BRAND PLACEMENT: IMPRESSION MANAGEMENT PREDICTIONS OF AUDIENCE IMPACT
By
JAMES A. KARRH
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTL\L FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1998 ACKNOWLEDGMENTS
A great many people have contributed in a variety of ways to my graduate studies which culminate in this work.
My primary professional acknowledgments belong to my committee chair,
Michael Weigold, for his patient guidance and support. Many thanks are also extended to the other committee members: Richard Lutz, whose support and example have inspired
for providing advice and me since I returned to graduate studies; Deborah Treise, friendship; Kent Lancaster, whose graduate seminar inspired my dissertation work; and
Mary Ann Ferguson, for providing both research insights and career advice. 1 am deeply indebted to each of them.
unflagging Next, I thank a number of former colleagues and friends who gave encouragement: Cynthia Frisby, Oscar Anderson, Charles Davis, and Tim Bajkiewdcz.
Jody Hedge, Margi Hatch, the rest of the graduate faculty and staff, and the advertising department faculty and staff were always helpful and supportive.
Kirsten Strausbaugh provided access to subjects and made the experiment
possible.
Two professors who provided inspiration and encouragement beyond the call of
duty are Drs. John Wright, II and Joseph Pisani. My students, both at UF and at Ithaca
College, also motivated me to continue this work.
ii part of this I am grateflil to the many fellow graduate students at UF who shared journey with me. My academic colleagues deserve thanks as well for their empathy,
encouragement, and help.
Finally, I attempt to adequately recognize the support of my family: my parents,
Tobe and Barbara Karrh; by brother Bill Karrh; and my many in-laws (especially my
brother-in-law Brandon Knott, who provided administrative help).
My deepest thanks go to my v^fe, Alison Knott Karrh, whose steadfast love,
enthusiasm, and encouragement have made everything worthwhile.
iii TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS ii
LIST OF TABLES vii
LIST OF FIGURES viii
ABSTRACT
CHAPTERS
1 INTRODUCTION 1
Why Study Brand Placement? 2
The Current ' State of the Art' 5 Organization of this Dissertation 8
2 LITERATURE REVIEW 9
Brand Placement 9 Brand Placement Defined 9 Brands in the Popular Culture 11 Advertisers as Programmers 13 Current Practice and the Growth of an hidustry .... 14 Reasons for Growth 16 Research in the Industry 19 The Market Value of Brand Placement 20 Legal and Social Issues 22 Regulatory Envirormient 22 Does Creativity Suffer? 25 Vulnerable Audiences and Special Product Categories ... 27 Prosocial Uses of Brand Placement 28 Audience Processing of Drama 29 Audience Response to Media Characters 31
iv Existing Research into Brand Placement 34 Content Analyses 34 Practitioners' Assumptions and Beliefs 35 Qualitative Explorations 36 Audience Knowledge and Inferences 37 Experimental Studies 38 Identity and the Self 43 The Importance of Identity 44 Identity and Impression Management 45 Definitions 45 Historical roots 46 Dramaturgical links 47 The value for social actors 47 The value for audiences 51 The Role of Symbols 54 Results of self-symbolizing efforts 55 Characteristics of effective symbols 57 Brands as Symbols 59 The importance of brands 59 Brands as communication devices 61 Audience and Marketplace Impact 63 Hypotheses 69
3 METHOD 73
Overview of Design 73 Subjects 74 Stimulus 76 Instrument 77 Priming Manipulation 79 Statistical Analyses 80 Pretest 81
4 RESULTS 84
Overview of the Experiment and Subjects 84 Assessing the Reliability of Measures 86 Assumptions 87 Memory for the Brands 90 Attitudes Toward the Brands and Purchase Intentions 93 Identity Traits Perceived in the Movie Characters 96 Identity Traits to be Appropriated by the Audience 98 Priming Manipulation 103 Effects of Prior Viewing of the Movie 105
V Tests of Hypotheses 106 Summary of Results 128
5 CONCLUSIONS 131
Limitations of the Study 132 Reliability and Validity 132 Internal Validity 133 External Validity 136
Construct Validity and Statistical Conclusion Validity . . . 137 Conclusions and Implications for Brand Placement Research ... 138
APPENDICES
A BRAND APPEARANCES IN /?£^Z/7T5/r£5 142
B QUESTIONNAIRES FOR MOTION PICTURE STUDY .... 146
REFERENCES 161
BIOGRAPHICAL SKETCH 173
vi LIST OF TABLES
Table page
2-1 . Prior Experimental Studies on Placement Effects 39
4- 1 . Correlations Between Pairs of Items Measuring Attitude Toward the Brand (Ab) 88
4-2. Correlations Between Pairs of Items Measuring Purchase Intention for the Brand (PI) 88
4-3. Unaided Brand Recall 91
4-4. Brand Recognition Scores 92
4-5. Attitude Toward the Brand (Ab) Mean Scores, Questionnaire 1 and Questionnaire 2 94
4-6. Purchase Intention Toward the Brand (PI) Mean Scores, Questionnaire 1 and Questionnaire 2 95
4-7. Mean Ratings ofEach Character By Identity Traits 97
4-8. Mean Ratings of Self By Identity Traits 100
4-9. Correlations Among Self-Description, Importance, and Need-to-show By Identity Traits 101
4-10. Scores on Ab and PI Scales for Primed and Unprimed Groups 104
4
vii LIST OF FIGURES
Figure page
2-1. Conceptual model 67
4- 1 . General Path Model for Brand / Character / Trait Combmations 1 07
4-2. Path Analysis Results for Saab / Michael / Successful Combination on Ab 1 09
on PI 110 4-3 . Path Analysis Results for Saab / Michael / Successful Combination
4-4. Path Analysis Results for Saab / Michael / Fun-loving Combination on Ab 113
4-5. Path Analysis Results for Saab / Michael / Fun-loving Combination on PI 114
4-6. Path Analysis Results for Saab / Michael / Materialistic Combination on Ab 115
4-7. Path Analysis Results for Saab / Michael / Materialistic Combination on PI 116
4-8. Path Analysis Results for Snickers / Troy / Responsible Combination on Ab 118
4-9. Path Analysis Results for Snickers / Troy / Responsible Combination on PI 119
4-10. Path Analysis Results for Snickers / Troy / Cool Combination on Ab 120
4-11. Path Analysis Results for Snickers / Troy / Cool Combination on PI 121
4-12. Path Analysis Results for BMW / Lelaina / Attractive Combination on Ab 123
4-13. Path Analysis Resuhs for BMW / Lelaina / Attractive Combination on PI 124
4-14. Path Analysis Results for Saab / Michael / Confident Combination on Ab 126
4-15. Path Analysis Results for Saab / Michael / Confident Combination on PI 127
viii School Abstract of Dissertation Presented to the Graduate of the of the University of Florida in Partial Fulfillment Requirements for the Degree of Doctor of Philosophy
BRAND PLACEMENT: IMPRESSION MANAGEMENT PREDICTIONS OF AUDIENCE IMPACT
By
James A. Karrh
August 1998
Chairman: Michael F. Weigold, Ph.D. Major Department: Journalism and Communications
growing Brand placement (sometimes referred to as product placement) is a
feature films, television practice in an increasing number of media vehicles, including
placement, a brand is included shows, video games, and popular music. Through brand
for some consideration from an as an inseparable part of a mass media program in return
advertiser.
has generally Research into the communication impact of brand placement
programs. Missing focused on audiences' memory for the brands that are included in
might actively process from prior studies is an explanation as to how audience members
impression brand information from a media program. The theoretical framework of
predicts management was used to offer hypotheses. A new model was presented that
audience member's changes in desire for the brand due to placement as a fimction of the
ix of the identity motivation to express identity characteristics, the perceived beneficiality
perceived characteristics, the degree to which the media character who uses the brand is
the brand is perceived to to possess the identity characteristics, and the degree to which help the audience member display the identity characteristics.
An experiment was conducted in which subjects watched a twenty-five-minute
undergraduate clip from a recent movie containing a number of brands; subjects were 153 students. Support was found for the four hypotheses advanced. The model predicted changes in desire for the brand in those instances of brand placement where the media character was perceived to strongly possess certain identity characteristics, the identity characteristics were desirable, and the character was clearly matched with a brand. The change in desire for the brand was constant across expensive and inexpensive brands.
The results of the study suggest brand placement will be most effective for the advertiser
desirable when the brand is directly paired with a character who displays one or more
only traits. The results are limited in their generalizability, however, since subjects saw
one program stimulus. It was suggested that future research in brand placement
incorporate a wider variety of program contexts and audiences.
X CHAPTER 1 INTRODUCTION
A recent episode of the current top-rated network television show--NBC's
Seinfeld-contained seven instances of real, brand-name products appearing in the 22- minute program (Darlin 1995). Over a decade ago, a newly introduced bite-sized candy-
"Reese's Pieces"~was used to entice a frightened alien out of hiding in the feature film
E.T.: The Extra-Terrestrial. The film quickly generated over $300 million in box-office
receipts, and Hershey executives were presumably enjoying the taste of a reported 65% jump in sales of Reese's Pieces (Reed 1989). Similarly, when actor Tom Cruise wore
Bausch and Lomb's Ray-Ban Wayfarer sunglasses in the 1983 film Risky Business and
the Aviator model in the 1986 movie Top Gun, sales reportedly soared both times
(Liebenson 1990). When Burt Reynolds drove a black Pontiac Trans Am (at high speeds)
in the 1977 film Smokey and the Bandit, both sales and lawsuits rose~as some drivers
blamed the movie's high-speed chase scenes for their subsequent high-speed driving zeal
(Solomon and Englis 1994a).
Brand-name products are sometimes included as an inseparable part of mass
media programs. In these cases, creative professionals involved with the programs (such
as producers, directors, writers, actors, or set designers) strategically use brands to create
particular impressions on the audience. For instance, in the use of realistic drama for
storytelling, contextual factors in the program are quite important in lending
1 verisimilitude; those contextual factors include product or brand usage, stylistic variations
in staging, and the physical attractiveness of actors (Solomon and Englis 1994b). On
other occasions, brand appearances are a promotional tool; the brands appear as a result
of cash, barter, or some other form of consideration supplied from an advertiser to the
program's creator(s). Such paid brand appearances (termed "product placement" or ;
"brand placement," as discussed below) are increasingly important to advertisers and
increasingly frequent occurrences for program audiences.
Why Study Brand Placement?
* A better imderstanding of the effects of brand placement is needed, for three
primary reasons. First, brand placement is increasingly important in practice. The
process of promoting paid brand appearances has become a big business. In motion
pictures, that business was estimated in the early 1990s to total over $50 million annually
(Elliott 1992a). For placement, though, cash payments represent only a small part of the
total market. When brand appearances are "paid," they are most often the result of a
barter (trade) arrangement. They also are frequently used as part of more lucrative
promotional tie-in arrangements between advertisers and program producers. A report
based upon viewpoints from major advertisers included the recommendation to
"recognize product placement in movies as a new medium" (Elliott 1992b). The current
trend of increasingly visible brands appears likely to continue. Some seers of
advertising's interactive fiiture predict that soon "we will see a greater fiision of publicity, advertising, and careful product placement in nearly every media outlet" {Wired \99A, p.
126).
Second, placement carries important social, legal, and public policy implications.
The practice has been the subject of both legal challenges and public calls for industry restraint on the part of industry, particularly in cases where more vulnerable audience
groups are involved. Fearing effects on younger television viewers, Action for Children's
Television and the United Church of Christ asked the Federal Communication
Commission in 1989 to investigate "in-program promotions"; they did so in response to a reported deal (later dropped) between CBS and Coca-Cola to exchange advertising for prominent placement in the children's series "TV 101" {Electronic Media 1991). The plan never came to fruition. For purposes of policy making—both for children and other
vulnerable audience groups—research is needed to specify what impact placements and
other hybrid message types may have, as well as ways to mitigate any harmful effects
(Balasubramanian 1994).
A final reason for this study is that placement represents an interesting and challenging area of research. Whether brand appearances in programs are "paid for" or not, the proliferation of brands in what were at one time less overtly commercialized programs raises questions about both the intent of program creators and the subsequent reactions of program audiences. Balasubramanian (1991) classifies brand placements as part of a new genre of marketing communication tools he terms "hybrid messages"; other examples of these messages include infomercials (program-length commercials), video news releases (VNRs), and program tie-ins. Combining elements of both advertising and publicity, hybrid messages are believed to potentially generate more persuasion than do 4
more traditional advertising appeals. For example, a video news release was found to carry more persuasive impact than a similarly constructed advertisement, and the degree of difference depended upon the credibility the audience gave to television news programs generally (Owen and Karrh 1996). Research into various combinations of
advertising and promotion is needed so that we may determine whether these messages gather more attention, are viewed as less threatening, are attributed to more credible sources, or work through other processes.
As a marketing communication tool, the integration of brand-name products into entertairmient represents a fertile area for understanding consumers and how they relate to brands. Brands can be useful tools for announcing and offering evidence for a
consumer's desired personal identity. While it has been said that "clothes make the
man"—that a person's attire signals to others one's social status and personality~this perception may also be extended to a variety of product types. We represent who we are to others with the cars we drive, the foods we eat, and the fiimishings in our homes.
Brands represent a potential vehicle for attaching ourselves to real or desired identity attributes. Moreover, people use the consumption behaviors they see in order to infer personal characteristics of others. Brands serve to differentiate offerings of the same product type, and the combination of easy availability and longstanding national advertising campaigns means that hundreds of brands are well-known and well- understood as social markers. As part of entertainment programs, the manipulation of brands can thus have implications for audiences' enjoyment of the program, their identification with program characters, and their judgment of the extent to which those brands are useful social markers for showing desired elements of their own identities. 5
Thus, in at least some instances, the manipulation of brands into entertainment vehicles may represent a particularly powerful brand communication device.
The Current 'State of the Art'
^ Despite the need for study into brand placement effects, research has not kept pace. There has been little empirical study of audiences 'responses to brands as part of mass media programming. Further, the few published studies addressing this technique offer varying measures of placement effectiveness. Even placement practitioners involved in the feature film industry~a narrowly defined slice of the business, centered around the Los Angeles area-reported using widely varying measures to gauge the effectiveness of this communication tool (Karrh 1995a; Turcotte 1995). A survey of film placement professionals found "a high degree of reliance upon subjective criteria for decision-making...placement does not yet use the same language or yardsticks of
evaluation as do other advertising and promotional activities" (Karrh 1995a, p. 187).
An approach to understanding the full persuasive potential of brand placements is needed. This approach should reflect the unique character of placement as a creative promotional tool. Since placements are inseparable from program content, they should be studied as an integral part of media stories. Placements may convey useful information about characters, scenes, and story development in addition to claims of brand utility or value. They also provide cues as to how the brand might signify or provide evidence of desired identity attributes which can be readily appropriated by audience members. 6
The rather Umited body of research into brand placement has focused upon more common yardsticks of marketing communication activities such as brand recall and attitude. These brand-specific effects are important and will be considered in this dissertation. However, brands as part of programs carry social information which goes
beyond the brand itself. Audiences watching mass media programs may often scan the programs for social cues; they may seek to associate themselves with one or more characteristics of desirable characters and avoid characteristics of undesirable ones.
While audiences may not be able to mirror all the desired attributes of a successful, glamorous actor, they may be able to use some brands as a link to some desired attributes.
This set of potential brand placement effects has not been considered in other research.
The present study recognizes the unique potential role of brand placements as carriers of social information within the context of a media program. This dissertation will use the framework of impression management, which predicts (among other things) the role of symbols such as brands in asserting and reinforcing personal identities. This
framework is valuable in making predictions about the type of persuasion which might
occur in this setting (and, perhaps, in explaining why prior studies have found little or no persuasive impact from placements). The inferences by audiences may translate into substantial marketplace impact.
These indications of value from the marketplace introduce a quandary. The small
number of studies into placement have found little or no effect on brand recall or brand attitudes from exposure to film-based placements. Industry sources, however, believe the impact of placements can be powerfiil and a few cases in the marketplace have apparently been substantial. Why the inconsistency? Perhaps previous studies examining 7
placements have either used insensitive measures of placement effectiveness or have
failed to consider critical contextual factors, or both. We must remember how audiences
use brands within the context of a movie, television show, or other media program. As
will be shown in Chapter 2, audiences typically use this type of brand cue as a window
into the identity of a media character. Audiences see the clothes a character wears, the
type of car he or she drives, or other brand cues as a way of becoming anchored in the
story and anticipating what will happen next. Audiences may also see these brands
matched with characters having desirable (or undesirable) characteristics and make
inferences relevant to their own identities. To date, research into brand placement has
treated the tool as simply another marketing communication exposure with reminder
value, ignoring the context of the brand's appearance.
