and then attribute the celebrity’s motive to the

compensation the celebrity obtains.

Role of Product Use Cue

 

The results of the study revealed weak effects of the product use statement.
Although the scores on the dependent variables were in the direction predicted in the
hypotheses (the presence of the product use cue produced more favorable advertising
outcomes than its absence), most of the differences between the two conditions were not
large enough to be statistically significant. However, consumers viewed the image
improvement to be more probable as the celebrity’s motive for the endorsement and
evaluated the endorsed brand to be more positive when they were exposed to the cue than
when they were not.

The effects of the product use statement were found to interact with the levels of
congruence for product and money attributions. Interestingly, the effects of the presence
of the celebrity’s personal statement of the product use on product and image attributions
were stronger especially under the moderate congruence condition. Consumers were
more likely to believe a celebrity endorser’s genuine affection for the product and the
need for image enhancement as the reasons why the celebrity promoted the product when
the product use statement was present within the ad than when it was not at all levels of
congruence. However, the difference between the presence and absence of the product
use cue was larger at the moderate level of congruence than the high and low levels of
congruence. This may be due to the ambiguity concerning the relationship between the

celebrity and the product endorsed. When the celebrity and product relationship is logical

74

(e.g., a tennis player for a tennis racquet) or unreasonable (a tennis player for a coffee
maker), consumers might easily assess the celebrity’s motive for endorsement. However,
when consumers are not certain about the celebrity and brand connection, they perhaps
rely on other cues or messages present; the celebrity’s product use claim was the cue in

this case.

Implications

Understanding the effectiveness of celebrity endorsers is an important issue for
both academics and practitioners. The findings of this study provide a number of
implications for both theory and practice. At the theoretical level, the research highlights
that consumers do not see celebrities as unidimensional individuals (Erdogan, Baker and
Tagg 2001). Instead of evaluating celebrities on a single dimension of credibility or
attractiveness, consumers view celebrities to represent a variety of meanings that are
drawn from the history of their careers.

Secondly, the findings confirm the importance Of the fit between a celebrity and
the product he or she endorses. The perceived expertise and experience using the product
a celebrity endorses played an important role in inducing congruent perceptions of the
relationship between the celebrity and the endorsed product. The overall congruence
between the celebrity and the product produced more credible perceptions of the celebrity
endorser, more favorable attitudes toward the ad and the brand, and higher behavioral
intentions. These three types of attributions of the celebrity endorser motive were
proposed as explanations for the positive influences Of the fit between the celebrity and

the endorsed product.

75

The Sheer cost to secure celebrities underscores the importance of the decision to
select right stars to promote products (Agrawal and Kamakura 1995) and thus advertising
practitioners have attempted to devise a checklist that can guide their decision-making
(Erdogan, Baker and Tagg 2001; Mieiak and Shanklin 1994). The dimensions of celebrity
images found in this research might help to understand consumer perceptions of
celebrities. In conjunction with the brand personality scale, the celebrity image scale
could also provide some criteria for evaluating celebrities in terms of their images against
the brand images.

Secondly, the findings of this study underlie that the fit between a celebrity and
the endorsed product matters more than other aspects of the endorsement. Consumers
would respond better when they perceived that the celebrity endorser has knowledge or
experience using the product. Thus, managers should realize that in using celebrities,
much of the success might lie in first selecting the right person and that consumers want

to see some kind of rational link between the product and the star endorsing that product.

Limitations
Several theoretical and methodological limitations merit discussion. First, the
generalizability of the celebrity image scale is limited to the celebrities and the subjects
used for this study. Although professional athletes appear in advertisements most often to
promote a variety of products, the use of one particular group of celebrities (i.e., famous
sports figures) confines the generalizability of the findings. Also, the scale developed in
the exploratory phase was not found to be robust with another professional athlete in the

confirmatory phase. This lack Of robustness of the scale might be due in part to the

76

subjects’ fatigue and boredom with the long list of items against which they were asked
to evaluate a celebrity’s image. Other similar research often used longer lists of traits or
items, but provided monetary incentives to the respondents. In addition, the scale was
tested on only student subjects. One limitation associated with factor analysis is the
potential difference in the meaning of the image items among distinct groups of people.
Therefore, to test the generality of the four celebrity image dimensions and the
measurement scale, the scale would need more analyses with various samples (e.g., male
vs. female, young vs. old) (Goldsmith and Flynn 2000).

Second, a concern with the manipulations of this research arises. The product
types used for the study (i.e., tennis racquet, Shampoo, and coffee maker) did not
precisely induce the planned pattern of the perceived expertise and experience using the
products although the overall congruence perceptions were created as planned. Perceived
expertise or experience was operationalized as a connection of a product to the celebrity’s
professional career and it was expected to determine the overall congruence perception of
the celebrity and product relationship. Although tested and selected in the pretest, the
shampoo and the coffee maker, which were intended to induce the moderate and low
congruence conditions respectively appeared to be problematic. Even though
Significantly lower scores were obtained on the product experience and overall
congruence for the coffee maker than the shampoo, the coffee maker produced a
moderate condition rather than an extreme condition of low congruence. Thus, the
significant differences between the two conditions on the dependent measures cannot be

interpreted without caution.

77

Perhaps both product categories created a moderate level of congruence, however
the different nature or characteristics of the products resulted in the discrepancies in their
effects. For example, a shampoo is a personal care product that is more image-oriented
and self-expressive than a coffee maker (Rossiter and Percy 1987). These product
characteristics could be the confounding factor in the experiment that was not controlled
for. Accordingly, future studies should examine the celebrity and product congruence
effects with various combinations of them and thus identify the factors that increase or
decrease the effectiveness of the celebrity endorsements.

Further, the manipulation of the product use cue was not strong although it was in
general successful. The statements of the product use were created to give subjects the
idea that the celebrity actually has used or uses the product. However, in an attempt to
make it realistic, the message was not too explicit or direct like the way it was
manipulated in the previous research (Speck, Schumann and Thompson 1988). Future
research could investigate other ways to induce the belief that the celebrity endorser uses
the brand him- or herself.

In addition, information on subjects’ product knowledge or usage, which might
influence their responses to the celebrity endorsements, was not obtained in the study.
For example, consumers who play tennis would pay closes attention to Pete Sampras’
recommendation of a particular tennis racquet brand. The effects of consumers’
experiences with products on their brand attitudes and purchase intention should be
assessed in future studies.

Finally, gender differences found in this study might be due to the fact that only a

male celebrity was used in the study. Females showed more positive responses to the

78

endorsements in terms Of the celebrity endorser credibility, attribution and brand attitudes
than males. This could be also due to females’ tendency to view celebrity endorsements
more favorably; however, the design Of this study cannot exclude the possibility that
females respond to male celebrity endorsers more positively compared to female

celebrities. Future studies Should investigate these gender effects more carefully.