While previous research into brand placement (addressed in Chapter 2) has posed
the question of whether to employ brand placement as part of the marketing
communications mix, this dissertation moves forward to address the question of how to
employ brands as part of media programming, primarily from the point of view of the
advertiser. In doing so, the present research employs a theoretical orientation which offers some new insight into understanding audience use of placements, as well as related tools which combine brand advertising and entertainment. The results should prove
useful to makers of public policy, sponsors embarking on niche programming efforts, creative professionals seeking to find an appropriate use of brands in storytelling, and advertisers using brand placements. 8
Organization of This Dissertation
Chapter 2, the Literature Review, provides an overview of relevant research from
several areas of study. First, the roles of identity creation and impression management
are examined, reviewing how individuals use various sources of information (including
symbolic information) to organize and present idealized versions of themselves. Brands
are introduced as a type of social symbol, useful for providing important information to
audiences. Next, research into audience effects and marketplace impact from brand
placements is reviewed. It is argued in this dissertation that a perspective gained from
studies of symbolizing efforts in impression management can be usefiilly applied to brand placement activities. Hypotheses are developed from these literatures.
Chapter 3, Methodology, discusses the operationalization of the hypotheses, the research design, procedures, measures, and the analyses to be performed. Chapter 4 summarizes the results of the study and conclusions and recommendations are offered in
Chapter 5. CHAPTER 2 LITERATURE REVIEW
Brand Placement
Brand Placement Defined
In addressing brand placement issues, it is important to define terms properly.
Note first that what are termed "brand placements" in the present study are more
commonly, but less correctly, called "product placements" in both trade and academic
articles. It is generally a particular brand, rather than a product type, which is highlighted.
Since the term "brand placement" better captures the essence of the activity, and since it has begun to appear in academic study (Babin and Carder 1996; DeLorme, Reid, and
Zimmer 1994; Karrh 1995a, 1994), that term will be used in the present study.
Beyond the use of different terms for placement, there have also been differing definitions offered in the literature. Balasubramanian (1994) defined a placement as "a paid product message aimed at influencing movie (or television) audiences via the planned and unobtrusive entry of a branded product into a movie (or television program)"
(p. 31). Steortz (1987) defined placement activities as "the inclusion of a brand name product package, signage, or other trademark merchandise with a motion picture,
television show, or music video" (p. 22). Baker and Crawford (1995) called placement
9 10
"the inclusion of commercial products or services in any form in television or film productions in return for some sort of payment from the advertiser" (p. 2).
While each of the above definitions touches upon important parts of brand placement, each also omits some important component. The Steortz (1987) definition
omits the paid nature of placement and considers only visual stimuli; placed brands may
be included in dialogue or background audio, for instance. Both the Balasubramanian
(1991) and Baker and Crawford (1995) definitions include the paid nature of placement
but limit the media involved to motion pictures and television shows. However, many
other media may be involved-including video games, Broadway shows, music videos,
and even novels. Further, brand placements cannot always be considered "unobtrusive."
Particular brands may be used to assert character identity, set the elements of a scene
(such as its time period), and otherwise serve rather overt functions. Brand placement is
better defined as the compensated inclusion of branded products or brand identifiers,
through audio and/or visual means, within mass media programming.
It is also useful to separate brand placement (a paid effort on the part of
advertisers) from simple "brand inclusion" in programs. The latter may be considered a
catch-all term for those brand appearances which are not sponsored or compensated in
any way. Friedman (1991) uses a similar distinction in partitioning what he calls "word-
of-author advertising," or the use of brand names within the texts of screenplays,
television shows, and other "popular cultural products": he calls "sponsored word-of-
author advertising" those uses of brand names considered to be commercially inspired
and "unsponsored word-of-author advertising" those which are not. Though in many
cases the difference may be invisible to audiences, and despite the fact audiences may not 11
make a distinction between brand placement and brand inclusion (Karrh 1995b), the goals
of the communication are different. The focus of the present research will be on
purposive, commercially inspired brand placement. So that the paid use of brands in
programs is put in an appropriate context, research into the appearance of brands in a
variety of media is also considered.
Brands in the Popular Culture
Brands appear in media vehicles beyond just movies or television shows. Music
videos can be a forum for paid placement, as when RC Cola appeared in a music video
for coimtry singer Louise Mandrell (Englis, Solomon, and Olofsson 1993). Another
singer, Barbara Mandrell, released an album entitled "No Nonsense" that was made with
the financial support of the manufacturer of No Nonsense pantyhose (Snyder 1992).
Video games are now used by advertisers such as Adidas, Coca-Cola, and Midland Bank
PLC. One London-based agency claims to have negotiated 100 video game placement
deals (Pope 1994). Even some novels are affected. In her novel Power City, author Beth
Ann Herman featured a Maserati as part of the Hollywood public relations community; a
Beverly Hills Maserati dealership threw a $15,000 party for Ms. Herman that attracted national TV coverage (Snyder 1992). Brands are even appearing in a variety of
Broadway shows (Marks 1995).
The paid inclusion of brands in an increasing variety of media seems to parallel a generally increasing commercial content in American mass media Consider the use of brands in everyday speech. Friedman (1985) examined best-selling novels since World
War II as a reflection of the commercial content of American popular language, since "the 12
language used in them was appropriate for the times in which they were published" (p.
928). The incidence of brand names appearing in a sample of popular novels increased over 500 percent between the late 1940s and early 1970s (Friedman 1985). Friedman
(1986) found a similar pattern of brand name occurrence in postwar popular American music; an examination of top-ten songs from 1946-1980 found most brand mentions
occurred during the 1970-80 period. There is also a wide use of brand names to create humor in popular culture writing such as novels, plays, songs, and mass-circulation newspapers and magazines (Friedman 1991).
Solomon and Englis (1994a) use the term "reality engineering" to describe the efforts of advertisers to shape the social images presented through popular cultural vehicles. As a result of these efforts, consumers today are more likely to both accept
brand images into areas of public life that were formerly commercial-free and to see
brands and their identifiers as a natural part of everyday life. As an example, while television talk-show hosts have for years been willing to incorporate advertising into their programs, Rosie O'Donnell has carried the practice to new heights. Brands such as
Listerine, Lucky Charms, and WebTV have been prominently featured in the show, including tie-ins to O'Donnell's charitable foimdation and giveaways to the show's audience (Johnson 1997). The heavy use of familiar brands—and the implicit signal that
O'Dormell is "like the rest of us"~may be one reason for the show's popularity.
The inclusion of brands into programming is part of the trend of consumption symbols increasingly permeating the popular culture. Commerically inspired brand
placement is also part of a larger trend—that of advertising which is often inseparable from entertainment vehicles. 13
Advertisers as ProRrammers
The 1 990s have seen the revival of an old idea: advertiser control over programming. There have been calls for advertisers to take a larger role in program production, "to make sure advertiser-supported programming can continue as an integral part" of the television industry (Advertising Age 1994). Several advertisers are doing so.
Promoting the re-design of its Pathfinder sport-utility vehicle, Nissan produced a weekly half-hour television show (titled Pathfinders) featuring celebrities on outdoor adventures; the show was syndicated to local markets and included advertising for both Nissan and other sponsors (Goldman 1995). Procter & Gamble is in a partnership agreement with
Paramount TV and airs "Home Court" on NBC and "Almost Perfect" and "Good
Company" on CBS (Savan 1996). Reebok has produced an annual dowoihill-bike racing program to air on ESPN plus a number of workout shows that air on ESPN and the Cable
Health Club (Jensen 1994). A new California firm even produces what it calls "song product placements," original, three- to four-minute compositions including the name of a sponsored product or service. These "Adsongs" are subsequently sold into commercial radio airtime and available for consumer purchase on compact disk and cassette (Wriite
Radio Adsong Information Page 1996).
Other, third-party producers are contributing to the trend as well. A new television production company is reportedly producing programs with direct input from
advertisers; its goal is to create "advertising that becomes part of the show without being
obtrusive but still communicates the advertiser's message" (Ross 1997). Alton
Entertainment produced a syndicated half-hour weekly program, "Main Floor," that 14
ostensibly highlighted fashion and beauty trends. Viewers were steered to merchandise
that sponsors paid to promote on the show, and sponsors had script approval. At least
one of the show's sponsors participated in this show because it was given more control
over the final product than would have been the case with video news releases (Agins
1994).
While not all advertisers wish to be a part of program production, an increasing
number do wish to include their brands seamlessly into programming. Feature films, in
particular, have included a resurgence in brand placement since the early 1980s.
Current Practice and the Growth of an Industry
Brand placement has played a role in film production for several decades. MGM
Studios even maintained a placement office as early as the 1930s, though the practice was
sporadic (Rothenberg 1991). As recently as a decade ago, the process of placing branded
products in feature films was described by one consumer goods marketer as a "Turkish
rug market" (Steortz 1987, p. 3). In earlier years, the use of brands as props was a
sporadic and unsophisticated process. Items were simply loaned or donated to film
makers, or the directors would purchase the goods they needed at local stores (DeLorme,
Reid, and Zimmer 1994). Even in the case of bigger-ticket items, the process was a loose
one. Auto manufacturers, for example, would often send entire fleets of new cars to
movie studios, with little planning and little control over how (and if) the vehicles were
used in movies (Risen 1989). Today, the process is more orderly and costs are rising.
One agent estimates the auto industry has dedicated nearly 15 million dollars worth of vehicles to films and television shows over the past 15 years (Bunish 1995). 15
Brand placements generally occur in feature films through one of two
intermediaries-placement agents and movie studios (J. Walter Thompson USA 1989).
Placement agents function as middlemen between advertisers and movie producers, and
may work on either an ad hoc or retainer basis. Studios typically have their own
departments to handle brand placement opportunities and may work with agents as well.
Today, studios solicit placements to serve any of several purposes: to acquire needed
props for their films; to lower the costs associated with film production; and to sometimes
establish a business relationship with an advertiser that might lead to more extensive~and
lucrative-promotional tie-ins. Most placements (over 85%) are estimated to be done on a
barter basis, where the product or service is simply traded in return for exposure in the
' , film(Bunish 1995). .
Advertising agencies are not heavily involved in arranging film placements,
unless the placement arrangement involves cross-promotion into media advertising.
There are several reasons for this: placement evolved fi^om corporate public relations
efforts; specialized placement agencies have come to dominate the market; the size of the
market overall is not attractive to some advertising agencies; and brand placement does
not fit easily into the typical media buying framework (Turcotte 1995).
Placement agreements generally call for producers to use products or services in
ways consistent with typical industry practice, and bar producers fi-om depicting products
or services in any way not acceptable to the relevant marketer (Entertainment Law and
Finance 1993). Contracts originate with studios, who have the rights (or "clearances") to
depict registered trademarks in their films (Turcotte 1 995). 16
Brand placement in feature films occurs within the rather unique structure of
Hollywood. The so-called "above-the-line" talent-producers, directors, writers, and stars-maintains virtually complete control over film production and thus keeps veto power over potential placements. "Below-the-line" parties such as prop masters, production designers, and set decorators are also important to the process, though they are usually prohibited by their studios from making deals themselves. In many cases, placement deals for feature films do not survive films' final cuts; one major marketer
estimated its placements fall through in about 30% of cases (Turcotte 1995).
The market for placement activity in movies now shows signs of maturation. A trade group of placement agents, studio representatives, and marketers (the Entertainment
Resources and Marketing Association, or ERMA) has grown to nearly 100 members
since its founding in 1991 and has established a code of ethics. A similar group-the
Entertainment Marketing Association—has been established in the United Kingdom around the placement of branded products in television shows and films (Curtis 1996).
One placement agent described the growth of the business with: "Corporate America has
gone from saying 'Don't bother to call us' to 'Tell us what the next "Ninja Turtles" is
going to be'" (CrisafiiUi 1995, p. 3).
Reasons for Growth
There are a number of reasons for the growth of brand placement as a paid
promotional device. One reason is a desire on the part of advertisers to take advantage of the special characteristics of movies, television, and other media. These programs have a
"shelf life" of decades (via syndication and the video rental and sales market) and often a 17
truly global reach. Heineken, for example, used a full integrated plan around the James
Bond adventure film Tomorrow Never Dies as the basis for its first major global promotion (Amdorfer 1997).
The physical environment of in-theater viewing may amplify the impact of showing a brand, since visual and auditory features can be presented in a particularly vivid and involving way; the wide screen, dark environment, and stereo sound, for
example, make for an absorbing in-theater experience (Morrison 1 994).
Many programs themselves also carry a strong persuasive power. Movies, for example, have the power to influence audience's social judgments, at least those made shortly after exposure. Forgas and Moylan (1988) used a field study, incorporating interviews with respondents who had just attended motion pictures, to determine whether the affective quality of a film influenced subsequent social judgments. Films had been previously coded as happy, sad, or aggressive in tone. Differences were found among the three movie-viewing groups in a series of four question categories—political judgments
(including ratings of two local political figures and the current state and federal governments), judgments of the likelihood of future events (a nuclear war, the future of the economy, and improvements in personal fortunes over the next year), judgments of responsibility and guilt (the appropriateness of severe punishment for drunken driving, heroin trafficking, and obstruction ofjustice), and judgments of satisfaction with work
and personal life. Those subjects viewing "happy" films rated political leaders and governments more positively, were more optimistic in predicting future events, were more lenient toward criminals, and reported being more satisfied with their personal lives than did subjects viewing sad or aggressive movies. Those viewing aggressive or sad 18
films did not differ from each other on subsequent social judgments, but in all four
categories both differed from those who had just viewed happy films. Further, the pattern
of differences was uniform across demographic categories. The Forgas and Moylan
(1988) study suggests audiences become highly enough involved in the movie viewing
experience to feel significant changes in their outlook toward life, at least for a short
period of time.
Movies may also prompt powerful affective reactions such as empathy (Davis et
al. 1987). Morrison (1994) studied the movie-viewing experience through a series of
depth interviews. The primary reported motivations for movie viewing were learning and
entertainment/escape—a finding which points to a receptive audience willing (or even
eager) to become highly involved in the stories and characters portrayed on-screen. This
heightened involvement may extend to processing of brands in films. According to one
report, about one-fifth of movie audiences say they actively look for brand-name products
in movies {Adweek's Marketing Week 1988). Audiences can have powerful responses to
a movie, and that affective response may transfer to any brand(s) included in the film. •
Another reason for brand placement's growth is the opportunity for "implied
endorsement" from celebrities, especially those who do not engage in more traditional
endorsement deals. In the film The Formula, actor Marlon Brando's character says to
another, "Have some Milk Duds, they're good for you." Brando does not endorse
products through advertisements (Balasubramanian 1991). One advertising agency report
considers placement "as powerful as a celebrity endorsement but more subtle ... it's conceivable that viewers will acknowledge and buy products used by idols on the big
screen" (J. Walter Thompson USA 1 989, p. 2). 19
Movie and television show producers may be particularly receptive to placement
deals because of the increasing cost of making shows. According to the Motion Picture
Association of American, the average cost of producing a movie is now $39.8 million;
with marketing costs, the total is $59.7 million, an increase of 9% from the prior year
(The Wall Street Journal 1997). Placements are used by producers in order to obtain
props at no charge or, on occasion, as significant sources of cash or marketing support.
A brand placement may also be a link to larger deals involving integrated
promotional arrangements. These tie-ins may include advertising, point-of-purchase
displays, product sampling, and other tools. Not only did Tom Cruise's character use
Apple Powerbook computers throughout the 1996 film Mission: Impossible, but Apple
reportedly spent $15 million to promote the film on television and in print (Caro 1996).
Chevrolet spent a reported $10 million around Days of Thunder, while BMW invested
$20 million on the total campaign for the launch of its Z3 roadster. The BMW campaign
included prominent placement in the James Bond film GoldenEye, as well as in most of
the film's trailers and television ads (Eisenstein 1997).
Research in the Industry
While brand placement has become a much larger business, the tools advertisers
use to evaluate its impact on audiences are still relatively unsophisticated. Measures are
memory-based, generally employing aided or unaided same-day and day-after recall.
CinemaScore, a firm which conducts surveys of opening-night moviegoers to gauge audience reactions to films (Morgan 1993; Fraser 1992), occasionally includes recall studies for clients on a contract basis. On the basis of these tests, advertisers are typically 20
pleased with 25% correct recall from a given placement. This expectation is for more
"plain-vanilla" placements, in which a brand is included as part of a scene but is neither in dialogue nor shown in use. Some in the industry claim prominent placements are much more memorable; one placement agent claims "research shows product placement
has a recall 10 times greater than a commercial" (Jacobs 1996, p. 17).
Still, most decisions are made according to instinct and a general desire to help a brand become noticed. A 1989 advertising agency report concluded, "numbers on
awareness or reach/frequency are lacking . . .and the overall evaluation must be
qualitative rather than quantitative" (J. Walter Thompson USA 1989, p. 8). A 1994
survey of placement agents, studio representatives, and marketing executives still found
"a high degree of reliance upon subjective criteria for decision-making . . . placement does not yet use the same language or yardsticks of evaluation as do other advertising and
promotional activities" (Karrh 1995a, p. 187). ,
The Market Value of Brand Placement
In addition to the anecdotal "success stories" of some brand placements in movies
{E.T.: The Extra-Terrestrial and Top Gun), there is ample evidence that substantial value may be attached to individual placements. Where a cash payment has been involved, the prices demanded for more mundane placements have risen in recent years. When Walt
Disney Company's Buena Vista Pictures division solicited payments for the movie Mr.
Destiny, it reportedly asked $20,000 to show a brand, $40,000 for a mention in dialogue, and $60,000 for the star to hold the product up for the camera {The Toronto Star 1991). 21
These figures represented an increase over the $10,000 to $40,000 20th Century-Fox asked eight years earlier (Harmetz 1983).
Many bigger-ticket arrangements are made more opportunistically and, depending upon the execution of the placement, involve much larger sums of money. Philip Morris reportedly paid $350,000 to have the James Bond character smoke Lark cigarettes in
License To Kill (Miller 1990), while Exxon is said to have paid $300,000 for a placement in Days of Thunder (Kanner 1993). Some contested deals also provide some insight into placement value. Reebok is seeking $10 million from TriStar Pictures in a lawsuit over
placement in the film Jerry Maguire ; the suit alleges the studio did not honor an agreement to spotlight Reebok favorably-despite $1.5 million Reebok invested in athletic gear, cameos by Reebok athletes, and a custom-produced "commercial" which was to appear as part of the film (Pener 1997). In 1990, Black & Decker sued Fox for $150,000, after an agreement to include a new cordless drill in Die Hard 2 fell through. The drill did not appear in the film's final cut, and the parties settled out of court (Busch 1996).
Advertisers may also take active steps to avoid being part of negative portrayals.
McDonald's, for example, has enlisted an agency to keep its products out of feature films that might "offend the family unit" (Time 1989). Coca-Cola found itself at the center of controversy when a placement deal with director Oliver Stone for the film Natural Born
Killers resulted in a Coke commercial edited into a violent murder scene (Johnson 1994).
Television production is far more restricted as to placement activities than is film production. Federal Communications Commission (FCC) rules require specific identification of paid placements, and the broadcast networks often attempt to limit brand appearances in shows. In the past, networks required brand names be "greeked," or '
22
masked, usually to a generic name such as "cola." Today, the networks require affidavits
stating producers have not accepted money for brand appearances in their shows.
However, placements made through an agency bypass those policies as well as FCC
regulations; a placement firm's fee for exposure on a hit television show is typically about
$20,000 (Warner 1995).
There may be even more implied value for in-program brand appearances.
Following an analysis that showed increasing numbers of brand mentions and
appearances on network affiliate programs, the editors of Advertising Age (1993) chided
television producers for apparently giving away valuable exposure, saying "Networks are
selling this time. If it can be gotten so easily for free, why pay for it?"
Legal and Social Issues
Regulatory Environment
Brand placement is occurring more frequently in both movies and television, but these two media are treated differently under the law. The U.S. Supreme Court has consistently recognized the unique nature of each medium and has taken a "hands-off approach to the feature film industry (Snyder 1992). Because they use public resources, television stations and networks face more stringent rules.
In the cases of broadcast and cable media, brand placement would have to comply
with sponsorship-identification requirements stated in Section 3 1 7 of the
Communications Act and enforced by the Federal Communications Commission. Any paid promotion must be specifically identified to the audience at the beginning and end of 23
any program over five minutes in length (Balasubramanian 1994). Moreover, broadcast
networks have generally adopted internal guidelines, through their ovra standards and
practices divisions, that limit brand placement in network programs; ABC reportedly bars
paid or bartered placement in any show (Snyder 1992). However, some exceptions exist,
hi a vaguely worded proviso, the Commimications Act states that no identification is
required when the exposure of a sponsored brand in a broadcast program is not "beyond
an identification reasonably related to the use of such service or property in the
broadcast" (Balasubramanian 1994). There is likewise no identification required for
feature films (films produced for theater showings) that are subsequently broadcast on
television. Other media, including movies, enjoy a higher level of First Amendment
protection than do broadcast media and cable; there is no identification provision for these media.
For some, the increasing use of brand placement is cause for concern. Because these brand appearances are embedded in (and inseparable from) program content, and since viewers are rarely informed when such brand exposures arise from paid efforts of advertisers, some call the practice deceptive. The reaction from some consumer activist groups has been heated during the 1990s. The Center for the Study of Commercialism
(CSC) has led a public fight against the practice, call placement a "powerful and deceitful advertising technique because consumers aren't told they're being advertised to" (The
Toronto Star 1991). Some researchers, as well, suggest placements may carry incremental impact over other promotional tools because audiences (particularly movie audiences) presumably do not consider themselves to be in an advertising environment and thus may be more susceptible to advertising appeals (Balasubramanian 1991). 24
Solomon and Englis (1994b) similarly surmise, "Product placement strategies pair products with attractive characters who may explicitly (through behavioral or verbal
means) or implicitly (when the product is simply part of the set) endorse a product. Since the audience perceives the movie as entertainment and not persuasion, they may not
generate counterarguments and thus the persuasive impact of the communication may be
enhanced" (p. 60). The CSC has not advocated a ban on placement, but rather has asked the Federal Trade Commission to require mandatory warnings for movie audiences when placements exist (Snyder 1992).
Any efforts to use the law to curb placement are sure to be controversial. Two reviews of brand placement in the context of the U.S. Supreme Court's decisions on commercial speech come to very different conclusions. Snyder (1992) maintained that,
for several reasons, placement in films is not commercial speech: the "speaker" is a movie studio which is primarily trying to produce noncommercial speech; films, in their entirety, do not propose a commercial transaction; films are not speech related solely to the economic interests of the speaker; films overall do not refer to a single product; and, in cases where films are "mixed" speech that intertwine product appearances (commercial speech) with a story (noncommercial speech), they simply reflect the commercialization of everyday America. Snyder concluded that movies retain their status as forms of noncommercial speech even when they contain paid brand appearances—the placed brands are transformed by the context of the movie to become an intrinsic part of a creative work.
On the other hand. Lackey (1993) argued that placement constitutes a form of commercial speech and could thus be restricted. He argued that the Court has recently 25
included "mixed" commercial and noncommercial speech (such as a movie that includes
brand placement) in its definition of speech that can serve to propose a commercial
transaction. Further, Lackey argued, placement is economically motivated by the
advertiser placing its brand. Addressing Snyder's (1992) contention that placed brands
are inextricably intertwined with a film for purposes of realism. Lackey (1993) counters,
"Cinematic realism is not a 'law of man or of nature' that makes it impossible to sell
products without producing movies, or to produce movies without placing products" (p.
285).
Does Creativity Suffer?
The trend toward including real-life products within program content continues to blur the lines between advertising, entertainment, and popular culture. "Marketing communication strategies are changing the fundamental process of how consumers apprehend social reality, especially insofar as they diminish the ability to distinguish
between editorial and commercial sources of information" (Solomon and Englis 1 994a, p.
2). For example, Guess? Inc.'s Guess Journal and Benetton Group's Colors are both publications produced in-house and featuring cover prices, mastheads, and bylines—just like independent magazines. Most "stories" feature the sponsor's brand (Kuntz, Weber,
and Dawley 1996). Increasingly, the marketing communication environment is moving from one dominated by discrete, easily identifiable commercial sources of information to one in which consumers are surrounded by marketplace information from a variety of sources. 26
Miller (1990) is among the critics who believe moviemaking has suffered from the practice of brand placement. Because producers often send scripts to placement
agents before beginning filming and may in some cases modify films around brands, the
"basic decisions of filmmaking are now often made, indirectly, by the advertisers, who
are focused not on a movie's narrative integrity but only on its viability as a means of
pushing products" (p. 198). Further, while movies are a haven for those firms paying to feature their brands, they are likewise skewed to portray competing brands in a negative light. In the film Missing, in which an American father searches Chile for his kidnapped son, "this haggard, loyal dad, while talking things out, takes rare (and noticeable) solace
in a bottle of Coke~whereas inside the nightmare stadium where the army does its
torturing and murdering there stands a mammoth Pepsi machine" (p. 191).
In at least one case, brand placement drove alterations in a movie according to distribution channels. In the domestically released version of Demolition Man, Taco Bell is, in the year 2032, the only fast-food chain on the planet. A Pizza Hut restaurant was substituted for overseas distribution, since Pizza Hut has a far more extensive overseas franchise network (the chains were co-owned); Pizza Hut paid the cost of re-editing the
film, including re-recording dialogue (Kuntzman 1996). Said one film crific, "From a
marketing standpoint, it makes sense. From any other standpoint, it's pretty deplorable"
(King 1993).
The practice of placement itself has been used as a creative lever. In the film
Wayne 's World, the lead character is urged by a producer to make changes in his local cable-access television show. But, Wayne, while snacking from a Pizza Hut box, announces he will "not bow to any sponsor." Wayne and his sidekick Garth then 27
conspicuously consume several well-known brands, including Reebok, Doritos, and a
Nuprin washed down by Pepsi {Adweek 1992). Nonetheless, many critics believe the
storytelling process itself has bowed to many sponsors, diminishing the value of creative
enterprises.
Vulnerable Audiences and Special Product Categories
Rather than decry all instances of brand placement, from the vivid to the banal,
many preserve their concern for special cases. Two such cases involve placement
directed toward children and the display of risky product types such as alcohol and
tobacco. While there is no published research on the impact of brand placement on
children, several studies have examined smoking and alcohol use in movies and
television shows.
A sample of daytime television dramas taken in 1991 found increased incidence
of alcohol consumption and display than had been the case five years earlier (Diener
1993). Tobacco consumption rates were significant but unchanged in the samples from
1986 to 1991 . Similarly, in popular movies, cigarette usage has not reflected changes in
society. A sample of top-grossing American films from 1960 to 1990 found no change in
the overall rate of tobacco use (Kazan, Lipton, and Glantz 1994). Further, the incidence
of smoking in these movies was nearly three times that of actual population data; the
authors cite these data as important because they believe "movies both shape and reflect
social values regarding tobacco use" (p. 998).
Of 18 movies showing nationwide in late 1996, all but one contained at least one incident of smoking; in 13 of the films, major characters smoke (Thomas 1996a). 28
Smoking in films is generally not brand-specific, however. A study by the American
Lung Association of films fi-om 1994-1995 found that 77% showed tobacco use in some
form but only 20% of the 103 films showed specific brands (Thomas 1996b).
Particularly troublesome to many are reports that tobacco firms have made major
efforts to place their brands in films, including movies popular with younger audiences.
Brown & Williamson Tobacco Corp., for example, reportedly spent $1 million to place
its cigarettes in movies including Where the Boys Are, Rocky IV, and Rhinestone Cowboy
(Adamson 1996). In 1990, 13 tobacco firms publicly stated their intention to cease paid
placements, as part of a voluntary move to curb tobacco marketing to youths, though
bartered or unpaid placement was not affected (Colford 1990). Today, some studios
refiise to put branded cigarettes in their films for fear they will be perceived as supporting
smoking among teens (Turcotte 1995).
Prosocial Uses of Brand Placement
There might be prosocial and/or nonbrand uses of in-program depictions as well.
Some industry leaders have explicitly called for prosocial uses of brand placement
(Advertising Age 1991). Balasubramanian (1991) pointed out the potential of in-program
depictions in serving socially usefial purposes. The most successful demonstration of this to date, the Harvard Alcohol Project, was an integrated agenda in which television writers were part of an effort to show socially responsible alcohol consumption in network programming. In particular, the scripts of many prime-time television shows were written to include the use of designated drivers. Within one year, there was a reported ten point 29
increase in the percentage of consumers who claimed to have used a designated driver
(Solomon and Englis 1994b).
Audience Processing of Drama
A general finding in studies of reactions to self-presentational tactics is that audiences are receptive to the images claimed by everyday social actors. "Across many studies, perceivers are quite consistent in their tendency to judge others as basically truthful and to believe the feelings or attitudes that others are trying to convey rather than
those that they really do hold" (DePaulo 1992, p. 220). In the particular environment of
motion pictures or other media programs, there is a further assiunption of audiences'
"willing suspension of disbelief in their desire to be swept up in the story (Coleridge
1817, quoted in Solomon and Englis 1994). This is a ripe setting for the display, and acceptance, of particular images and identities.
Deighton, Romer, and McQueen (1989) argue that the use of drama in advertising
and brand promotion is both potentially powerful and evocative of a different type of
information processing than that from a more reasoned argumentation. "When a drama is successful, the audience becomes 'lost' in the story and experiences the concerns and
feelings of the characters. When an argument is successful, the audience weighs the
evidence and then yields to it" (p. 335). They further argue that, in the case of drama advertising, the degree of realism perceived by the audience increases empathy. Wells
(1989) differentiates "lecture" (a direct address from the speaker to the audience) from
"drama" (an indirect address to the audience, where characters speak to one another) in 30
the context of television advertising. He likewise asserts that dramatic presentation involves a unique processing style from the audience. "Viewers learn from dramas by observation and inference, just as they learn from stories they hear and from other things
that happen to them every day . . . when a drama rings true, the viewer migrates into the
drama, draws conclusions from it, and applies those conclusions to the management of
life" (p. 15).
To this point, it has been argued that several general processes involved in claiming an identity or forming an impression of others are likely to apply to symbolizing efforts in media programs. However, we must acknowledge there are also substantial differences between responses to media characters and impression formation in interpersonal settings. First, in a mass media situation, audiences have no expectation of actually meeting and interacting with the characters they perceive. In interpersonal settings, audiences do interact with social actors. Second, the information available to
audiences in mass media and interpersonal settings is typically different. Through media characterizations, an audience may be exposed to a character's private thoughts or be able
to see how a character behaves when he or she is alone, for example; such opportunities are not available in interpersonal situations. Third, media characterizations are scripted, edited, and constructed in quite systematic ways that are designed to produce a particular
impression efficiently. Our exposure to others' behavior in interpersonal settings is less structured and often unplanned (Hoffner and Cantor 1991).
Despite these differences, though, Hoffrier and Cantor (1991) concluded in their review that "the cognitive processes involved in forming impressions of characters and
real people appear to be highly similar" (p. 65). Further, they concluded that the theories 31 . .
and findings in one area can be used to shed light on those in the other. Indeed, since
media settings are characterized by a potentially greater quantity of information (which is
more carefully structured) than is the case in interpersonal settings, impression management processes would appear to be of great importance in understanding responses to media characters and their identity characteristics.
Audience Responses to Media Characters
Hoffner and Cantor (1991) present a model describing how audiences form impressions of media characters and respond to characters' experiences. Their model
posits three major factors: (1) specific audience characteristics, which influence how available information about characters is used; (2) the sources of information available to audiences about media characters and how those sources of information are likely to affect impressions; and (3) the implications of character perceptions for audience responses to media characters and programs. Each of these three parts of the model are discussed in turn.
Audience characteristics that affect responses include (a) prior knowledge and experience, including general knowledge about media presentations and familiarity with specific characters, (b) assumptions about the characteristics of other people, including both person categorization processes such as stereotyping and the existence of strongly held prototype information for "person types" such as intellectuals, criminals, or blue- collar workers, and (c) the individual's developmental level, which affects the importance attached to physical appearance, the extent to which characters' behaviors are well 32
understood and easily evaluated, and the skill in recognizing and responding to emotional and nonverbal cues. One characteristic particularly salient for the present research is the ease with which audiences categorize media characters according to some prototype. Not
only do audiences use prototypes in order to understand characters, but there is evidence
that program casting is done in accord with commonly held cliches or stereotypes associated with a role (Turow 1978). "In other words, characters' physical appearance, clothing, and other observable characteristics are often used as a 'short hand' approach to
character development" (Hoffner and Cantor 1991, p. 79).
A second major factor identified by Hoffner and Cantor (1991) is that program audiences have at their disposal a variety of information sources about characters.
Characters' physical appearance, including their level of attractiveness, physique, and manner of dress affect audience evaluations. Other sources include speech characteristics such as vocal qualities and language use, characters' overt behavior, gender or perceived
social role, or characters' nonverbal behaviors. Audiences' use of these information
sources is likely to be skewed in a number of ways. As discussed earlier, audiences skew
their use of trait information about others toward negative or extreme information
(Skowronski and Carlston 1 989) as well as behavior or symbols that have meaning easily inferred or retrieved (Beike and Sherman 1994).