Future Research

 

Future research should continue in this area. First, the celebrity image dimensions
and the scale development need more research efforts. Although this research identified
the four dimensions of the celebrity images and developed the scale to measure the image
dimensions, the results were not convincing when the scale was re-tested on a different
sample with a different celebrity. TO develop a more reliable and valid scale to measure
the celebrity images, a variety of celebrities that have different careers and project diverse
images should be tested. In addition, consumer perceptions of celebrity images would
vary with their demographic characteristics such as gender and age. Thus, the robustness
of the scale should be assessed with different groups of subjects. The relationship of the
celebrity image dimensions to other aspects or characteristics of celebrities also warrants
future research. In particular, it would be interesting to explore how a celebrity’s images
relate to credibility perceptions of the celebrity and how they contribute to the
effectiveness of the celebrity’s product endorsements.

The construct of congruence deserves more attention. As examined in this study, a
celebrity endorser’s expertise or experience with regard to the product the celebrity

endorses clearly plays an important role in producing congruent perceptions of the

79

celebrity and product relationship. The connection of the product to the celebrity
endorser’s professional career provides an Obvious, logical link between them that
consumers easily notice. However, the creation of the overall congruence perceptions
might not be a function of this single factor. There might be other ways on which
celebrities are matched with products for endorsements. Image could also be a basis for a
fit between a celebrity endorser and a product (Till and Busler 1998). Future research
should investigate the role Of image as a match-up factor and the relative contributions of
the several bases to the creation Of congruent perceptions of the celebrity and product link
and the formation of the subsequent attitudes and behavioral intentions. With regard to
this issue, a variety of products should be used.

There is also an opportunity to explore the role of involvement in the congruence
effects (Ohanian 1990; Till and Busler 2000). Perhaps under low involvement, the
perceived image congruence between the celebrity endorser and the endorsed product is
important in generating the effects of the endorsement because the celebrity’s image
might contribute to consumer assessments of the overall congruence due to the readily
available knowledge about the celebrity’s inherent characteristics of fame and popularity.
On the contrary, the celebrity’s expertise-based congruence with the product and the
subsequent credibility assessments might be more important in leading to the
endorsement effects under high involvement. Related to the issue, another question arises
as to how the attributions consumers make about the celebrity endorser’s motive vary
with different levels of involvement or other situations.

An interesting finding of this study is related to the role ofcelebrity/product

congruence in influencing consumer assessments of celebrity endorsers’ motives. When

80

consumers recognized no obvious, logical relationship between a celebrity endorser and a
product that the celebrity endorses, they were more likely to attribute the celebrity’s
motive to the monetary gains in return for the endorsement rather than the celebrity’s
affection for the product or need for image enhancement. An interesting issue would be if
these findings pertain to dead celebrities whose computer-generated images have been
widely used in advertisements. The celebrities used for this study are all living and active.
Therefore, it would be interesting for future research to investigate consumer responses to
the endorsements of deceased celebrities in terms of the endorser motive attributions and
the endorser credibility.

Another interesting question is derived from the idea of parasocial interaction. If
perceived expertise plays a role in driving the congruent perceptions and the endorsement
effectiveness, would “pseudo-expertise” work the same way? For example, would an
actor who has played a doctor on television have similar effects to those a real doctor has
when both endorse a painkiller? How believable is this kind of endorsement? Is the
effectiveness dependent upon consumer relationships with the television figure?

A concern with the product use cue manipulation calls for future research. There
might be better ways in which consumers strongly perceive or believe that the celebrity
actually uses the product he or she endorses. Or, there might be other means to create
more congruent perceptions of the initially subtle link between a celebrity and a product
and thereby making the endorsement more effective. Answers to these questions will
Shed insight into the strategic options available to practitioners after selecting a star to

endorse their products.

81

Finally, little is known about the celebrity endorsement process across cultures.
Collectivist cultures appreciate interdependence, conformity, and similarity whereas
individual cultures value independence, autonomy, and uniqueness (Markus and
Kitayarna 1991). Celebrities who represent popular cultural meanings (McCracken 1989)
might play different roles depending on consumers’ different cultural orientations.
Although studies examining Asian countries have found a higher frequency of celebrities
used in ads than in the United States (DiBenedetto, Tamate and Chandran 1992; Cutler,
Javalgi and Lee, 1995; Lin 1993), no research has explored differences in the effects of
celebrity endorsers between cultures. Future research should examine how celebrity
image dimensions and the roles of the attributions of the celebrity endorser motive and

the celebrity credibility in the celebrity endorsement process vary with culture.

APPENDIX A.

TABLES

83

Table 1.

Factors Analysis Results of Celebrity Images

 

 

Items Factor 1 Factor 11 Factor III FactorIV
Genumeness Competence Excrtement Sociability
Comfortable/ Uncomfortable .754
Pleasant/ Unpleasant .816
Organized/Unorganized .769
Rational/ Emotional .618
Modest/Vain .747
Sophisticated/Naive .740
Wise/Stupid .877
Socially responsible/Irresponsible .859
Caring/Uncaring .877
Family-oriented/Not family-oriented .732
Down-to-earth/Arrogant .743
Strong/ Weak .75 l
Confident/Apprehensive .817
Enthusiastic/Not enthusiastic .740
Detennine/Undetermined .707
Athletic/Not athletic .659
Rugged/Delicate ~876
Excitable/Calm 758
Dominating/Submissive -740
Masculine/Feminine J9]
Public/Private .792
Bold/Shy .764

 

84

Table 2.

Factor Loadings ofCelebrity Image Dimensions

 

 

Factors Indicators Unstd. Std. R2
Factor 1 socially responsible/socially irresponsible* 1.00 .59 .35
Genuineness wise/stupid .98 .55 .30
pleasant/unpleasant .82 .53 .28
comfortable/uncomfortable .97 .59 .35
sophisticate/na'r've 1 .00 .60 .36
Factor 11 strong/weak* 1.00 .78 .60
Competence confident/apprehensive .67 .57 .33
enthusiastic/not enthusiastic l .05 .71 .50
determined/ undetermined .83 .61 .3 7
Factor Ill rugged/delicatc* 1.00 .48 .23
Excitement excitable/calm 1.12 .54 .29
dominating/submissive 1.19 .64 .41
masculine/ feminine 1.78 .47 .21
Factor IV public/private* 1.00 .51 .26
Sociability bold/shy 1.50 .91 .83

 

* Reference Indicator
All coefficients are significant (p<.001).

85

Table 3.