Finally, Hoffner and Cantor (1991) differentiate the types of potential audience responses to media characters. Audiences may be attracted to characters in three interrelated ways: liking of the character, perceived similarity to the character, and desire to be like a character. The general findings of research into character attractiveness are that audiences hold more favorable evaluations toward characters who are seen to possess 33
desirable attributes and who are seen as more similar to themselves. These components
of attraction enhance the desire to be like characters, to attend more closely to character-
relevant events, and to imitate those characters' appearance and behaviors (Hoffner and
Cantor 1991). Other audience responses include emotional ties to characters, including
empathy, emotion sharing, or revulsion, and, in some cases, long-term emotional
involvement.
Audiences may also attempt to adopt and/or imitate elements from media
depictions and characters. Social learning theory (Bandura 1977) describes the general
instances when new response tendencies can be learned through observation and
modeling. Tan (1986) adopted the social learning model to explain the influence of
media programs (specifically, television shows with violent content) on subsequent
audience behavior. Where attributes of the program are positive, simple, and effective in
helping one achieve goals, those attributes are more likely to be attended to and
subsequently used by audiences.
While we may note ways in which some audience members seemingly seek to
emulate all facets of particular actors or characters in the media, it is more likely the case that audience members try to simply mirror certain attributes of those characters~the cool glamour of "James Bond," the uninhibited sexuality of Madonna, or the casual hip of
Jerry Seinfeld, for example. Brands are socially visible, readily accessible, and easily understood links to those desired characteristics. 34
Existing Research into Brand Placement
The body of research into brand placement is still rather small. Those in the
industry have done little research on placement effectiveness. Movie studios do not offer testing, and placement agencies only occasionally hire outside research firms to do testing for their clients. Only a few advertisers, such as Anheuser-Busch and AT&T, regularly conduct tests of their placement efforts (Turcotte 1995). A stream of academic research has begun to appear, however. These studies can be classified into: content analyses of brand placement in various media; surveys of practitioners; qualitative studies of how audiences interpret placements; studies of audience knowledge of and beliefs about placement; and experiments testing memory, attitude change, and/or purchase intention.
Content Analyses
Some researchers have attempted to categorize the appearance of brands from controversial product types. As mentioned above, Diener (1993) examined the use of
alcohol and tobacco in daytime TV dramas, while Hazan et al. (1994) found a much higher level of cigarette use in popular movies than exists in the general population.
Englis, Solomon, and Olafsson (1993) compared brand appearances and other examples of consumption behavior in music videos from the United States and Sweden, finding higher levels in the American sample; the authors attributed the result in part to the more frequent use of brand placement in the United States. In an analysis of the top 25
American movies fi-om 1991, Sapolsky and Kinney (1994) found automobiles and foods 1
35
as the product categories most represented. The films contained an average of over 1 brand appearances each.
A 1 993 study examined brand or company appearances in television programming through an analysis of four network stations over a 24-hour period (Fawcett 1993). Over
1 ,000 mentions were found, which represented a 1 0% increase from a comparable study three years earlier. In half of the cases, the brand or company was both shown and the name spoken; in just over a third of the cases (35%), the brand or company name was heard but not seen. The brand mentions occurred most often during news and talk shows, especially evening news (40% of cases), morning news (16%), and talk shows (15%).
The categories most represented were automobiles (18% of cases), business names
( 1 5%), and movies and plays (1 3%).
Practitioners' Assimiptions and Beliefs
One useful way of approaching the value of brand placement is through the beliefs of practitioners. Karrh (1995a) surveyed the membership of ERMA during 1994 in order to gauge practitioners' beliefs about the executional and brand characteristics judged most
likely to lead to effective placements, as well as how "effectiveness" is currently
measured. While there are strong limitations to the study (it was based upon 23 respondents, and was skewed toward the agency side of the business), some results were interesting. For instance, practitioners rated "portraying the product or service in a favorable light" as the most important executional characteristic among twelve options presented; this is consistent with Schlenker's (1980) association principle. The practitioners surveyed in the Karrh (1995a) study believed memory-based measures of 36
effectiveness—unaided recall and recognition—to be most appropriate for evaluating
placement "success."
Pardun and McKee (1996) surveyed a different group, advertising agency media
directors. Seventy percent of these media directors considered themselves extremely or
somewhat knowledgeable about placements. This respondent group (n=89) generally
predicted a larger role for placements in future media plans. Asked to list the factors
most important in making the brand placement decision for motion pictures, the
respondents listed "national viewing potential," the "price of the placement," and the
"theme of the movie" as the most important of ten factors suggested, with the "action in
the movie," "international viewing potential," and the "movie producer" as least
important.
Qualitative Explorations
Other researchers have used qualitative tools to gauge the meanings audiences
ascribe to brand placement. DeLorme, Reid, and Zimmer (1994) conducted focus groups
of college-aged frequent moviegoers and reported three emergent themes: (1) an
appreciation of realism, as participants enjoyed the realism placements can bring to
movies, noticed and disliked excessive brand exposure, and particularly disliked generic
products in movies; (2) notice of the familiar, since participants consistently mentioned the attention they gave to familiar brands in movies; and (3) relation to characters, in that
shared brands can contribute to the relationship between audiences and characters. The authors concluded, in part, that through comparing their brand usage to that of a movie character, "viewers can validate their interpretations of the character, their purchasing 37
behavior, and their own identity" (p. 14). Wooten (1995), after employing a combination of focus groups and long interviews with college students, reported eleven categories of results: a general lack of concern over brand placements; an appreciation of realism in films; a liking for subtlety; a dislike for generic or masked brands in movies; active
judgments as to the 'fit' of a brand with a scene; greater acceptance of brands within
"blockbuster" or other overtly commercial films; more liking of brand placements than of
fiill promotional tie-ins; a general view that most brands in movies are not noticed by audiences; a lack of understanding of the placement process; the view that placements rarely impact marketplace decisions; a feeling that disclaimers or warnings for audiences would be distracting for audiences; and a belief that consumers would use their own judgment to correct any marketplace excesses involving placement.
Audience Knowledge and Inferences
Studies of audiences' knowledge of and/or beliefs about placement activities help address the public policy concerns mentioned above. Balasubramanian (1991), for one, has called for research on the effects of specific information on audience attributions, asking "(i)f consumers are informed of the persuasive intent behind hybrid messages,
does this knowledge affect their processing of the messages?" (p. 24). Similarly, Kinnear and Root (1988), in their review of public policy issues relating to consumer deception, pointed out the need for "more consumer-oriented research" to determine the actual
"comprehension and perception" of promotional messages in the study of deception
issues (p. 46). 38
Some activists who oppose placements in feature films have suggested a
"remedy" of explicitly identifying brand placements to audiences. Karrh (1995b) tested
the effect on audience attributions when written descriptions of four placements in a
movie clip were provided. Not surprisingly, both recall and recognition of the four placed
brands were higher when notification was provided prior to viewing than when no
notification was given, and higher still when the notification was provided immediately
after viewing. However, subjects' perceptions of the contribution of the brands to
characters and story were unaffected by identification condition. One explanation for the
minimal attribution effects was that Karrh' s sample (college students) already had a
general prior belief that brand appearances in movies are the result of a promotional
effort. This belief, at least among younger movie viewers, is seemingly a consistent
finding. In a survey of college students, Nebenzahl and Secunda (1991) found most
respondents did not object to brand placement, were "tired of traditional commercials,"
and would prefer less obtrusive advertising tools such as placements. Fewer than one-
fourth of the respondents took an ethical stance against placements.
Experimental Studies
Several experimental studies have involved the incremental impact of brand
placements; these studies are summarized in Table 2-1 . The results can be categorized as memory-based effects, attitude-based effects, and effects on reported purchase intention. 39
Table 2-1 Prior Experimental Studies on Placement Effects
Author(s) Topic Method Ind. Var. Dep. Var. Results
Babin and placement two-group exposure to brand salience, increased salience Carder (1995) effectiveness posttest-only Rocky III brand attitude for about 1/4 of (n=108) (in entirety) brands tested; no attitudinal effects
Baker and placement one-group exposure to recall, purch. int., high aided and
Crawford ( 1 995 ) effectiveness posttest-only; Wayne 's World attitudes toward unaided recall,
students (n=43) practice 16% higher p.i. for placed brands over reported favorite
Karrh (1995b) impact of three-group manipulation of recall, recall and recog. placement experiment identification recognition, highest under identification (n=260) (none; prior to attributions ident. after on memory, viewing; after of placement viewing; most attributions viewing) intent attributional items unaffected
Karrh (1994) placement experimental exposure to brand salience, among five placed effectiveness and control clip of Raising brand evaluation brands, one
groups (n=76) Arizona (incl. showed salience five brands) effect; salience
sig. when controlling for product category familiarity
Ong and Meri placement two-group exposure to recall, purch. int., positive attitudes (1994) effectiveness posttest-only; Falling Down attitudes toward toward practice; field study or Point ofNo practice, brand no impact of (n=74) Return attitude familiarity on
evals. or p.i.;
recall rates 5-78%
Sabherwal impact of two-group verbal versus recall, brand- sig. impact on all
et al. (1994) visual and/or experiment visual encoding differentiating three DVs in verbal (n=62), no (visual; both); thoughts. visual+verbal presentation control. clip of Days of complexity of condition posttest-only Thunder thoughts
* Vollmers and placement experimental exposure to two recall, brand only brand recall Mizerski (1994) effectiveness and control movie clips attitude, attitude affected groups (n=71) toward paired actor
Steortz(1987) placement one group exposure to day-after recall mean 38% recall effectiveness posttest-only; one of six across films and field study movies brands (n=304) 40
In general, the studies-all using feature films or clips of films as the experimental
stimulus-have found mild effects on audience memory for the placed brand. Audience
recall of placed brands is the measure most widely used and cited by practitioners to
represent brand placement effectiveness. Several studies have focused on placement
recall as well. One field study, employing telephone interviews of 304 respondents who
had seen a feature film the day before, found 38% correct day-after recall across 29
brands from six feature films (Steortz 1987). Ong and Meri (1994) used intercepts
outside of a theater to gain more immediate recall measures, asking 75 respondents about
brands seen in Point ofNo Return or Falling Down. In the case of Falling Down, a
familiar brand (Coke) was prominently featured; over 78% of respondents correctly
recalled the brand. Other brand appearances were recalled in rates ranging from 4-40%.
Eleven percent of respondents recalled no brands from the movie they had just seen. In a
study using 43 U.K. students as respondents, Baker and Crawford (1995) had the group
watch Wayne 's World in its entirety (a film that addressed placement in a tongue-m-cheek
fashion, as discussed above). Unaided recall rates were 90% for Pepsi, 65% for Pizza
Hut, 63% for Reebok, and 45% for Doritos. Only one of the 43 respondents recalled
Nuprin, a brand unfamiliar in the UK. Aided recall rates ranged from 100% for Reebok
to 48% for Nuprin.
Babin and Carder (1996) focused on brand recognition, through a study of 98
undergraduates who viewed either Rocky III or Rocky Fin its entirety. Both films had 36
brand appearances. Eighteen of the brands in Rocky III and 12 of the brands in Rocky V had recognition rates of over 30%, a cutoff that some professionals use to mark a placement as successfiil (Steortz 1987). 41
While the studies mentioned above measured recall from movies seen in their
entirety, some others have used a movie clip as a stimulus (where, other things equal, we
might expect higher memorability for placed brands). Karrh (1994) foimd an impact on
brand salience~the memorability of a brand relative to other brands in its product
category-from some placements seen in a half-hour clip of Raising Arizona. Prior
familiarity with a relevant product category moderated the salience-raising impact of
placement exposure, as only brands from relatively unfamiliar categories became more
memorable. Using a ten-minute clip from Days of Thunder, Sabherwal, Pokrywczynski,
and Griffin (1994) tested unaided recall and brand associations between those subjects
exposed to only visual placement (a logo for the Hardee's fast-food chain) versus those
exposed to the visual plus a verbal mention. Recall of the Hardee's logo was 65% in the visual-plus-verbal condition and 43% in the visual-only condition.
Tests of attitude change have revealed no impact from placements. In a study of audience reaction to very short clips of Gorillas in the Mist and Mr. and Mrs. Bridge,
Vollmers and Mizerski (1994) found no differences in either attitude toward the brand
(Ab) or attitude toward a paired actor between experimental and control groups.
Similarly, Karrh (1994) found no changes in the evaluation of placed brands, even when those brands were made more memorable.
Finally, there is some evidence of placement impact on short-term purchase intention. Baker and Crawford (1995) found evidence that brand appearances in a feature film may impact short-term purchase intention. In their study of 43 students who viewed
Wayne 's World in its entirety. Baker and Crawford (1995) found a 16% higher reported purchase intention for placed brands over the brands respondents had previously noted as 42
their "favorites." This resuh is interesting in that it ties brand placement more directly to marketplace activity.
Ron (1996) had undergraduate subjects compete for prizes by playing a video game. The study manipulated brand inclusion (the experimental condition had a game based on Pizza Hut, while the control condition involved banana bread) and involvement
(based upon the value of a prize for completing the game successfully). No difference was found between the groups in brand attitude or purchase intention for Pizza Hut.
Subjects in the higher-involvement conditions generated more favorable thoughts, and those thoughts were more strongly correlated with attitudes toward the game.
Overall, then, experimental studies have found consistent but mild effects on audience memory from brand placement. In addition, some studies report effects on purchase intention while none have found significant impact on audience attitudes toward placed brands or the programs themselves. Perhaps we should not be surprised, since placements typically involve only a few exposures (or even a single exposure) to familiar
brands; ceiling effects might limit significant findings. Still, the results seen in placement studies contrast with the beliefs of many practitioners that true "home runs" can and do occur through brand placement. Different research approaches may lead to greater insight into brand, audience, and/or placement characteristics that produce stronger
communication effects. The present study posits that a brand placement is most likely to
be effective when it provides a clear and attainable link to desired identity
characteristic(s) displayed by a media character. When such a link is not present, the brand appearance becomes another piece of visual clutter—either drawing some fairly
brief attention (in which case some memory effects might occur) or none at all. Ubiquity 43
in placement, like ubiquity in other advertising formats, is not likely to be an effective
strategy.
Identity and the Self
This dissertation addresses the creation and maintenance of identity in media
programs through the strategic use of brands; it is primarily concerned with audiences'
subseuent inferences as to how those brands might be used to create and maintain a
positive identity for themselves. Therefore, a useful starting point is a discussion of identity, including the functions of identity for both the individual and his or her intended audience.
Through an identity, an individual becomes part of the social matrix of interpersonal relationships (Schlenker 1980). Thus, the establishment and maintenance of identity is a necessary and usefiil part of social life. The perception of one's identity may influence both the individual's effectiveness in social interactions and the way the individual views himself or herself People claim specific identities in order to gain expected social rewards or to avoid social sanctions, or both. Identities are considered more desirable for an individual when they are beneficial and consistent with one's beliefs about themselves, or accurate (Schlenker and Weigold 1992). The control of information-intentional or not--about the identities of individuals lies at the center of social and interpersonal processes. 44
The Importance of Identity
Identity creation and maintenance serves important mediational functions in social
processes. Identities also perform important functions for the individual social actor.
Individuals' actions tend to be guided by the desire to minimize punishments and
maximize rewards.
People often seek to protect their identities from negative actions and events by
assigning cause to external forces. The tools used to deflect threats to one's desired
identity include lessening one's apparent responsibility for negative occurrences,
reframing performances, and directing internal responsibility to the outside (Snyder
1985). Identities also can help in securing social benefits; others will seek to be
associated with people who present a positive image through their apparent competence,
grace, humor, or other positively valenced attributes.
Identities are not only for external audiences, however. The internal audience for
an identity reflects the individuals' values and standards, though a perception of the
external audience's standards will be incorporated as well. A wealth of research has
supported a general motivation for the individual to preserve self-esteem; as for identities,
it is a common assumption that most people actively seek to preserve positive ones
(Snyder 1985).
The self-view is also important in the processing of information from the social
environment. There is evidence that individuals' cognitive systems operate according to
a "self-verification" principle, through which individuals' responses to feedback on themselves are based on a (necessarily subjective) judgment as to whether this feedback 45
is consistent with their self-view (Swann, Griffin, Predmore, and Gaines 1987). Identities
thus are both filters of information from the envirormient and "faces" through which
social benefits and costs accrue to the individual. They represent the way we see
ourselves as well as they way others see us.
Identity and Impression Management
Definitions
A number of terms have arisen to describe the process of interpersonal
information control. Some terms are incorrectly used interchangeably, so it is useful to
clarify them first. Self-identification is the process or result of showing oneself to be
some particular type of person, thereby fixing and expressing an identity for both oneself and others; impression regulation subsumes self-identification by referring to the goal- directed activity of controlling information about some object or event, which may
include the self (Schlenker and Weigold 1992). Self-presentation is the activity of regulating identity about oneself primarily for real or imagined others; impression management (IM) subsumes self-presentation by also considering those activities of information regulation that are not directly self-relevant (Schlenker and Weigold 1992).
Research into IM has increased dramatically in depth and scope in recent years as social psychologists have recognized impression management as a fiindamental part of interpersonal processes. 46
Historical Roots
Widely regarded as the inspiration for IM research is Erving Goffman. Goffman's
1 959 book The Presentation ofSelf in Everyday Life stimulated a tradition of research
that has grown in scope and sophistication since the topic was introduced (Arkin and
Shepperd 1990). Goffman was influenced by the sociological perspective known as
symbolic interactionism, which had its own roots in the philosophy of pragmatism.
According to the pragmatist view, truth is not absolute but rather relative to the needs and
interests of organisms; pragmatists believe "the reality of the world is not merely
something that is 'out there' waiting to be discovered by us, but is actively created as we
act in and toward the world" (Hewitt 1991, p. 9). In tracing the development of symbolic
interactionist thought, Hewitt (1991) centers on the work of George Herbert Mead. Mead
argued that the mind and behavior are inescapably linked, and, further, that himian
behavior is socially coordinated. Mead departed from pure behaviorism, taking the view
that both internal, mental events and the social nature of human life are necessary to
explain human intelligence. Symbolic interactionists further assert that the most
important gestures for people are fundamentally linguistic ones, including complex acts of communication that precede and accompany overt behavior. These acts of speech- referred to as "significant symbols"~have the important property of arousing the same or
similar response in the one using them as they do in those to whom the symbols are directed. The significant symbol, then, "not only affords humans a degree of control over their own conduct that other animals do not possess but also gives them a form of
consciousness not found elsewhere: consciousness of self (Hewitt 1991, p. 12). 47
Dramaturgical Links
Goffman (1959) asserted that to be a human in everyday social interactions is to
somehow be involved in a life of marked similarity to the stage; social life can be
understood as a series of performances. In other words, one's activities of impression
regulation may be thought of as the adoption of a role; others may be thought of as an
audience to be won over. The model does not assume that such role-playing is
necessarily deceitful or contrived; social "actors" may simply make different parts of their
"true" selves more salient to different audiences. People may often edit information
about themselves according to their goals and the expectations, values, and abilities of the
audience (Schlenker and Weigold 1992). Further, an explicit recognition of this social
role-playing on the part of the actor is not necessary. This model of self-definition and
behavior is useful in explanation, regardless of whether individuals explicitly see
themselves or their behavior in such a dramaturgical fashion (Messinger, Sampson, and
Towne 1990). Arkin and Shepperd (1990) cite William James as having foreseen the
analysis of self-presentational behavior over a century ago. James (1 890) stated that a
social actor "has as many different social selves as there are distinct groups of persons
about whose opinions he cares" (Vol. 1, p. 294).
The Value for Social Actors
Cooley (1909) stated that an individual's self-concept is largely a result of social interactions. His idea of "the looking-glass self asserts that one's sense of self comes from perceptions or imaginative processes during interaction with others. Cooley 48
proposed that the self-concept develops over time, as people infer characteristics about
themselves through the actions of others toward them. Further, people may (either
consciously or not) affect the actions of others~and their sense of self-through their own
behaviors. The self, then, is primarily a reflected self composed of three elements: the
imagination of our appearance to the other person(s); the imagination of that person's judgment of our appearance; and some sort of affective result such as pride or shame
(Cooley 1909).
The current view of self-identification processes recognizes that these activities
are dynamic and goal-directed. Self-identifications may be viewed as socially constructed transactions involving the actor, audience, and situation. In order to be desirable, self- identifications must be believable—considered reasonably accurate by the audience given available evidence, or at least perceived to be so by the actor. They must also be beneficial; the self-identification must be regarded by the actor as helping him or her to achieve relevant goals. For example, identifications might not always be purely positive
ones. People may adopt a self-effacing manner, for example, if that leads to (or is believed to lead to) a particular goal.
Wicklund and Gollwitzer (1982) make a distinction between two types of goals.
Self-defining goals are those which require some sort of social acknowledgment of status in order to be satisfied (such as being recognized as an accomplished musician or a loving parent). Objective, or non-self-defining, goals are those which have one clear criterion by which the individual knows whether the goal has been attained (such as breaking 80 in a
round of golf or assembling a piece of furniture). Non-self-defining goals may still have a social component—as in a case where one seeks feedback as to whether a piece of 49
furniture has been assembled adequately. There is an important distinction between the
two types of goals. Self-defining goals have multiple, social criteria for goal attainment;
in the case of non-self-defining goals, the goal comes to a close when the goal is attained.
Self-defining goals require public acknowledgment of some status or quality. The claim
to possessing that status or quality does not go away, however, when the acknowledgment
is achieved—rather, "the person's claim to possessing the self-defining quality persists"
(Wicklund and GoUwitzer 1 982, p. 6). This means the pursuit of self-defining goals is an
ongoing process with multiple criteria available for helping the actor achieve a sense of
success.
The social claim made to an identity is an assertion as to how a person washes to
be considered and treated. That claim may be the adoption of a new or different identity
or the reinforcement of an existing identity. We thus have a world in which people may
influence the way they are perceived by others through the use of behavior and symbolic
information. How might individuals display information about the way(s) in which they
wish to be viewed by others? According to Schlenker (1980), people shape self-relevant
information through a principle of personal association with desirable images and the
avoidance of association with undesirable images. This "association principle" thus links
individual and image.
Claims to an identity may be made directly, with the presentation of information
about one's own attributes and accomplishments, or indirectly, through the presentation of similar information about one's associates. An example of association applied to
indirect claims is the phenomenon of "basking in reflected glory" of others (Cialdini,
Finch, and De Nicholas 1990). Cialdini and his colleagues cite the same associational 50
mechanism in explaining why this phenomenon occurs. They claim simple links are
made by audiences and that social rewards come about from those links—people do
"systematically seek to associate themselves with positive rather than negative entities in
the eyes of observers because these observers tend to see associated objects as alike" (p.
198). Buss and Briggs (1984) concluded, "The impression management model of social
behavior is likely to be appropriate when there are clear economic rewards, when the
context is public, and when the relationship is superficial" (p. 1313). Actors are thus
likely to engage in active impression management behaviors when there are incentives for
having their desired identity publicly recognized and when there are few other sources of
information available to the audience about the actor.
Wicklund and Gollwitzer (1982) describe a process of "symbolic self-
completion," through which individuals use a variety of means to present idealized
versions of themselves. The motivation for this arises because people often find
themselves in a condition of perceived incompleteness, where they feel they have not
adequately presented symbolic evidence of their desired self Further, as was discussed earlier, self-defining goals have multiple symbolic ways of being characterized.
Wicklund and Gollwitzer (1982) propose that the failure to possess a symbolic indicator of this idealized self leads to compensatory display of other indicators. For example, should a young physician perceive she is not given full credence in her role as a respected medical practitioner, she might augment her presentation to include wearing a white
jacket, carrying a stethoscope (even when it is not needed), and displaying medical credentials in her office. 51
In their review of impression management studies, Leary and Kowalski (1990)
present a two-component model of impression management processes for the social actor.
The first component, impression motivation, concerns the degree to which a person is
motivated to influence the way in which others perceive him or her. Impression
motivation is affected by the relevance of the impression to one's goals, the value of
desired outcomes, and the current gap between the person's current image and the one he
or she wants to convey; each of these has a positive relationship to motivation, so that
higher judged relevance, more desired outcomes, and a wider perceived image gap all
contribute to higher impression motivation levels. The second component, impression
construction, addresses the types of identities actors attempt to create and why people
adopt one IM tactic over another. The impression construction process is affected by the
person's self-concept, desired and undesired identities, role constraints, the audience's
values, and the person's perception of how he or she is currently regarded.
The Value for Audiences
The process of identity creation and reinforcement, including the strategic use of
symbolic information, is of value to social actors. Information as to an individual's status
and desired identity is important to audiences as well.
Audiences use both verbal and nonverbal information for practical and usefiil
reasons. Information about an individual frames a given social situation, letting others
know what to expect from him or her (Goffman 1959). Therefore, by giving indications about an individual's identity, socially available information reduces uncertainty for the audience. Because "audiences virtually never have access to an actor's private beliefs" 52 ' ;
(Schlenker and Leary 1982, p. 91), audiences must generally rely on available public
information, including behavior and symbols, in order to make accurate inferences about
others.
Audiences need not be visible to the actor or even real; imagined audiences,
which could range from a generalized other to specific and important others (like parents,
siblings, or admired heroes) may also affect the behavior of actors (Schlenker 1980).
Audiences do impose constraints on actors, however. The information in a
communication must be in line with the audience's knowledge and value systems, using
language, symbols, and evidence that is both easily understood and readily accepted
(Schlenker and Weigold 1992). Therefore, actors must consider and adapt to the
characteristics of their intended audience in order to convey identity effectively.
Audiences may follow any of several processes in forming impressions of social
actors. One such process, relevant to the present study, is the use of trait and behavioral
information to form an impression of an individual. Several recent models of this
inference process suggest that two stages might be followed by an audience in assigning
traits to the individual. In general, the initial stage is rather automatic, based upon trait
beliefs stored in memory, and is difficult to disrupt or circumvent. The second stage is more conscious, controlled, demanding of resources, and considerate of situational
factors' impact of the actor's behavior. When the audience is cognitively busy, the first stage of more automatic inference will occur but the more conscious stage may not- meaning the audiences will be biased toward dispositional inferences (Beike and Sherman
1994). In many cases, then—especially those cases in which audiences are processing 53
many stimuli—people will give great weight to verbal or nonverbal behaviors based solely
on what those behaviors imply on face value.
Impression management processes can vary along six dimensions as identified by
Schlenker and Weigold (1992): (1) the amount of cognitive effort spent by the social
actor in presenting information; (2) the extent of the actor's consciousness in creating or
maintaining impressions; (3) the extent to which the actor's behavior is perceived as
deceptive or authentic; (4) whether the behavior in question is relatively automatic or
controlled; (5) whether the behavior is perceived as motivated by relatively pure
objectives (presenting an accurate impression) or more base ones (attempting to exploit or
fool others); and (6) the type of audience(s) to whom the behavior is targeted.
Social actors do not have free reign to establish themselves in others' eyes as they would so choose; there are often constraints on an actor's effectiveness in conveying a desired identity. According to DePaulo (1992), there may be constraints on the formation of intentions (such as cultural and situational norms) or constraints on the production of nonverbal behavior from intentions (such as ability, practice, experience, the controllability of some nonverbal behaviors, personal style, and motivational and emotional constraints).
In many cases, though, identity characteristics can be effectively communicated to relevant audiences—and the means of doing so need not be explicit. In a review of studies on nonverbal expression and impression management, DePaulo (1992) concluded that the nonverbal behaviors of individuals can be effective in conveying identity characteristics
to audiences. Further, the effectiveness of the communication is enhanced when the desired image to be displayed is closer to the image social actors hold of themselves. 54
Therefore, while deception might be a concern for some observers, there is nothing about
impression management that is inherently false or deceptive (Schlenker and Weigold
1 992). Some still criticize impression management models as supposedly emphasizing
false impressions—Buss and Briggs (1984) use the term "pretense" to describe what they
see as the deliberate attempt of people to portray characters at odds with reality—but the
identities people project are often quite consistent with their perceptions of themselves
(Leary and Kowalski 1990). Overall, "people can succeed in claiming nonverbally many,
though not all, images and. . . they are best at claiming those identities that are closer to
their 'true selves,' as they perceive them" (DePaulo 1992, p. 235). Nonverbals are thus a rich and more or less accurate source of information about individuals to audiences.
The Role of Symbols
Symbols represent an important type of nonverbal communication tool. Solomon
(1983) defines a symbol as a stimulus with a learned meaning and value. The
presentation of symbolic information is often based on a belief that the symbol(s) displayed will prompt an appropriate response from relevant others. "The purpose of (a)
symbol as it has evolved is to stir up a readiness in the commimity to respond to the
symbol...the symbol is effective as long as it causes the community to acknowledge the
person's self-definition" (Wicklund and Gollwitzer 1982, p. 5). So, by Wicklund and
Gollwitzer's definition of a self-defining goal (those that require public acknowledgment of some status or quality), symbols are, much as Cooley (1909) described, directed toward others and others' anticipated reactions. 55
One particular type of symbol, defined by its function, is used to reinforce the claim of a particular image. According to Wicklund and Gollwitzer (1982), a symbol of
completeness is a "word, gesture, behavior, or physical entity that potentially signals to
others one's self-definitional attainment" (p. 33, emphasis added) Thus, a symbol may be considered as evidence that a desired self has been achieved. The presence or absence of
proper symbols is vital to self-definition; "the construction and preservation of a self- definition depends heavily on the person's use and possession of symbols of
completeness" (Wicklund and Gollwitzer 1982, p. 33).
So, when does "completeness" occur? The individual actor attains a sense of completeness about a self-definition when symbols are instrumental in leading to community or audience acknowledgment of the actor's desired status (Wicklund and
Gollwitzer 1982). This definition implies that societal consensus and feedback are
important components of the development of self Further, it is implied that symbols must be visible or public enough so that their meaning may be effectively transmitted.
Schlenker (1980) differentiated two types of symbols according to both use and more objective characteristics. "Props" are "movable objects that affect a performance and
might project relevant symbolic information," while "scenery" is defined as symbolic
information which "consists of relatively permanent backdrops for performances" (p.
268).
Results of Self-Symbolizing Efforts
An actor's attempt at self-symbolizing will have one of three results-success, no effect, or failure (Wicklund and Gollwitzer 1982). In the case of a successful effort, a 56
more idealized version of the self is conveyed to others; "the audience is at that point
prepared to assume that the self-symbolizing individual would likely manifest any of a
number of possible symbols associated with the self-definition" (p. 169). For example, a
person successfully presenting a symbol of wealth would be assumed by his or her
audience to be both wealthy and more likely to have other indicators of wealth as well.
The self-symbolizing effort may also produce no effect. In this case, the desired
inference is not produced in the audience. The audience may see the symbolizing effort
as an instrumental act only. For example, an aspiring ruimer may wear jogging shoes to
work, hoping to link the symbols to her self-definition as an athlete, but the audience
reaction at work is that she is only trying to be comfortable. Alternatively, the audience may see the symbolizing effort as merely "acting" and thus avoid a self-definitional
inference (for example, an audience may see someone who attends graduate school and carries lots of books as "playing student").
Finally, a particular symbolizing effort may backfire. It may produce an effect, though one which is counter to the goal of the actor. In such a case, the actor may suffer
a "loss of completeness" as the audience thinks the actor is compensating for some weakness (Wicklund and Gollwitzer 1982).
These three potential results carry different implications for the audience's
interpretation of the symbolizing effort. In the first case—that of a successfiil use of symbol(s)~the audience uses the symbol as an input into the attribute(s) of the actor.
Further, the attributes conveyed are in line with the actor's goals. In the second case,
where the symbol produces no effect, the symbol is not a useful input for the audience. In the third case, where the symbol produces a negative effect, the audience sees a mismatch 57
between the symbol and known or presumed attributes of the actor. Audiences are Ukely
to consider the actor as disingenuous, attempting to use the symbol to convey a false
impression.
Characteristics of Effective Symbols
What characteristics of symbols are likely to contribute to the production of a
desired effect? Wicklund and GoUwitzer (1982) point out there is a reason for
consistency to be important to audiences— it is highly adaptive for others to be able to
comprehend that a person can or cannot perform certain behaviors. "To the extent that
anyone is going to communicate successfiilly with others, and react to them consistently,
it would be crucial that their symbols hang together in a positive way" (p. 169).
Schlenker (1980) also notes consistency as a key symbol characteristic: "inconsistencies between (a) verbal and nonverbal behavior and (b) personal appearance, props, and
scenery are Ukely to cast doubt on the validity of the entire performance" (p. 268).
Examples of inconsistent symbolization efforts might include a minister reading pornographic material or (an example from the Monty Python comedy troupe) a
supposedly macho lumberjack who professes to dressing in women's lingerie. It is true that "people recognize that such inconsistencies refute the remainder of their projected identities, and they try to keep any such inconsistent information about themselves from
coming to the attention of their usual audiences" (p. 268). Inconsistent symbolization thus is likely to lead an ineffective conveyance of identity and a negative impression of the communication effort with the audience. 58
Hoffher and Cantor (1991) extended the likely impact of consistent symbolization
into the realm of audience response to media characters. They note that while "verbal,
nonverbal, and situational cues to emotion are usually consistent in mass media
portrayals, this is not always the case" (p. 75). Further, "the judged reasons for
inconsistencies among emotional cues should have an influence on the impressions
viewers form of characters (e.g., neurotic, dishonest, sarcastic, polite)" (p. 75).