Correlations of Latent Constructs

 

 

 

Constructs l 2 3 4 5 6 7

Celebrity Images

1. Genuineness 1 .00

2. Competence .85 1.00

3. Excitement .66 .90 1.00

4. Sociability .25a .48 .44 1.00
Celebrity Credibility

5. Attractiveness -.67 -.67 -.47 -.08b 1.00

6. Trustworthiness -.74 -47 -.42 .03b .39 1.00

7. Expertise -.55 -.62 -.56 -.16b .27 .48 1.00
ap<.05
bp>.05

All the other correlations are significant (p < .01).

86

Table 4.

Analysis of Variance: Celebrity/Product Match and Product Use Cue Effects
on Overall Congruence Perception

 

 

 

 

 

Sum of Mean
Factor Squares d.f. Square F p

Congruence condition 303.98 2 151.99 87.74 .000
Product use cue 6.00 1 6.00 3.46 .064
Congruence x Product use cue 3.72 2 1.86 1.07 .344
Error 341.27 197 1.73

Table 5.

Overall Congruence
Means and Std. Deviations
n mean" SD

High Congruence/Cue 34 5.98 .92
High Congruence/No Cue 34 6.02 1.07
Moderate Congruence/Cue 34 3.99 1.39
Moderate Congruence/No Cue 34 3.49 1.65
Low Congruence/Cue 33 3.45 1.40
Low Congruence/No Cue 34 2.88 1.34

 

a . . .
average of an eight-Item. seven-pomt scale

87

Table 6.

Factor Analysis Results of Attributions Of Endorser Motives

 

Items

Factor I

(Money) (Product)

Factor 11

Factor 11]
(Image)

 

1. The endorser does the ad solely for the money.

9.

. The endorser does the ad only because he is paid to

do it.

. The endorser appears in the ad only to receive the

compensation.

. The endorser agreed to do the ad because of the

characteristics of the brand.

. The endorser appears in the ad because he likes the

brand.

. The endorser does the ad because he really believes

in the brand.

. The endorser promotes the brand because it is

consistent with his image.

. The endorser promotes the brand to receive

attention.
The endorser becomes associated with the brand to
improve his image.

10. The endorser appears in the ad to enhance his

public image

1 l. The endorser appears in the ad to get exposure.

.826

.882

.855

.748

.733

.800

.762

.828

.765

.857
.818

 

88

Table 7.

Analysis of Variance: Congruence and Product Use Cue Effects
on Attribution to Money

 

 

 

 

 

 

 

 

Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 27.14 2 13.57 9.66 .000
Product use cue .53 l .53 .38 .539
Congruence x Product use cue 2.38 2 1.19 .85 .430
Error 276.71 197 1.41
Table 8.
Analysis of Variance: Congruence and Product Use Cue Effects
on Attribution to Product
Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 87.73 2 43.87 35.45 .000
Product use cue 3.60 l 3.60 2.91 .089
Congruence x Product use cue 7.29 2 3.64 2.95 .055
Error 243.77 197 1.24
Table 9.
Analysis of Variance: Congruence and Product Use Cue Effects
on Attribution to Image
Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 80.10 2 40.05 40.18 .000
Product use one 4.46 1 4.46 4.48 .036
Congruence x Product use cue 5.57 2 2.78 2.79 .064
Error 196.39 197 1.00

 

89

Table 10.

Attributions to Money, Product, and Image
Means and Std. Deviations

 

 

Moneya Productl’ ImageC
n mean SD mean SD mean SD
High Congruence/Cue 34 4.86 1.22 4.43 1.07 4.60 1.20
High Congruence/No Cue 34 4.70 1.23 4.65 1.13 4.73 .85
Moderate Congruence/Cue 34 5.49 1.06 3.35 1.15 3.56 1.17
Moderate Congruence/No Cue 34 5.85 .89 2.65 1.08 2.89 1.16
Low Congruence/Cue 33 5.21 1.38 3.52 1.28 3.66 1.01
Low Congruence/No Cue 34 5.32 1.28 3.20 .96 3.31 1.33

 

a . .

average of a three-item. seven—pornt scale
a . .

average of a four-item, seven-pomt scale
a . .

average of a four-item, seven-pornt scale

90

Table 1 1.

Analysis of Variance: Congruence and Product Use Cue Effects
on Endorser Attractiveness

 

 

 

 

 

 

 

 

Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 1.91 2 .95 .88 .415
Product use cue 4.03 1 4.03 3.74 .055
Congruence x Product use cue 2.14 2 1.07 .99 .372
Error 212.08 197 1.08
Table 12.
Analysis Of Variance: Congruence and Product Use Cue Effects
on Endorser Trustworthiness
Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 7.18 2 3.59 4.12 .016
Product use cue .10 1 .10 .12 .729
Congruence x Product use cue 1.89 2 .94 1.11 .332
Error 167.88 197 .85
Table 13.
Analysis of Variance: Congruence and Product Use Cue Effects
on Endorser Expertise
Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 95.78 2 47.89 31.36 .000
Product use cue 1.70 l 1.70 1.11 .293
Congruence x Product use cue .51 2 .26 .17 .846
Error 300.81 197 1.53

 

91

Table 14.

Endorser Credibility Perceptions
Means a and Std. Deviations

 

 

Attractiveness Trustworthiness Expertise
n mean SD mean SD mean SD
High Congruence/Cue 34 4.60 .88 4.95 1.05 6.41 .82
High Congruence/No Cue 34 4.70 1.26 5.08 .78 6.11 .90
Moderate Congruence/Cue 34 4.74 1.02 4.87 .84 4.97 1.43
Moderate Congruence/No Cue 34 4.57 1.05 4.56 1.11 4.78 1.40
Low Congruence/Cue 33 4.73 .95 4.54 .81 4.76 1.31
Low Congruence/No Cue 34 4.16 1.03 4.58 .91 4.71 1.42

 

a . .
average of a five-item, seven-pomt scale

92

Table 15.

Analysis of Variance: Congruence and Product Use Cue Effects
on Attitude toward the Ad

 

 

 

Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 25.31 2 12.65 8.08 .000
Product use cue 1.27 1 1.27 .81 .369
Congruence x Produce use cue 6.26 2 3.13 2.00 .138
Error 307.04 196 1.57
Table 16.

Analysis of Variance: Congruence and Product Use Cue Effects
on Attitude toward the Brand

 

 

 

Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 17.41 2 8.70 8.11 .000
Product use cue 4.33 1 4.33 4.04 .046
Congruence x Produce use cue 3.45 2 1.72 1.61 .203
Error 211.31 197 1.07
Table 17.

Analysis of Variance: Congruence and Product Use Cue Effects
on Purchase Intent

 

 

Sum of Mean
Factor Squares d.f. Square F p
Congruence condition 122.42 2 61.21 31.25 .000
Product use cue .00 1 .00 .005 .999
Congruence x Produce use cue 3.21 2 1.62 .82 .443
Error 385.90 197 1.96

 

93

Table 18.

Ad Attitude, Brand Attitude and Purchase Intent
Means and Std. Deviations

 

 

Ad Attitude ’ Brand Attitude “ PI 3
It Mean SD mean SD mean SD
High Congruence/Cue 34 4.53 1.30 5.02 1.14 4.02 1.43
High Congruence/NO Cue 34 4.69 1.19 4.98 .96 4.23 1.16
Moderate Congruence/Cue 34 3.79 1.25 4.39 1.17 2.40 1.61
Moderate Congruence/No Cue 34 3.81 1.41 4.22 1.01 2.05 1.22
Low Congruence/Cue 33 4.28 1.11 4.82 .78 3.28 1.50
Low Congruence/No Cue 34 3.63 1.23 4.17 1.11 3.14 1.39

 

a . .
average of a five-item, seven-pornt scale
a . .
average of a four-item, seven-pornt scale
a . .
average of a three-item, seven-pomt scale

94

APPENDIX B.

FIGURES

95

Figure 1.

Effects of Congruence on Attributions

 

 

 

—0- 'Money

— — Product

f- 1- Image

 

 

 

Racquet Shampoo Coffee Maker

96

 

 

Figure 2.

Effects ofCongruence on Ad Effectiveness

 

5.5 4

5-
4.5‘ k\__' 4—

 

 

4 ‘ s ‘— - — I 4 —."' 'Aad
3.5 T ‘ — -Ab
3 -1 \ , ‘ ' i' PI

 

 

 

2.5 a ‘ '

1.5 -

 

 

 

Racquet Shampoo Coffee Maker

97

Figure 3.

Effects of Congruence and Product Use Cue on Product Attribution

 

 

+Cue
-- No cue

 

 

 

 

 

 

 

2 . r
Racquet Shampoo Coffee Maker

98

Figure 4.

Effects ofCongruence and Produce Use Cue on Image Attribution

 

 

 

--Nocue

 

 

 

 

2 I I
Racquet Shampoo Coffee Maker

 

99

APPENDIX C.

SAMPLE QUESTIONANIRE - STUDY 1

100

SURVEY

The following questions are about your reactions to a professional athlete. Please
mark an “X" in the space closest to the word that best reflects W
W. Please read carefully and answer the questions. If you have any
questions, please raise your hand.

ANDRE AGASSI

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rugged Delicate
Excitable Calm
Uncomfortable Comfortable
Dominating Submissive
Thrifty Indulgent
Pleasant Unpleasant
Complex Simple
Organized Unorganized
Rational Emotional
Youthful Mature

Formal Informal
Orthodox Liberal
Contemporary Noncontemporary
Colorless Colorful
Modest Vain
Aggressive Defensive
Naive Sophisticated
Usual Unusual
Competitive Cooperative
Conservative Liberal

Strong Weak
Awkward Graceful
Modern Old fashioned
Conformist Nonconformist
Introvert Extrovert
Adventurous Timid

Passive Active
Confident Apprehensive
Sporty Businesslike
Entertaining Not entertaining

 

101

ANDRE AGASSI (conr'b)

Impulsive

 

Enthusiastic

 

Leader

 

Private

 

Fun

 

Stupid

 

Creative

 

Determined

 

Hard-working

 

Charismatic

 

Socially

 

responsible

 

Canng
Religious

 

Athletic

 

Well-dressed

 

Ethical

 

Misunderstood

 

Lonely

 

Mysterious

 

Shy

 

Family-oriented

 

Down-to-earth

 

 

Trendy
Dependent

 

Insecure

 

Tense

 

Masculine

 

Likable

 

Unfamiliar

 

Popular

 

Unsuccessful

 

102

Deliberate

Not enthusiastic
Follower

Public

BOHng

Wise

Uncreative
Undetermined
Lazy
Uncharismatic
Socially
irresponsible
Uncaflng
Unreligious
Not athletic
Not well-dressed
Unethical

Well-understood
Sociable

Not mysterious
Bold

Not family—oriented

Arrogant
Not trendy
Independent
Secure
Relaxed

Feminine
Not likable
Familiar
Unpopular

Successful

The following questions are about your reactions to a professional athlete. Please
mark an “X” in the space closest to the word that best reflects W
W. Please read carefully and answer the questions.

ANNA KOURNIKOVA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rugged Delicate
Excitable Calm
Uncomfortable Comfortable
Dominating Submissive
Thrifty Indulgent
Pleasant Unpleasant
Complex Simple
Organized Unorganized
Rational Emotional
Youthful Mature

Formal Informal
Orthodox Liberal
Contemporary Noncontemporary
Colorless Colorful
Modest Vain
Aggressive Defensive
Naive Sophisticated
Usual Unusual
Competitive Cooperative
Conservative Liberal

Strong Weak
Awkward Graceful
Modern Old fashioned
Conformist Nonconformist
Introvert Extrovert
Adventurous Timid

Passive Active
Confident Apprehensive
Sporty Businesslike
Entertaining Not entertaining

 

103

Impulsive
Enthusiastic
Leader
Private

Fun

Stupid

Creative
Determined
Hard-working
Charismatic
Socially
responsible
Caflng
Religious
Athletic
Well-dressed
Ethical

Misunderstood
Lonely
Mysterious

Shy
Family-oriented

Down-to-earth
Trendy
Dependent
Insecure

Tense

Masculine
Likable
Unfamiliar
Popular
Unsuccessful

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANNA KOURNIKOVA (counuueo)

Deliberate

Not enthusiastic
Follower

Public

Bonng

Wise

Uncreative
Undetermined
Lazy
Uncharismatic
Socially
irresponsible
Uncaring
Unreligious
Not athletic
Not well-dressed
Unethical

Well-understood
Sociable

Not mysterious
Bold

Not family-oriented

Arrogant

Not trendy
Independent
Secure
Relaxed

Feminine
Not likable
Familiar
Unpopular
Successful

5mm

 

 

 

This part is to get some W. Please mark the appropriate
answers.
1. What is your gender? Male Female
2. What is your age? years
3. What is your major?
4. What is your year in college?
(1) Freshman (2) Sophomore (3) Junior
(4) Senior (5) Other (please specify)
5. What is your current marital status?
(1) Single (2) Married (3) Divorced
(4) Widowed (5) Other (please specify)
6. How do you describe your ethnic origin?
(1) Caucasian (2) African-American (3) Asian

(4) Hispanic (5) Native American (6) Other (please specify)

Thank you!

105

APPENDIX D.