Symbols may also differ in the amount of useful information they convey in a
particular setting, relative to the presence of other symbols or other pieces of identity-
relevant information. Information about individuals differs in its diagnosticity, or its
value in distinguishing between people or types of people (Beike and Sherman 1994).
For example, negative information is more diagnostic than positive information for
inferences concerning others' morality, while positive information is more diagnostic for
inferences about ability. Extreme behaviors are judged more diagnostic than more
moderate ones for both morality and ability inferences (Skowronski and Carlston 1989,
1987). While the type of information available about others is important to judgments, so
is the amount of information available in a given enviroimient.
When other information about an attribute of the actor is present, the incremental
impact of an attribute-relevant symbol is reduced. Schlenker and Leary (1982) manipulated the level of other relevant information known by an audience through a
series of written scenarios. When information about an actor's claim was known but information about the actor's performance was not, evaluations of the actor were more directly related to the nature of the claim. When the audience knew how the actor actually performed, however, evaluations of the actor were related to the accuracy of the 59
claim. It is likely that the presence of multiple relevant cues will lessen the information
value attached to any single cue, and that those cues could include performance
information or other symbols. There is more use to the audience for a given relevant
symbol under conditions of more uncertainty about the actor's "true" characteristics.
Two characteristics of a more effective symbol in a given setting, then, are the
symbol's consistency with other pieces of relevant information and the incremental value
of the symbol relative to other environmental cues.
Brands as Symbols
The Importance of Brands
One class of symbols important in everyday social interactions is that of consumer goods. Those goods are identified by brands, which are different than "products." A product is some idea, good, service, or combination that is some element of exchange. A brand is a name, term, sign, symbol, design, or combination that a company uses to
identify its product and differentiate it from those of competitors (Bearden, Ingram, and
LaForge 1995). For instance, soft drinks and automobiles are products, while Coca-Cola,
Dr. Pepper, Pontiac, and BMW are brands within those product types.
Branding is considered important because of both the consistency and quality cues that brands communicate to consumers and the psychological benefits of a brand that might be conveyed as well. Brands are even used in defining person types-the
"Marlboro Man" has been used in advertising to typify an image of the rugged. Western male, for example, while Magnavox's motto of "smart, very smart" is intended to convey 60
an image of Magnavox television buyers as likely to make intelligent choices.
McCracken (1993) uses an anthropological perspective in arguing that a brand is "a
bundle or container of meanings" (p. 125); in his research, McCracken says the
incremental value of a brand arises from its ability to add meaning to a consumer good.
Brands are, indeed, symbolically important. In his landmark article "Symbols For
Sale," Levy (1959) pointed out the important symbolic value attached both to
consumption patterns generally and to some brands more specifically. Levy's major
premise-one likely to be even more profound now, in a period of more numerous but less
easily differentiated brands, than when the point was made almost forty years ago~was
that consumers buy things not only for their functions (what those things can do) but also
for their symbolic value (what those things mean). In an increasingly affluent
marketplace, where consumers have more resources, product and service choices are more varied, and the pursuit of bare essentials for daily life (some base level of food, clothing, and shelter) is not a concern for most, consumption itself takes on a more
". symbolic importance. As Levy (1959) wrote, . . all commercial objects have a symbolic character, and making a purchase involves an assessment-implicit or explicit-
of this symbolism, to decide whether or not it fits'' (p. 1 19, emphasis added). The idea of
"fit" was explored along dimensions of gender, symbolic age, social class, self-control
and self-indulgence, and degrees of public formality; for example, since "smoothness" is
socially understood as a more feminine concept, it is not surprising that girls have a general preference for smooth peanut butter while boys tend to prefer chunky varieties.
In each case. Levy (1959) made the point that particular brands not only serve as ways to display and reinforce one's self-concept, but also that the symbolic value of brands is 61
used by people to differentiate brands and make inferences about their users. In short,
"people use symbols to distinguish" (p. 120).
Solomon and Englis (1994b) quote Daniel Boorstin, who, is his 1967 book The
Decline ofRadicalism, observed, "The material goods that historically have been the symbols which elsewhere separated men from one another have become, under American
conditions, symbols which hold men together. . . the breakfast food we eat, the coffee we
drink, the automobile we drive to work—all these and nearly all the things we consume become thin, but not negligible, bonds with thousands of other Americans."
Brands as Communication Devices
For symbols to be usefiil evidence to the actor of a self-view and to ease social performance, as well as to be usefiil information for audiences, their implications for an actor must be broadly and consistently interpretable. Brands meet this requirement in that mass media advertising and years of visible use in the marketplace have made many
brands easily understood. Branding is not only a point of competitive differentiation, but also a usefiil process of social marking. Solomon (1983) asserts, "a significant portion of
consumption behavior is actually social behavior—and vice versa" (p. 319); he also gives some motivation behind the use of marketplace symbols by the actor in creating identity
by saying "product symbolism is often consumed by the social actor for the purpose of defining and clarifying behavior patterns associated with social roles. The consumer often relies upon the social information inherent in products to shape self-image and to
maximize the quality of role performance" (p. 320). According to Solomon (1983), people may use brand-name products in an a priori fashion, as antecedents to behavior. 62
with the motivation of a desire for role definition, resulting in enhanced role performance
or more confidently held self-attributions or self-images. People may also use brands to
satisfy need arousal, with resulting need satisfaction, brand purchase, and/or impression
management (communication) aims.
The impact on interpersonal communication, then-both to the self and to others-
is an important element to branding. The focus of the present research is simply a logical
extension of Goffman's view to media characterizations. People use symbols such as
brands in order to both infer the identity characteristics of others and display desired
identity characteristics of themselves. Brands may, as has been shown, serve as easily
understood links between others and the self Further, when others have identity
characteristics which are desired and are represented in some fashion through specific
brand use, the brand symbol may serve as a bridge for attaining those desired identity
characteristics. As Hoffner and Cantor (1991) have shown, "others" may be characters in the mass media.
Friedman (1991) summarized a number of his studies on the uses of brand names in popular literature and other media by writing, "brand names, in particular, are often used by authors for their symbolic communication value, and especially if these terms represent consumer products and services, such as automobiles and magazines, with high ratings the on psychological dimension of value expressiveness" (p. 117). Wells (1995) used a sample of television characterizations to examine gift-giving behavior. He noted the connection of television depictions to reality in the use of props: "In comedy and drama, a sense of verisimilitude is an artistic and economic asset, and producers strive 63
mightily to attain it . . . television stories are likely to include literal depictions of real
objects . . . and reproduce real-world behavior" (p. 308).
Holbrook and Grayson (1986) analyzed consumption patterns shown in the
motion picture Out ofAfrica and noted the importance of details in establishing and
conveying character identity: "... a film's central themes may resound in its use of major
consumption symbolism, but . . . much of the film's artistic meaning will also hinge on its
more minor references to symbolic consumer behavior" (p. 375). These seemingly more
minor symbolic uses are important "for the insights they provide into the development of
plot and character" (p. 376). As to consumer goods and character development, Holbrook
and Grayson (1986) further state, "consumption patterns contribute many nuances of
various characters' thoughts and feelings" (p. 379).
Audience and Marketplace Impact
The present study assumes audience members may use brands as symbols to infer identity, then translate that inference into their own roles as social actors. In analyzing patterns of product and brand use in the motion picture Out ofAfrica, Holbrook and
Grayson (1986) use an approach related to abductive inference. "In abductive inference, one reasons fi-om a rule and a result to a case. For example: elite people generally drive expensive cars (the rule); in this work of art, Jane drives a Cadillac (the result); therefore,
Jane is probably elite (the case)" (p. 375). In such a case, seeing another person (Jane)
paired with a brand that conveys status leads to an inference that Jane has status. 64
Extending this reasoning, one could also learn from the movie that driving a Cadillac
would help one meet a personal goal of conveying status to others.
As stated earlier, qualitative research (DeLorme, Reid, and Zimmer 1994; Wooten
1 995) has highlighted realism and character identification as important program viewing
experiences associated with brand placements. Further, a 1989 report fi-om a leading
advertising agency noted two particular advantages in the use of brand placement: that
"the product is part of a story, and as such attributes can be clearly translated to the
audience"; and that the "the product is used by a star-as powerful as a celebrity
endorsement but more subtle. In society's zeal to imitate stars, it's conceivable that
viewers will acknowledge and buy products used by idols on the big screen" (J. Walter
Thompson USA 1989, p. 2). The report also noted, however, that "traditional measures
of advertising success cannot be applied to product placements" (p. 3) and "the overall
evaluation must be qualitative rather than quantitative" (p. 8). However, it is the view of the present research that if relevant dimensions of brand placement and desired identity can be specified, a quantitative evaluation can be made. To date, there has been only
some speculative work advancing some of these ideas (Solomon and Englis 1994). The proposed dissertation would operationalize this general idea in a more structured and testable way.
hi those cases where brands are an integral means of displaying a media character's identity, several factors emerge fi-om prior research and current practice as likely to contribute to the brand placement's persuasive impact. Individual characteristics among audience members (or, audience characteristics) will play a role, as those with higher levels of motivation and ability to process the media depictions, as well as the 65
tools to do so, are more likely to be persuaded (Leary and Kowalski 1990). In the present
study, "motivation" will be considered particularly as the motivation to display certain
identity characteristics as well as the motivation to process information about the brand
and its potential utility as an expression of that identity. Since the identification goals
considered here are self-defining ones—and thus part of an ongoing process that is never
truly "finished" (Wicklund and Gollwitzer 1982)-we may assume that most people
maintain at least a minimal level of motivation in most social settings.
Certain characteristics of the placed brand are also likely to have an impact on the
communication effectiveness of a particular brand placement. First, as mentioned above,
any symbolic communication attempt must be in line with an audience's knowledge and
value systems in order to be understood and accepted (Schlenker and Weigold 1992).
Symbols are chosen that are meaningfiil to relevant others, in the expectation that the
symbol will be effective in prompting an appropriate response (Wicklund and Gollwitzer
1982). Brands generally meet this criterion since they are well understood as social
symbols. Of course, some brands will be more familiar and meaning-imbued than others,
and thus more likely to be effective as communication devices. Brand characteristics
such as familiarity to the relevant audience will impact upon the communication
effectiveness of the symbolizing effort. The brand's attainability for subsequent display
by that audience will also affect the marketplace impact of the brand placement, since brands that are not available or affordable cannot be appropriated for later use by the audience members.
The desiredpersonal traits to be represented through the symbolizing effort are
also particularly important. For example, when the desired image to be displayed is close 66
to that which social actors already hold of themselves, the symbolizing effort is more
effective (DePaulo 1992; Leary and Kowalski 1990). The valence of an identity attribute
matters, since people generally seek to associate themselves with desirable, more positive
images and to avoid associations with undesirable images (Schlenker 1980). The valence
of an identity attribute matters to the individual who wants to claim a certain image,
though the valence might not be interpreted in the same way by all; a young man might
find a "bad boy" image desirable even though many others would see that image as
negative. People seek to create an identity for themselves that is beneficial, producing
desirable social outcomes, as well as one that is believable fi-om the perspective of the
social actor (Schlenker 1980).
The brand/trait match will also impact upon the effectiveness of the
communication, since the particular brand placement occasion represents a window
through which audiences can make a relevant connection. When the "window" is small-
when the brand is shown only briefly and indistinctly~an audience will have little or no
opportunity to make inferences. Similarly, when the brand is not connected to a specific
character (and thus to clear identity attributes), few inferences about the brand can be
drawn. Practitioners mention the opportunity audiences have to see the brand and its
connection to desirable characters as important factors they consider in deciding upon
particular brand placement opportunities (Karrh 1995a).
Where people see a desired trait (depicted through a media character) paired with
a particular brand, they may judge that brand to have utility in conveying the relevant trait to others. For instance, if someone wants to be seen as "cool," and sees a "cool" movie character who rides a Harley-Davidson motorcycle, that person will perceive Harley- 67
Davidson as a brand that reflects "cool" to the outside world. If the desire to be cool is sufficiently strong, the individual will likely have a more positive view of Harley-
Davidson, a greater desire to buy a Harley-Davidson, and, if the brand is attainable, a
more powerful intention to acquire it. See Figure 2-1 below.
DESIRED BRAND PERSONAL TRAIT(S) CHARACTERISTICS * believable to others * brand carries meaning as * beneficial to individual social symbol * attainable (affordable and available) for individual
INDIVIDUAL'S CHARACTERISTICS * motivation to display or BRAND/TRAIT MATCH claim identity traits (BRAND PLACEMENT) * brand linked to display of desired trait(s)
UTILITY OF BRAND IN CONVEYING DESIRED TRAIT(S)
H4a,b
DESIRE FOR BRAND * attitude toward brand (Ab) * purchase intention (PI)
Figure 2-1 Conceptual Model 68
This conceptual model includes two dependent variables-attitude toward the brand (Ab)
and intention to purchase the brand (PI). Brand attitudes, and changes in brand attitudes,
are frequently used in both research and practice to gauge the effectiveness of marketing
communication efforts. Attitudes may change as a result of brand placement efforts if
program audiences change their perceptions of the brand's utility in conveying relevant
identity characteristics. Whether brand attitude changes will result in increased
marketplace activity-sales of the brand-depends upon the assumed link between
attitudes and behavior.
Fishbein and Ajzen (1975), in a model subsequently termed the "theory of
reasoned action," argued that volitional human behaviors can be predicted almost
exclusively from individual beliefs and attitudes. According to this theory, the immediate
determinant of a person's behavior is that person's intention to perform the behavior.
Intentions, in turn, are a fiinction of both the person's attitude toward the behavior and the person's subjective norm, or perceptions of the social pressures to perform (or not perform) the relavant behavior. The theory of reasoned action explains attitudes toward behaviors rather than attitudes toward objects-so the theory operates at a different conceptual level than would attitudes toward particular brands.
Both brand attitudes and intention to purchase the brands are considered as dependent variables in the present study. Brand attitudes are important to consider because they tap an effect specific to the brand and individual; they do not consider other factors, such as competing brands or resource constraints. Behavioral intentions are also important to consider because, as Fishbein and Ajzen (1975) have demonstrated, intentions play the central role in predicting voluntary behaviors. It is the behavior of 69
brand purchase that is most important to marketers who consider brand placement
activities.
The characteristics mentioned above come from research into how people go
about claiming and displaying personal identity characteristics, their motivation and
ability to learn about symbols for asserting desired identities, their use of brands as
meaningful social symbols, and how they might use identity-relevant brand cues that
appear in mass media programs for their personal benefit. Following this research, the
Model assumes people seek information about identity maintenance and display on most
occasions, and will be better at learning when they have the tools and ability to do so.
Further, people seek to display characteristics they find desirable and plausible. The
Model also reflects assumptions that brands will be more meaningfiil for communicating
identity when they are already familiar to a relevant social audience, and that a given brand placement will be more effective in communicating identity when audiences notice
the brand and link it to desirable attributes of a particular character.
Hypotheses
The hypotheses for the proposed study all concern relationships among the constructs mentioned above and illustrated in the Model.
The first hypothesis concerns the link between an individual's motivation to show one or more identity traits and the characteristics of the trait(s). People are more likely to want to show characteristics of themselves that are both believable to others and beneficial (Schlenker and Weigold 1992). Where an individual perceives a trait to be 70
both believable and beneficial, he or she will be motivated to display that trait to others.
This motivation may set the stage for an audience member to scan the environment-
including the media—for cues as to ways to display the trait(s).
HI: The believability and beneficiality of an identity trait are positively
associated with the motivation to display the trait to others.
Hypothesis Two concerns an individual audience member's motivation to claim
identity characteristics, and thus to more consciously scan the environment for cues to
help him or her do so. While we can assume a relatively involved audience for many
media programs, those individuals with a particular interest in conveying identity traits
will likely pay more attention to relevant cues. Therefore, more motivated audience
members will perceive more identity-relevant information from a given brand placement
than will less motivated audience members. -
H2: Individuals with a higher level of motivation to display identity traits
conveyed in a brand placement perceive more utility of the brand to convey the relevant
traits. =
Hypothesis Three concerns the relationship between the trait/character pairing (in
a brand placement) and the perceived utility of a particular placed brand in conveying a trait. As discussed above, brand placements may be uniquely effective communication
devices due to their power in connecting the brand to particular identity traits.