SAMPLE QUESTIONNAIRE - STUDY 2

106

SURVEY

INSTRUCIONS: The following questions are about mm

mm. Please read carefully and answer the questions. If you have any
questions, please raise your hand.

m
Below is a set of word pairs. Please mark an ‘X' in the space closest to the word that

best reflects personalities or images of W you

saw in the ad. For example, if you think the endorser is convincing, you might
respond like this:

 

 

 

 

 

 

 

 

 

 

 

Convincing X Unconvincing
ENDORSER

Attractive Unattractive
Not classy Classy
Handsome Ugly
Plain Elegant
Sexy Not sexy
Undependable Dependable
Honest Dishonest
Unreliable Reliable
Sincere Insincere
Untrustworthy Trustworthy
Expert Not an expert

 

Inexpenenced

 

Knowledgeable

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Expenenced
Unknowledgeable

Unqualified Qualified
Skilled Unskilled
Unbelievable Believable
Likable Not likable
Unfamiliar Familiar
Successful Unsuccessful
Masculine Feminine
Ethical Unethical
Rugged Delicate
Excitable Calm
Uncomfortable Comfortable
Dominating Submissive
Pleasant Unpleasant
Naive Sophisticated

 

107

ENDORSER (Cont'd)

Organized

 

Rahonal

 

Modest

 

Strong

 

Confident

 

Enthusiastic

 

Private

 

Determined

 

Canng

 

athletic

 

Wise

 

Shy

 

 

Family-oriented

Down-to—earth

 

Socially

 

responsible

Unorganized
Emotional
Vain

Weak
Apprehensive

Unenthusiastic
Public
Undetermined
Not caring

Not Athletic

Stupid

Bold

Not family—oriented
Arrogant

Socially
irresponsible

mu.
How likely' IS it that each of the following statements about WM;

WM' IS true? Please circle the answers that most reflect
your opinions.
Extremely Extremely
Likely Unlikely
1. The endorser actually uses the brand. 1 2 3 4 5 6 7
2. The endorser has used the brand. 1 2 3 4 S 6 7
3. The endorser is able to assess the qualities of the brand. 1 2 3 4 5 6 7
4. The endorser knows a lot about the brand. 1 2 3 4 5 6 7
5. The endorser’s performance or activities are related to the brand. 1 2 3 4 S 6 7
6. The endorser is competent to discuss the brand 1 2 3 4 S 6 7

based on his professional experiences.

108

Below is a set of word pairs. Please mark an ‘X' in the space closest to the word
which best reflects your feelings about them-and advertised in the ad you saw.

AERIUS Tennis Racquet

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Good Bad
Unpleasant Pleasant
Favorable Unfavorable
Boring Interesting
Like Dislike
Agreeable Disagreeable
Believable Unbelievable
Unconvincing Convincing
Credible Not credible
Untrustworthy Trustworthy
Low quality High quality
Useful Useless
Not beneficial Beneficial
Valuable Worthless
m

1. The next time you consider purchasing a tennis racquet, what is the likelihood

that you would consider the advertised brand, “Aerius”?
Extremely Likely 1 2 3 4 5 6 7 Extremely Unlikely

2. Please indicate the extent to which you agree or disagree with each of the

following statements.

Strongly Strongly
Agree Disagree
(1) The next time I consider purchasing a tennis racquet, 1 2 3 4 5 6 7
I would inquire about “Aerius.”
(2) The next time I purchase a tennis racquet, 1 2 3 4 S 6 7

I would buy “Aerius.”

109

Below is a set of word pairs. Please mark an ‘X’ in the space closest to the word that
best reflects your feelings about W.

 

 

 

 

 

 

 

 

 

ADVERTISEMENT
Good : : : : : : Bad
Unpleasant : : : : : : Pleasant
Favorable : : : : : : Unfavorable
Boring : : : : : : Interesting
Like : : : : : : Dislike
Uninformative : : : : : : Informative
Believable : : : : : : Unbelievable
Unconvincing : : : : : : Convincing
Credible : : : : : : Not credible
Untrustworthy : : : : : : Trustworthy

 

ma.
How likely is it that each of the following statements about W
W' Is true? Please circle the

answers that most reflect your opinions.

Extremely Extremely
Likely Unlikely
1. The endorser agreed to do the ad 1 2 3 4 5 6 7

because of the characteristics of the brand.

The endorser does the ad solely for the money.

The endorser promotes the brand to receive attention.

The endorser appears in the ad because he likes the brand.

The endorser becomes associated with the brand to improve his image.
The endorser does the ad only because he is paid to do it.

The endorser promotes the brand because it is consistent with his image.

The endorser does the ad because he really believes in the brand.

PWNP‘WPP’N

The endorser appears in the ad only to receive the compensation.
10. The endorser appears in the ad to enhance his public image.
11. The endorser does the ad to get free products.

HHHHHHHHi—AHH
NNNNNNNNNNN
wwwwwwwwwww
AAAAAPPAPAA
U1U‘IUIUWU'IUIU'1U1U1UIU'I
GONONONO‘O‘O‘O‘ONO‘CD
\1\l\l\1\l\l\l\l\1\l\l

12. The endorser appears in the ad to get exposure.

mu.
Below is a set of word pairs. Please mark an ‘X’ in the space closest to the word

which best reflects your feelings about WW.
TENNIS RACQUETS

 

 

 

 

Unimportant : : : : : : Important

Of concern to me : : : : : : Of no concern to me
Irrelevant : : : : : : Relevant

Means a lot to me : : : : : : Means nothing to me
Doesn’t matter : : : : : : Matters to me

 

110

SectiongL
Below is a set of word pairs. Please check the word within each pair that best

describes how you feel about MW
RECESS). in the ad you saw.

ENDORSER and TENNIS RACQUET

 

 

 

 

 

 

 

Compatible : : : : : : Not compatible
Bad fit : : : : : : Good fit
Relevant : : : : : : Irrelevant
Incongruent : : : : : : Congruent
Good match : : : : : : Bad match
Inconsistent : : : : : : Consistent
Appropriate : : : : : : Inappropriate
Ineffective : : : : : : Effective

 

MILL
How likely is it that each of the following statements about W
W IS true? Please circle the answers that most reflect your
opinions.

Extremely Extremely
Likely Unlikely

1. The endorser actually uses a tennis racquet. I 2 3 4 5 6 7

2. The endorser has used tennis racquets. 1 2 3 4 S 6 7

3. The endorser is able to assess qualities of tennis racquets in general. 1 2 3 4 S 6 7

4. The endorser knows a lot about tennis racquets in general. 1 2 3 4 S 6 7

5. Tennis racquets are related to the endorser's performance or activities. 1 2 3 4 5 6 7

6. The endorser is competent to discuss tennis racquets in general 1 2 3 4 S 6 7

based on his professional experiences.
$331211.
1- Did you recognize WW?

(1) No (2) Yes

 

2. MES, please write down the person's name and profession in the space below.

Name:

 

Profession:

 

111

Minus.
This part is to get some information about yourself. Please mark the appropriate
answer.

 

 

 

1. What is your gender? Male Female
2. What is your age? years
3. What is your major?
4. What is your year in college?
(1) Freshman (2) Sophomore (3) Junior
(4) Senior (5) Other (please specify)
5. What is your current marital status?
(1) Single (2) Married (3) Divorced
(4) Widowed (5) Other (please specify)
6. How do you describe your ethnic origin?
(1) Caucasian (2) African-American (3) Asian
(4) Hispanic (5) Native American (6) Other (please specify)

Thank you!

APPENDIX E.

STIMULUS MATERIALS - STUDY 2

113

A. High Congruence / No Product Use Cue

pete sampras. ..
...thirtccn gram 1 Siam trophies...

nothing...but the best.

the best tennis racquet

 

114

B. High Congruence / Product Use Cue

pete sampras...
...thil‘teen grand sIam trophies...
nothing...but the best.

     
 

"When I use Aerius,
I make the best shots.

I am a winner every time.

   

115

C. Moderate Congruence / No Product Use Cue

Pete sampras...
...thirteen grand slam trophies...

nothing...but the best.

the best hair care product
- men's shampoo and conditioner

 

116

D. Moderate Congruence / Product Use Cue

pete sampras...
...thirteen grand skim trophies...

nothing...but the best.

     
 

"When I use Aerius,
I'm the best I can be.
I am a winner every time."

M“
an. n.

AGRIUS 7"

~. the best hair care product
,..-‘ \ - men‘s shampoo and conditioner

 

117

E. Low Congruence / No Product Use Cue

pete sampras...
...thirtCCn grand shim trophies...

nothing...but the best.

the best coffee maker

 

118

F. Low Congruence / Product Use Cue

pete sampras. ..
...thirteen grand 513m trophies...

nothing...but the best.

     
 

"When I use Aerius,
I make the best coffee.
I am a winner every time."

at t
rm.“
3, 9?

K \ the best coffee maker

 

119

BIBLIOGRAPHY

Aaker, Jennifer L. (1997), “Dimensions of Brand Personality,” Journal of Marketing
Research, 34 (August), 347-356.

Agrawal, J. and Kamakura, W. A. (1995), “The Economic Worth of Celebrity Endorsers:
An Event Study Analysis,” Journal of Marketing, 59 (3), 56-63.

Alperstein, Neil M. (1991), "Imaginary Social Relationships with Celebrities Appearing
in Television Commercials," Journal of Broadcasting and Electronic Media, 35
(1 ), 43-58.

American Heritage Dictionary of the English Language, the (1980), 3rd ed., Boston:
Houghton Mifflin.

Atkin, Charles and Martin Block (1983), "Effectiveness of Celebrity Endorsers," Journal
of Advertising Research, 23 (1), 57-61.

Baker, Michael J. and Gilbert A. Churchill, Jr. (1977), “The Impact of Physical Attractive
Models on Advertising Evaluations,” Journal of Marketing Research, 14 (4), 538-
555.

Basil, Michael D. (1996), "Identification as 3 Mediator of Celebrity Effects," Journal of
Broadcasting and Electronic Media, 40 (4), 478-495.

Bem, Daryl (1972), “Self Perception Theory,” in Advances in Experimental Social
Psychology, Vol. 6, Leonard Berkowitz, ed., New York: Academic Press.

Caballero, Marjorie J., James R. Lumpkin and Charles S. Madden (1989), “Using
Physical Attractiveness as an Advertising Tool: An Empirical Test of the

Attraction Phenomenon,” Journal of Advertising Research, August/September,
16-22.

Caughey, J. (1984), Imaginary Social Worlds: A Cultural Approach, Lincoln: University
of Nebraska Press.

C ocanougher, A. B., and G. D. Bruce (1971), “Socially Distant Reference Groups and
Consumer Aspirations,” Journal of Marketing Research, 8, 379-381.

Crimmins, James and Martin Horn ( 1996), "Sponsorship: From Management Ego Trip to
Marketing Success," Journal ofAdvertising Research, 36 (4), 11-21.

Cutler, Bob D., Rajshekhar G. Javalgi and Dongdae Lee (1995) ”The Portrayal of People
in Magazine Advertisements: The United States and Korea,” Journal of
International Consumer Marketing, Vol. 8 (2), 45-58.

Dean, Dwane Hal (1999), "Brand Endorsement, Popularity, and Event Sponsorship as
Advertising Cues Affecting Consumer Pre-Purchase Attitudes," Journal of
Advertising, 28 (3), 1-12.

Debevec, Kathleen and Easwar Iyer (1986), "The Influence of Spokespersons in Altering
a Product's Gender Image: Implications for Advertising Effectiveness," Journal of
Advertising, 15(4), 12-20.

Devevec, Kathleen and Jerome B. Keman (1984), “More Evidence on the Effects of a
Presenter’s Physical Attractiveness: Some Cognitive, Affective and Behavioral
Consequences,” in Advances in Consumer Research, Vol. 1 l, T. C. Kinnear, ed.
Provo, UT: Association for Consumer Research.

Debono, Kenneth G. and Richard J. Hamish (1988), “Source Expertise, Source
Attractiveness, and the Processing of Persuasive Information: A Functional
Approach,” Journal of Personality and Social Psychology, 55 (4), 541 -546.

DiBenedetto, Anthony, Mariko Tamate and Rajan Chandran (1992) “Developing

Creative Advertising Strategy for the Japanese Marketplace,” Journal of
Advertising Research, 32 (1), 39-48.

Eagley, Alice H., Wendy Wood and Shelly Chaiken (1978), “Causal Inferences About
Communicators and Their Effect on Opinion Change,” Journal of Personality and
Social Psychology, 36, 424-435.

Erdogan, B. Zafer (1999), "Celebrity Endorsement: A Literature Review," Journal of
Marketing Management, 15 (4), 291-314.

Erdogan, B. Zafer, Michael J. Baker and Stephen Tagg (2001), “Selecting Celebrity
Endorsers: The Practitioner’s Perspective,” Journal of Advertising Research,
May/June, 39-48.

F eick, Lawrence and Robin A. Higie (1992), “The Effects of Preference Heterogeneity
and Source Characteristic on Ad Processing and Judgments about Endorsers,”
Journal of Advertising, 21 (2), 9-.

Fiske, Susan T. (1982), “Schema-triggered Affect: Applications to Social Perception,” in
Affect and Cognition: The Seventeenth Annual Carnegie Symposium on
Cognition, Magaret Sydnor Clark and Susan T, Fiske, eds., Hillsdale, NY:
Lawrence Erlbaum Associates, 55-78.