H3: Perceived links between a trait and a media character (who is paired with a brand) are positively associated with the perceived utility of the brand for conveying the
desired trait. 71
A fourth hypothesis set addresses the potential marketplace impact on a brand from the effective link to desired identity characteristics represented in a movie (or other media program). Through their use in the media, brands may come to represent more powerfully certain social characteristics and thus may become more desirable to program audiences; that enhanced desirability will be represented through a more positive brand attitude and a more powerful desire to acquire the brand.
H4a: Individuals who perceive more utility of a brand for conveying a desired identity characteristic also desire that brand to a greater extent than do other individuals, as reflected in a more positive brand attitude (Ab).
H4b: Individuals who perceive more utility of a brand for conveying a desired identity characteristic also desire that brand to a greater extent than do other individuals, as reflected in a stronger purchase intention (PI).
The Model shows more possible relationships than those which are specified as
hypotheses. In some cases, there is little support for these possible relationships in the
literature; in other cases, simple time sequence makes a link improbable. The present
study will test all possible links through a path analysis, as is discussed in Chapter 3.
However, the research reviewed above suggests some potential links are not significant.
Consider one potential link shown in the Model—that between an individual's
motivation to show identity traits to others and the desire for the relevant placed brand.
A higher level of motivation might lead to higher awareness of or memory for the brand placement (because of more goal-directed processing of the media environment), but not necessarily to brand-specific effects. As has been argued above, a brand placement should include identity-relevant information to be most useful for that audience member 72
(and thus for the placement to be most effective in terms of driving desire and perhaps
sales). So, even for individuals who are particularly motivated to display identity traits,
the context of the placement is important for driving brand-level effects. This is an important point of investigation of the present study. CHAPTER 3 METHOD
Overview of Design
The present study primarily measures (rather than manipulates) a number of constructs at the individual subject level. All subjects were exposed to a feature film containing a number of brands paired with different characters (or, alternatively, a long
clip of the film if time constraints occur). Brand-level effects, including attitude toward the brand (Ab) and purchase intention (PI), were measured both prior to and following exposure to the film, so that the incremental effect of the brand appearances in the film
may be assessed. This is the "one-group pretest-posttest design" described by Cook and
Campbell (1979).
The link between individual motivations to display identity traits and response to
a given brand placement is important. It is also particular to the individual and situation.
While the literature suggests individuals are relatively active processors of social
information, it is still likely the rare occasion when motivation is high and the individual
is exposed to a brand placement with great personal relevance.
The present research seeks to identify those brand placement occasions (though they may be rare) which lead to greater impact with program audiences. To do so, motivation could be primed with some subjects so that the motivational link could be identified. A prime would naturally amplify the incidence of such motivation as
73 74
compared to "real-world" settings, but would enable us to recognize the link. Therefore,
the present study amplifies the H2 link with a priming condition. One group of randomly
assigned subjects read instructions which are designed to serve as a motivational prime.
Subjects in this primed condition could thus be compared to subjects who have not been
primed as a test of the motivational link.
All subjects in the study saw the same stimulus. As Reeves and Geiger (1994)
discuss, there are several advantages to this approach in the study of audience responses
to media messages. First, subjects have an opportimity for comparison; differences
between more or less desirable depictions might not be apparent unless those depictions
can be compared at the same occasion. Further, where in a more real-world case
audiences are likely to see varying depictions anyway, this approach offers an advantage
related to ecological validity. Subjects are also less likely to guess the research question
(and adjust their responses), since "(t)here are so many things happening in messages that
it is hard for a subject to pinpoint the highlighted feature" (p. 175). On more practical
bases, this approach offers the enhanced statistical power inherent in within-subject
designs as well as reduced set-up time for the study. This procedure has also been used
successfully in prior brand placement studies (Karrh 1995b, 1994).
Subjects
This study involves audiences' use of brand symbols in programming and the results of different patterns of symbolizing on processing of the program and subsequent potential marketplace impact. These effects are seen as universal in scope in that they are
i 75
likely to be applicable to any group. A convenience sample of students was employed for
this study. While the use of students as subjects may often be inappropriate, "there are
good arguments that students can be used in those cases where the situation is the same or
where they have knowledge and experience comparable with that of the population being
studied" (Shuptrine 1975, p. 390).
The use of students is appropriate in this particular instance since the target
audience for much of Hollywood's production is college-aged, hi terms of movie theater
attendance, those aged 1 8-24 are considered the prime target for movie makers. Some
34% of people aged 1 8-24 report going to the movies at least once per month, compared
with a 20% rate for all U.S. adults. Those aged 1 8-24 are often more than twice as likely
as the population as a whole to have seen specific movies within the past six months.
Movie makers generally "make movies aimed to please young audiences" (Dortch 1 996,
p. 4). The 12-20 year-old age segment is currently the fastest-growing American movie
audience, with over 347 million admissions in 1996, a jump of 25% over the prior year's
total {The Wall Street Journal 1997). But, while these two published surveys of college
students found a general acceptance of brand placement as a part of feature films, it
remains an open question whether the views of college-aged samples-likely a media-
sawy group—are shared by other parts of the population.
The students recruited for this study came fi-om an introductory advertising class at the University of Florida. 76
Stimulus
Several criteria were deemed important in selecting a stimulus. First, the movie
should appeal to the college-aged subjects in this proposed study, both in its story
(whether the audience would be interested) and its characters (whether the audience
would relate to them). Second, the movie should feature several characters who differ in
the identity attributes they display as well as the relative desirability of those attributes.
Third, a number of familiar brands should be prominently shown in the movie. Finally,
those brands should vary in their attainability (likely in the form of cost) for the audience
members. This is needed for statistical variability so that conclusions could be drawn
between relatively attainable and non-attainable brands.
The stimulus for this study-which meets the criteria listed above~is the 1994
feature film Reality Bites. This romantic comedy featured an ensemble cast of yoimg
characters balancing personal relationships and the start of their professional careers. The
film stars: Winona Ryder as a television production assistant ("Lelaina Pierce") who
yearns to produce a video docimientary about her friends; Ethan Hawke as a usually
unemployed, brooding but brilliant rebel ("Troy Dyer"); Ben Stiller as a sucessful but
square and materialistic video executive ("Michael Grace") who becomes involved with
Lelaina; Janeane Garofalo as Lelaina' s roommate ("Vickie Minor"), a sexually
promiscuous manager of a Gap store who is obsessed with the 1 970s; and Steve Zahn as
"Sammy," who struggles with his own sexuality. A number of familiar brands (including
Diet Coke, Rolling Rock beer. Snickers candy bars. Gap clothing stores, and BMW, Ford, and Saab automobiles) are prominently featured in the film. The characters are also 77
shown with a mix of more or less desirable attributes. A listing of all brand appearances
in Reality Bites is presented as Appendix A.
Instrument
Subjects completed separate questionnaires prior to and following exposure to the
film clip. Table 3-1 shows the constructs tapped through the first questionnaire (Oi) and
the second questionnaire (O2):
Table 3-1 Constructs Measured in the Questionnaires
Questionnaire One (O i) Questionnaire Two (O?)
demographic characteristics: sex; liking of / attention to movie major in school; year in school; family income prior viewing of the movie (yes/no, how many times) movie use: monthly movie attendance; monthly VCR rental; favorite movie type(s) unaided recall of brands in movie attitude toward the brands (Ab) recognition of brands in movie purchase intentions (PI) perceived identity characteristics of movie characters self-view (believability) of relevant identity characteristics perceived utility of brands in conveying identity characteristics self-view (beneficiality) of relevant identity characteristics attitude toward the brands (Ab) perceived "gap" between identity purchase intentions (PI) characteristics held and those desired 78
The questionnaires are shown as Appendix B. These constructs are represented by
questionnaire items in the following ways:
attitude toward the brand (Ab) : two seven-point semantic differential scales (anchored by
good—bad and like--dislike);
purchase intention (PI): two seven-point semantic differential scales (anchored by /
definitely intend to buy--I definitely do ml intend to buy and / would buy ifI could--I
would not buy even ifI could);
self-view of identitv characteristics : one seven-point scale for each of nine identity
characteristics (anchored by describes me well—does not describe me);
importance assiRned to identitv characteristics : one seven-point scale for each of nine
identity characteristics (anchored by an important quality to show others—not an
important quality to show others);
perceived gap between characteristics desired and held: one seven-point scale for each of
nine identity characteristics (anchored by / need to show this trait to others—I do not need
to show this trait to others);
liking of the movie : two seven-point Likert scales (agreement with / liked the movie and
the characters were interesting)
attention paid to the movie : one seven-point Likert scale (agreement with the movie held my attention);
brands noticed from the movie : one open-ended question asking respondents to list "any brand-name products used in the movie"; 79
brand recognition from the movie : one seven-point scale for each of 1 1 brands (anchored
by was not in the movie--was definitely in the movie), representing both brands that
actually appeared in the movie as well as competing brands that did not appear (so that a
measure of accuracy may be computed for each respondent in each product category);
perceived identity characteristics in movie characters : one seven-point scale for each of
the five major characters along each of nine identity characteristics (anchored by
describes that character well—does not describe that character well);
perceived utility of brands in coveying identity characteristics : one seven-point scale for
each of nine identity characteristics (anchored by the brand would show others I am
Priming Manipulation
The present study posits that brand placements are more persuasive when there is a sort of intersection between a placement which conveys identity-displaying information
(a function of the context of the brand placement in the program) and audience members who are seeking brands as symbols to convey elements of their identities (a function which varies from person to person). In order to effectively make comparisons in this study, sufficient numbers of subjects should be motivated to scan the movie for identity- relevant information.
It is not clear that, in most situations, most movie audience members will be looking for identity-conveying brands. Many, if not most, audience members may be motivated by many other things (including passive entertainment). While more passive 80
processes could and often do lead to persuasion, the literature reviewed earlier suggests
that the greatest impact of a brand placement, other things equal, will come with those
audience members who are in fact more active processors of the brands in a movie or
other program. A large body of research has shown that the way in which ambiguous
behavior is interpreted and judged depends upon the construct which is most accessible at
the time information about the behavior is received (Stapel, Koomen, and van der Pligt
1997). If identity display were made more accessible for a group of subjects, just prior to
their viewing of the movie, then more subjects would interpret the movie in terms of
impression management goals. For purposes of the present study, then, a group of
subjects were primed to be aware of identity-conveying information.
Subjects in the "primed" condition had the following message included, in large
bold letters, as the last page of their first questionnaire:
"As you watch the film (clip), we ask that you keep the following in mind Imagine there are certain things about you that you would like other people to know. These other people could be goodfriends, people you see only occasionally, or even strangers.
Imagine that other people don 't know as much about your personality as you would like—
" and that you are seeking ways to show others the 'real you '.
Statistical Analvses
Two different types of statistical analyses were used to test the hypotheses. First, path analysis, a method for studying patterns of relationships among several variables, was conducted; the method has been used in several studies of media impact that involve 81
a number of associated variables (for example, Basil 1996; Wanta and Hu 1994). Path
coefficients test the strength and direction of influence fi"om one variable to another.
Where the reliabilities among sets of items are high, scales will be constructed for
purposes of analysis. Multiple regression analyses are performed, since there are multiple
dependent variables that are believed to occur in a clear temporal ordering (Basil 1996).
Then, through entering Pearson zero-order correlations, the path relationships shown in
the Model may be identified. Hypotheses One, Two, Three, and Four were tested
according to whether the relationships are statistically significant and in the direction
specified.
Second, a comparison was made of the primed group to the non-primed group,
regarding the impact of brand placements on perceptions of the brands' power to convey
relevant traits (H2). This comparison was made through tests of means, controlling for
sex, fi-equency of movie viewing, class in school, and program of study. A single-item
manipulation check was also included in the second questionnaire.
Pretest
A pretest was conducted, using a clip from Reality Bites and a prior version of the questionnaires, with a group of 36 Ithaca College communication students in the spring of
1997. The clip was of the first 25 minutes of the film, the same clip used in the ftill study. The purposes of the pretest were: to determine if college-aged respondents would understand and follow the questionnaire; to help in judging whether these respondents identified with and paid attention to this movie; and to determine whether identity 82
characteristics coded by the author from Reality Bites were likely to be perceived
similarly by respondents in the present study.
Along this last point, a set of open-ended questions were included in the pretest;
the questions asked respondents to give up to five adjectives describing each of the five
main characters in the film. Through an analysis of these responses, it was possible to
determine whether the traits believed to be associated with these characters are in fact
perceived by college students watching the film.
The questionnaire used in the pretest contained ten traits: successful; fiin-loving;
responsible; loose; honest; artistic; materialistic; intelligent; square; and confident. Only
one trait outside of these ten was mentioned by more than two respondents in the pretest.
However, on the basis of several open-ended responses, some changes were made to the
traits listed in the questionnaire now presented as Appendix B. The adjectives "honest"
and "loose" were deleted; the former was dropped because respondents did not feel it could be effectively conveyed through brand use, while the latter was a source of
confiision to several respondents. The trait "square" was changed to "cool," both to state a quality in the affirmative and to match language from the respondents. Finally,
"attractive" was added, since it was mentioned on an umprompted basis from ten
respondents. After the pretest results, then, the original list of ten identity characteristics was limited to nine.
The pretest also provided information as to the reliability of the questionnaire items. Correlations between the two measure of brand attitude (Ab) for different brands ranged from .82 to .91, suggesting these items may be reliably combined for analysis. As 83
stated earlier in the discussion of H3, purchase intention (PI) is likely to be affected by a brand's attainability (availability and cost). Not surprisingly, the correlations between the two PI items ranged from .90 for a low-cost item (Snickers) to .57 for a high-cost brand
(BMW). These pretest results confirm that Ab and PI may not be interchangeable, or even similarly sensitive to brand placement effects, across different brands and product types. CHAPTER 4 RESULTS
Overview of the Experiment and Subjects
The experiment was conducted during a single administration in December 1997.
Subjects saw the first twenty-five minutes of the film Reality Bites. One hundred fifty-
three subjects completed the experiment, all of whom were undergraduate students in an
introductory advertising course. Fifty-six percent of the subjects were female and 44%
were male. Fifty-two percent of the subjects were majors in mass communication, while
20% reported a in major business administration or accounting, 12% were liberal arts
majors, and 16% reported some other major. Fifty-seven percent of the subjects were ;
juniors, 22% were seniors, and 21% were sophomores. Most subjects (56%) reported
annual family income of $60,000 or more. Six percent of subjects reported annual family i
income of $50,000 to $59,999, 8% reported family income of $40,000 to $49,999, 3% percent reported family income of $30,000 to $39,999, 4% reported family income of
$20,000 to $29,999, and 5% reported family income below $20,000. Nineteen percent of subjects did not know or did not report annual family income.
' ' 1 Most subjects were relatively fi-equent movie viewers. When asked how often they attend movies in theaters in an average month, only 4% reported never attending movies in theaters. Forty-two percent attend movie theaters once in an average month, while attend 34% movie theaters twice per month, 14% attend movie theaters three times
84 85
per month, and 7% attend movie theaters four times or more in an average month. When
asked how often in an average month they watch rental movies on a videocassette
recorder (VCR), said never, 3% 29% said once, 34% said twice, 19% said three times,
and reported 20% watching rental movies more than three times in an average month.
Subjects were also asked which types (genres) of movies are their favorites;
subjects had the option of listing more than one type. Ninety-two percent of subjects
indicated comedies as a favorite type. Several other movie types were popular as well-
68% of subjects indicated action movies, 50% indicated dramas, 45% indicated romance
movies, indicated 45% mysteries, 25% indicated animated movies, 13% indicated
biographies, and 9% indicated musicals or concert films.
Fifty-eight percent of subjects reporting having seen the movie Reality Bites prior
to the experiment, while 42% had not seen the movie before. Of those who had seen the
movie before, had seen it once, 44% 23% had seen it twice, 9% had seen it three times,
and 24% reported seeing the movie four or more times in the past.
Most subjects reported liking and paying attention to the movie clip. On a one-to-
seven-point scale, anchored by "I strongly disagree" and "I strongly agree" respectively,
subjects tended to agree with several statements tapping interest in and liking of the movie clip: "I liked the movie" (M = 6.33, SD = 1.15); "the characters were interesting"
= = (M 6.32, SD 1.22); "the story was believable" (M = 6.15, SD = 1.23); and "the movie held my attention" (M = 6.39, SD = 1.19).
Chapter 3 outlined how several items, tapping attitude toward a brand and purchase intentions toward a brand, would be used to measure the dependent variables; 86
also outlined were the free recall item and several interval-level items that are
hypothesized to serve as predictors. The following section reports the results of the
procedures used to assess the reliability of the pairs of items used as dependent variables.
Assessing the Reliability of Measures
The dependent variables of attitude toward the brand (Ab) and purchase intention
(PI) were both measured with pairs of items. If a pair of items reliably measures the sam(
underlying construct, then responses to both items should be consistent for the same
individual. This requirement, referred to as the "internal consistency" of items designed
to measure the same construct, must be tested before the items can be reliably combined.