Fiske, Susan T. and M. Pavelchak (1986), “Category-based Versus Piecemeal-based
Affective Responses: Developments in Schema-triggered Affect,” in Handbook of
Motivation and Cognition: Foundations of Social Behavior, R.M. Sorrentino and
E. T. Higgins, eds., New York: The Guilford Press.

F riden, Jon B, (1984), “Advertising Spokesperson Effects: An Examination of Endorser
Type and Gender on Two Audiences,” Journal of Advertising Research, 24, 33-41.

F olkes, Valerie S. (1988), "Recent Attribution Research in Consumer Behavior: A

Review and New Directions," Journal of Consumer Research, 14 (March), 548-
565.

F orkan, J. (1980), "Product Matchup Key to Effective Star presentations," Advertising
Age, 51 (October 6), 42.

Friedman, Hershey H. and Linda Friedman (1979), "Endorser Effectiveness by Product
Type," Journal of Advertising Research, 19 (5), 63-71.

Gabor, Andrea, Thorton Jeannye and Daniel P. Wienner (1987), “Star Turns That Can
Turn Star-Crossed,” U. S. News and World Report (Dec.7), 57.

Goldsmith, Ronald E. and Leisa R. Flynn (2000), “Gender Differences in Self-Image
Described by Malhotra’s Self-Concept Scale,” Psychological Reports, 86, 1213-
1 21 7.

Gwinner, Kevin P. and John Eaton (1999), "Building Brand Image Through Event
Sponsorship: The Role of Image Transfer," Journal of Advertising, 28 (4), 47-5 7.

Horton, D., and R. Wohl (1956), “Mass Communication and Para-Social Interaction:
Observations on Intimacy at a Distance,” Psychiatry, 19, 215-229.

Hovland, Carl 1., Irving K. Janis, and Harold H. Kelley (1953), Communication and
Persuasion, New Haven, CT: Yale University Press.

Hovland, Carl 1. and Walter Weiss (1951), “The Influence of Source Credibility on
Communication Effectiveness,” Public Opinion Quarterly, 15 (Winter), 635-650.

Joreskog (1979), “Basic Ideas for Factor and Component Analysis,” Advances in Factor
Analysis and Structural Equations Models, K. G. Joreskog and D. Sorbom, eds.,
Cambridge, MA: Abt Books, 5-20.

Joseph, W. Benoy (1982), “The Credibility of Physically Attractive Communicators: A
Review,” Journal of Advertising, 11 (3), 15-24.

Kahle, Lynn R. and Pamela M. Homer (1985), "Physical Attractiveness of the Celebrity

Endorser: A Social Adaptation Perspective," Journal of Consumer Research, 1 1
(March), 954-961.

Kaikati, Jack G. (1987), "Celebrity Advertising, A Review and Synthesis," International
Journal of Advertising, 6, 93-105.

Kamen, Joseph, Abdul Azhari, and Judith Kragh (1975), "What a Spokesman Does for a
Sponsor," Journal of Advertising Research, 16 (2), 17-24.

Kamins, Michael A. (1989), "Celebrity and Noncelebrity Advertising in a Two-Sided
C ontext," Journal of Advertising Research, J une/J uly, 34-42.

Kamins, Michael A. (1990), "An Investigation into the 'Match-Up' Hypothesis in
Celebrity Advertising: When Beauty May be Only Skin Deep," Journal of
Advertising, 19 (1), 4-13.

Kamins, Michael A. and Kamal Gupta (1994), "Congruence Between Spokesperson and
Product Types: A Matchup Hypothesis Perspective," Psychology and Marketing,
11 (6), 569-586.

Kamins, Michael A., Meribeth J. Brand, Stuart A. Hoeke, and John C. Moe (1989),
"Two-sided Versus One-sided Celebrity Endorsements: The Impact on

Advertising Effectiveness and Credibility," Journal of Advertising, 18 (2), 4-10.

Kelman, Herbert C. (1961), “Processes of Opinion Change,” Public Opinion Quarterly,
25 (Winter), 57-78.

Keller, Kevin Lane (1993), "Conceptualizing, Measuring, and Managing Customer-
Based Brand Equity," Journal of Marketing, 57 (1), 1-22.

Kelley, Harold H. (1972), “Causal Schemata and the Attribution Process,” in Attribution:
Perceiving the Causes of Behavior, Edward E. Jones et al., eds., Morristown, NJ:
General Learning Press, 151-174.

King. Thomas R. (July 5, 1989), “Credibility Gap: More Consumers Find Celebrity Ads
Unpersuasive,” Wall Street Journal, B5.

Langmeyer, Lynn and Walker, Mary (1991a), “A First step to Identify the Meaning in
Celebrity Endorsers,” Advances in Consumer Research, Vol. 18, Rebecca R.
Holman and Michael R. Solomon, eds., 364-371.

Langmeyer, Lynn and Walker, Mary (1991 b), “Assessing the Affects of Celebrity
Endorsers: Preliminary Findings,” Proceedings of the American Academy of
Advertising, Rebecca R. Holman, ed., 32-42.

Lin, Carolyn (1993), “Cultural Differences in Message Strategies: A Comparison
between American and Japanese Commercials,” Journal of Advertising Research,
July/August, 40-48.

Lynch, James and Drue Schuler (1994), "The Matchup Effect of Spokesperson and
Product Congruency: A Schema Theory Interpretation," Psychology and
Marketing, 11 (5), 417-445.

MacKenzie, Scott B. and Richard J. Lutz (1989), "An Empirical Examination of the
Structural Antecedents of Attitude Toward the Ad in an Advertising Pretesting
Context," Journal of Marketing, 53 (April), 48-65.

Maddox, James E. and Ronald W. Rogers (1980), "Effects of Source Expertness,
Physical Attractiveness, and Supporting Arguments on Persuasion: A Case of

Brains Over Beauty," Journal of Personality and Social Psychology, 39 (August),
235-244.

Malhotra. Naresh K. (1981), “A Scale to Measure Self-Concepts, Person Concepts, and
Product Concepts,” Journal of Marketing Research, 18 (November), 456-464.

Markus, Hazel and Shinobu Kitayama (1991), “Culture and the Self: Implications for
Cognition, Emotion and Motivation,” Journal of Personality and Social
Psychology, 35 (2), 63-78.

Mathur, Lynette Knowles, Ike Mather and Nanda Rangan (1997), “The Wealth Effects
Associated with a Celebrity Endorser: The Michael Jordan Phenomenon,” Journal
of Advertising Research, 37 (3), 67-73.

McCracken, Grant (1989), "Who is the Celebrity Endorser? Cultural Foundations of the
Endorsement Process," Journal of Consumer Research, 16, 310-321.

McDaniel. Stephen R. (1999), “An Investigation of Match-up Effects in Sport

Sponsorship Advertising: The Implications of Consumer Advertising Schemas, ”
Psychology and Marketing, 16 (2), 163-184.

McGuire, William J. (1985), "Attitudes and Attitude Change," Handbook of Social
Psychology, Gardner Lindzey and Elliot Aronson, eds., Vol. 2, NY: Random
House, 233-346.

Meyers-Levy, Joan and Alice M. Tybout (1989), "Schema Congruity as a Basis for
Product Evaluation," Journal of Consumer Research, 16, 39-54.

Mieiak, Alan R. and William L. Shanklin (1994), “Choosing Celebrity Endorsers, ”
Marketing Management, 3 (3), 51 -59.

Misra. Shekhar and Sharon E. Beatty (1990), "Celebrity Spokesperson and Brand

Congruence: An Assessment of Recall and Affect," Journal of Business Research,
21 (2), 159-173.

Moore, David J ., John C. Mowen, and Richard Reardon (1994), "Multiple Sources in
Advertising: When Product Endorsers Are Paid by the Advertising Sponsor,"
Journal of the Academy of Marketing Science, 22 (3), 234-243.

Mowen, John C. (1979), "On Product Endorser Effectiveness: A Balance Model
Approach," Current Issues and Research in Advertising, ( ) 41-57.

Mowen, John C. and Stephen W. Brown (1981), “On Explaining and Predicting the
Effectiveness of Celebrity Endorsers,” Advances in Consumer Research, Vol. 8,
Kent M. Monroe, ed., 437-441.

Ohanian, Roobina (1990), "Construction and Validation of a Scale to Measure Celebrity

Endorsers' Perceived Expertise, Trustworthiness, and Attractiveness," Journal of
Advertising, 19 (3), 39-52.

Ohanian, Roobina (1991), "The Impact of Celebrity Spokespersons' Perceived Image on
Consumers' Intention to Purchase," Journal of Advertising Research, 31 (1), 46-
54.

O'Mahony, Sheila and Tony Meenaghan (1997/98), "The Impact of Celebrity
Endorsements on Consumers," Irish Marketing Review, 10 (2), 15-24.

Park, Chan W. and V. P. Lessig (1977), “Students and Housewives: Differences in
Susceptibility to Reference Group Influence,” Journal of Consumer Research, 4,
102-1 10.

Patzer, Gordon L. (1983), “Source Credibility as a Function of Communicator Physical
Attractiveness,” Journal of Business Research, 11 (2), 229-241.

Petty, Richard E. and John T. Cacioppo (1986), Communication and Persuasion: Central
and Peripheral Routes to Attitude Change, New York: Springer-Verlag.

Petty, Richard B, John T. Cacioppo and David Schumann (1983), “Central and Peripheral
Routes to Advertising Effectiveness: The Moderating Role of Involvement,”
Journal of Consumer Research, 10, 135-146.

Petroshius, Susan M. and Kenneth E. Crocker (1989), “An Empirical Analysis of
Spokesperson Characteristics on Advertisement and Product Evaluation,” Journal
of the Academy of Marketing Science, 17 (3), 217-225.

Rossiter, John R. and Larry Percy (1987), Advertising and Promotion Management. New
York; McGraw-Hill.

Shimp, Terence (2000). Advertising, Promotion: Supplemental Aspects of Integrated
Marketing Communications, 5‘h ed., Fort Worth, TX: The Dryden Press.

Singer, Benjamin D. (1983), "The Case for Using "Real People" in Advertising,"
Business Quarterly, 48 (Winter), 32-27.

Sirgy, M. Joseph (1982), “Self-concept in Consumer Behavior: A Critical Review,”
Journal of Consumer Research, 9, 287-300.

Sparkman, Richard M., Jr. (1982), "The Discounting Principle in the Perception of
Advertising," Advances in Consumer Research, Vol. 9, Andrew Mitchell, ed.,
277-280.

Sparkman, Richard M., Jr. and William B. Locander (1980), "Attribution Theory and
Advertising Effectiveness," Journal of Consumer Research, 7 (December), 219-
224.

Speck, Paul Surgi, David W. Schumann, and Craig Thompson (1988), "Celebrity
Endorsements — Scripts, Schema and Roles: Theoretical Framework and
Preliminary Tests," Advances in Consumer Research, Vol. 15, H. K. Hunt, ed.,

69-76.

Stafford, Marly Royne (1998), “Advertising Sex-typed Services: The Effects of Sex,
Service Type, and Employee Type on Consumer Attitudes,” Journal of
Advertising, 27 (2), 65-82.

Stemthal, Brian, Lynn W. Phillips, and Ruby Dholakia (1978), "The Persuasive Effect of
Source Credibility: A Situational Analysis," Public Opinion Quarterly, 42, 285-
314.

Stemthal, Brian, R. R. Dholakia, and C. Leavitt (1978), "The Persuasive Effect of Source
Credibility: Tests ofCognitive Response," Journal of Consumer Research, 4,

252-260.

Stever, Gayle S. (1991), “The Celebrity Appeal Questionnaire,” Psychological Reports,
68, 859-866.

Swartz, Teresa A. (1984), “Relationship between Source Expertise and Source Similarity
in an Advertising Context,” Journal of Advertising, 13 (2), 49-55.

126

Taylor, Shelley E. and Jennifer Crocker (1981), “Schematic Bases of Social Information
Processing,” in Social C ognition: The Ontario Symposium on Personality and
Social Psychology, Vol. 1, E. Tory Higgins, C. A. Herman, and M. P. Zanna, eds.,
Hillsdale, NJ: Lawrence Erlbaum Associates, 89-134.

Till, Brian D. and Michael Busler (1998), "Matching Products with Endorsers:
Attractiveness Versus Expertise," Journal of Consumer Marketing, 15 (6), 576-
586.

Till, Brian D. and Michael Busler (2000), “The Match-Up Hypothesis: Physical
Attractiveness, Expertise, and the Role of Fit on Brand Attitude, Purchase Intent
and Brand Beliefs,” Journal of Advertising, 29 (3), 1-13.

Till, Brian D. and Terence Shimp (1998), “Endorsers in Advertising: The Case of
Negative Celebrity Information,” Journal of Advertising, 27 (1), 67-82.

Tripp, Carolyn, Jensen, Thomas D., and Carlson, Les (1994), “The Effect of Multiple
Product Endorsements by Celebrities on Consumer Attitudes and Intentions,”
Journal of Consumer Research, 20, 4, 535-547.

Zaichokowski, Judith Lynne (1985), “Measuring the Involvement Construct,” Journal of
Consumer Research, 12 (December), 341-3 52.

' 1tuttiitiitttigi1