The Pearson product-moment correlation, commonly symbolized as r, is used to
test internal consistency; this statistic is considered a valid description of the relationship
between two sets of scores when the data represent ratio or interval measurements, the
relationship between the scores can be assumed to be linear, and the distributions of the
variables are assumed comparable (Wimmer and Dominick 1997).
The two items designed to measure attitude toward the brand (Ab) were seven- point scales. One item was anchored by "good" and "bad," while the other item was anchored by "I like it" and "I dislike it." The items were scored so that higher scores corresponded to more positive attitudes. The two items designed to measure purchase intention for the brand (PI) were also seven-point scales. The task of measuring purchase intentions for brands which ranged from the very inexpensive (for example. Snickers candy bars) to the very expensive (for example, BMW automobiles) could be 87
problematic; purchase intention likely involves both the desire to acquire a brand and the
ready resources to do so. The first item tapping PI was anchored by "I definitely intend to
buy it" and "I definitely do not intend to buy it," while the second item was anchored by
"I would buy it if I could" and "I would not buy it even if I could." It was hoped that this
second PI item, in particular, would help reflect the desire to purchase relevant brands
even when the money or opportunity to do so is not immediately at hand.
The Pearson r was calculated for the pairs of items tapping Ab and PI for each
brand during both questionnaires; Tables 4-1 and 4-2 give the results of those
calculations.
Since all pairs of measures were significantly and positively correlated, we may
assume the pairs of items meet the requirements for internal consistency. Therefore, in an
examination of the dependent variables, the pairs of items (tapping Ab and PI for each
brand) may be combined to form single measures more sensitive than are single items
alone.
Assumptions
The present study is concerned not only with the effects that exposure to brand placements may have in driving brand attitude change and enhancing purchase intention, but also with identifying the processes underlying such changes. As such, structural analytical models (such as path analysis) are well-suited for assessing these data.
Structural models are particularly usefiil for analyzing data collected in more natural settings, uncovering underlying processes in experiments, describing the pattern of . .
88
Table 4-1 Correlations Between Pairs of Items Measuring Attitude Toward the Brand (Ah)
Quest. 1 Quest. 2 orana r r Chevrolet .82 .88 Audi .68 .80 omCKerS .76 .84 iviiiKy way .81 .82 Budweiser .83 .88 Rolling Rock .85 .89 Diet Coke .89 .86 Diet Pepsi .89 .86 BMW .86 .69 Ford .82 .86 Saab .84 .85
Note: All r values significant at p<.000 1
Table 4-2 Correlations Between Pairs of Items Measuring Purchase Intention for the Brand (?])
Quest. 1 Quest. 2
Brand r r
Chevrolet .84 .8'l Audi .68 .67 Snickers .92 .92 Milky Way .94 .89 Budweiser .95 .93 Rolling Rock .86 .92 Diet Coke .97 .94 Diet Pepsi .91 .93 BMW .46 .38 Ford .82 .77 Saab .70 .75
Note: All r values significant at p<.000 1 89
interrelationships among a body of variables, and, where correctly applied, enabling
researchers to choose between competing models and explanations (Reis 1982).
Path analyses have three basic aspects. The first aspect, the path diagram, is some
pictorial representation of a system of simultaneous equations; it shows the assumed
relationships among all variables. The second aspect, the equations relating correlations
or covariances to parameters, were outlined by Sewall Wright in the early part of this
century. The third aspect is the power of path analysis to distinguish direct, indirect, and
total effects of one variable on another. Direct effects are those not mediated by any other
variable. Indirect effects operate through at least one other variable, while total effects
combine direct and indirect effects (Bollen 1989).
A path analysis is quite analogous to a series of regression analyses, and thus it
contains the same limitations as regression analysis. Regression analyses do not show
causality; structural models, even those which explain a large proportion of variance,
provide no proof of causality. However, where there are plausible competing models,
structural techniques can compare the models' ability to account for some given set of
data. From a path analysis, we may be able to validly conclude that the data are more nearly consistent with one set of constructs than with another; this permits some causal inference based upon disconfirmation (Reis 1982). Stated another way, "the purpose of path analysis is to determine if the causal inferences of a researcher are consistent with the data" (Bollen 1989, p. 38). So, while conducting a path analysis as part of the present study is entirely consistent with the goals of the study, it will not prove one explanation for the data that excludes all others. 90
The present study contained one manipulated variable. A general motivational
prime was included for approximately half the subjects, as discussed earlier. Individuals'
motivation to look for identity-relevant cues in a media program was assumed as an
important variable for understanding brand placement effects. In a completely natural
setting, however, only a small number of individuals might have a high level of
motivation at that time to look for identity-relevant cues in a movie clip. So that the
motivational link could be explored, an overt prime was included; this enables a test of
whether an overt prime would enhance subjects' awareness of their own self-
presentational goals and thus lead to deeper processing of identity-relevant brand
information. A random assignment procedure was closely followed so that the two
experimental groups (primed and control) may be assumed independent.
So that the manipulation of an external motivational prime could be checked,
subjects were asked their level of agreement with the statement, "Before I watched the
movie, I was asked to look for ways to express my personality to others." The strongest
level of agreement was scored as seven while strongest disagreement was scored as one.
This check was significant (F (1, 152) = 25.78, p < .0001). Subjects in the primed
condition had a mean score of 6.48 = (SD .95) while those in the control condition had a
mean score of 3.45 (SD = 1 .73).
Memory for the Brands
As mentioned in Chapter 2, most of the prior research into brand placement effects has centered on memory. Unaided recall, aided recall, and recognition tests have 91
all been employed as dependent measures in these studies. Audience memory for a brand
appearance is indeed important, as it may be related to the audience's depth of processing
of the brand information. In the present study, both unaided recall and recognition were
used to gauge subjects' memory for the brand appearances in the movie clip.
In the unaided recall measure, subjects were asked to take about one minute and
list any brand-name products they remembered from the movie; they were also asked not
to come back to that list at a later time. Overall, subjects correctly recalled a mean of
4. 1 brands from the movie clip. 7 Subjects in the primed condition correctly recalled a
mean of 4.40 brands, not different than the 3.95 brands correctly recalled on average by subjects in the unprimed condition (F (1, 151) = 2.3, p < .14). The recall results by brand are listed in Table 4-3.
Table 4-3 Unaided Brand Recall
No. subiects
Brand recalling Pet. recall BMW 136 88.9% Snickers 72 47.1 Ford 61 39.9 Pringles 61 39.9 Gap 58 37.9 Saab 56 36.6 Rolling Rock 51 33.3 Diet Coke 43 28.1 Domino's 41 26.8 Evian 41 26.8 Minute Maid 8 5.2 Memorex 6 3.9 Shell 4 2.6 92
After subjects were asked to spend about one minute with the unaided recall item,
they were subsequently shown a list of brands and asked to indicate, on a seven-point
scale, their level of confidence that each brand was included in the movie clip. For each
brand that appeared in the movie clip, a competing brand that did not appear in the movie
clip was also included in this brand recognition item. This afforded an opportunity to
gauge subjects' accuracy in recognizing a brand rather than simply a product or service
category (for example, separating subjects who specifically recognized BMW fi-om those
who remembered some type of foreign-made sedan and guessed at the brand). Table 4-4
shows the results of this brand recognition measure.
Table 4-4 Brand Recognition Scores
Brand Mean SD BMW (#) 6.92 0.68 Snickers (#) 5.86 1.89 Ford (#) 5.44 2.05 Saab (#) 5.22 2.14 Diet Coke (#) 5.19 1.96 Rolling Rock (#) 5.10 2.04 Budweiser 3.56 1.88 Audi 2.97 1.88 Diet Pepsi 2.72 1.64 Milky Way 2.69 1.61 Chevrolet 2.39 1.39
Note: Brands marked with (#) were shown, mentioned, and/or used in the movie clip.
Subjects were accurate in their brand recognition. The six listed brands that were actually included in the movie clip were the six most confidently recognized brands; the bottom five in recognition were not part of the movie clip. 93
Tests between competing brands also showed subjects' accuracy in recognizing
brands from the movie clip. BMW was recognized to a greater degree than was Audi (F
(1, 152) = 61 1.6, < .001); Saab also p was recognized more than was Audi (F (1, 151) =
99.4, < .001). Ford was recognized E over Chevrolet (F (1, 152) = 227.1, p < .001),
Snickers was recognized over Milky Way (F (1, 152) = 230.4, p < .001), Rolling Rock
was recognized over Budweiser (F (1, 152) = 42.64, p < .001), and Diet Coke was
recognized over Diet Pepsi (F (1, 152) = 141.6, p < .001).
The presence or absence of the prime had no effect on brand recognition. Tests of
mean recognition scores showed no significant difference between the primed and
unprimed groups for any of the eleven listed brands.
Attitudes Toward the Brands and Purchase Intentions
While subjects in the present study had a reasonably strong and discriminating memory for brands in the movie clip, it remained to be seen whether measurable, market- related effects occurred for those included brands. All brands were tested for changes in
Ab and PI from the first to the second questionnaire. Table 4-5 shows the pattern of changes in subjects' attitude toward the brands after exposure to the movie clip. 94
Table 4-5 Attitude Toward the Brand (Ah) Mean Scores. Questionnaire 1 and Questionnaire 2
Brand Ab. questionnaire 1 Ab, questionnaire 2 BMW(#) 12.73 13.49 (*) Saab(#) 9.54 11.61 (*) Ford(#) 9.29 8.35 (*) Chevrolet 9.85 7.78 (*) Audi 9.06 8.60 (*) Snickers (#) 11.12 11.68(*) Milky Way 9.58 9.04 Rolling Rock (#) 8.14 9.82 (*) Budweiser 9.48 9.16 (*) Diet Coke (#) 7.15 8.16 (*) Diet Pepsi 5.90 5.93
Notes: Brands marked with (#) were shown, mentioned, and/or used in the movie clip. Ab scores marked with (*) changed significantly (p < .01) from the furst to the second questiormaire.
As shown in Table 4-5, of the six brands in the movie clip, all reflected some
change in Ab after subjects saw the movie clip; five of the brands in the movie clip
showed a more positive brand attitude while one showed a more negative brand attitude.
Qf the five brands not in the movie clip, three showed a decrease in Ab, one showed an
increase, and one showed no change.
A similar comparison was made with the two-item purchase intention scale.
Those results are shown in Table 4-6. 95
Table 4-6
Purchase Intention Toward the Brand (?D Mean Scores. Questionnaire 1 and Questionnaire 2
Brmid PI. questionnaire 1 PI. questionnaire 2 BMW(#) 10.93 12.28 (*) Saab(#) 7.61 10.42 (*) Ford(#) 8.16 7.61 (*) Chevrolet 6.69 6.46 Audi 6.93 6.79 Snickers (#) 10.29 1 1 .07 (*) Milky Way 8.65 8.29 (*) Rolling Rock (#) 7. 1 8 9.16 (*) Budweiser 9.37 8.90 (*) Diet Coke (#) 6.46 7.42 (*) Diet Pepsi 5.24 5.41
Notes: Brands marked with (#) were shown, mentioned, and/or used in the movie clip. PI scores marked with (*) changed significantly (p < .01) from the first to the second questionnaire.
The pattern seen in brand attitudes was even more pronounced for purchase
intentions. Qf the six brands in the movie clip, all reflected a significant change in PI
after subjects saw the movie clip; five of those six brands reflected an increase in
purchase intention while one showed a decrease. Qf the five brands not included in the
movie clip, only two showed a change in PI and both of those were decreases.
Exposure in the movie thus led to changes in brand attitudes and purchase intentions for all six brands. For only one brand (Ford) was this a negative change. For the Ford brand, the context of exposure in the movie clip was brief and a bit negative. No such pattern was apparent for the brands that were not part of the movie clip—and where changes occurred, they were more often negative than positive. 96
The significant changes in Ab and PI for brands in the movie clip, coupled with a
milder and negative pattern in the same dependent variables for brands not in the movie
clip, follows the hypothesized model. If the brands in the movie clip were not evaluated
differently by subjects after exposure to the movie clip than they were prior to exposure,
then inclusion in the clip movie would have done nothing for the brand. Similarly, if the
pattern of changes in Ab and PI was the same for brands not included in the movie clip as
for those in the movie, then the changes in Ab and PI could be assumed to have occurred
from a testing effect alone.
Since a unique and largely positive contribution to the brands resulted from
inclusion in the movie clip, the analysis next moves toward an explanation of that
contribution to the brand.
Identity Traits Perceived in the Movie Characters
Implicit in the present study is the idea that not only will brand placement often
lead to a more positive evaluation of the relevant brand, but also that the change in
evaluation will be due in significant part to audience members' use of identity-relevant information surrounding the brand placement. As a first step in assessing the
hypothesized model, it must be determined whether subjects distinguished among the movie characters on the basis of traits.
Subjects rated each of the five characters named in the movie clip along nine identity traits: successfiil; responsible; artistic; fim-loving; cool; intelligent; confident; attractive; and materialistic. These nine traits were developed from a pretest described 97
above. Subjects rated each character and trait along a seven-point scale, anchored by
"does not describe that character well" and "describes that character well." Table 4-7
shows the mean rating for each character along each of the nine traits.
Table 4-7
Mean Ratings of Each Character By Identity Traits
Character "successful" "responsible" "artistic" "fun-loving" "cool 5.12'' Lelaina 6.08' 5.92' 5.69" 5.65' Troy 1.90' 1.72' 5.80' 5.04' 5.58' Michael 6.14' 5.71' 3.53'' 3.74" 4.07' Vickie 4.29' 4.47" 4.99" 6.31' 5.57' Sammy 3.26'' 3.55" 3.94' 5.42" 4.41"
Character "intelligent" "confident" "attractive" "materialistic" Lelaina 6.31" 5.20' 6.16' 2.84" Troy 5.32' 4.78" 5.31" 2.03' Michael 5.88'' 5.58' 4.67' 5.89' Vickie 5.07' 4.92" 4.20'' 3.27" Sammy 4.75" 3.88' 4.09" 3.02"
Note: Entries (in the same column) with different superscripts are significantly different at the .05 level by Scheffe procedures.
In general, subjects made clear distinctions among the characters according to
individual traits. Only one character ("Lelaina") was rated on the upper half of the scale
on all positive traits and on the lower half of the scale on the more negative trait
("materialistic"); other characters were considered a mix of more positive and more negative identity traits.
Among the five characters, "Michael" received the highest rating as being most
"successful" and "materialistic," and the lowest rating on being "f\m-loving" and "cool."
"Troy" received the lowest score among the characters on being "responsible" but was 98
also judged to be the least "materialistic." "Lelaina" was judged as the most "intelligent"
and "attractive" among the five characters. "Michael" and "Lelaina" were judged
together as highest on being most "responsible," while "Lelaina" and "Troy" were judged
to both be the most "artistic."
Rather than judging characters only along some single dimension, subjects
appeared to make finer distinctions among the characters, as different characters received
different ratings on different traits. Examined in another way, the individual traits were
not viewed as equally discriminating among the characters. For example, across the five
characters the range of mean scores by trait ranged from 4.36 for "responsible" and 4.24
for "successful" to 1.51 for "cool" and 1.56 for "intelligent."
Identitv Traits to be Appropriated bv the Audience
While subjects made important distinctions among the movie characters, it
remained to be seen whether subjects were also motivated to display certain identity
characteristics themselves. According to the Model, in the absence of such motivation,
there is little likelihood that audience members would be persuaded by the use of brands
in the movie clip. However, it was expected that subjects would have a relatively high
level of motivation to meet self-defining goals and, thus, that persuasion might occur.
On three separate items, subjects indicated their beliefs concerning each of the same nine traits that were used to describe the movie characters; subjects made ratings along seven-point scales, with higher scores indicating greater agreement. The self- description item was anchored by "does not describe me" and "describes me well." The 99
item ofjudged importance was anchored by "not an important quality to show others" and
"an important quality to show others." The need-to-show item was anchored by "I do not
need to show this trait to others" and "I need to show this trait to others." Table 4-8
shows the results from these three items along each of the nine traits.
Interestingly, on the self-description group, the subjects in this study generally
rated themselves above the scale midpoint on all nine items. Subjects believed that "fun-
loving" and "intelligent" described them most accurately among the nine items; to a
lesser extent, subjects believed themselves to be responsible, successful, confident, cool,
and attractive as well.
Most of these nine traits were also viewed as being important for public display.
Subjects believed it is particularly important to show others that they are confident, fun-
loving, responsible, and intelligent. To a slightly lesser extent, they also believed it is
important to show others they are successful, attractive, and cool.
Finally, subjects rated each of the nine traits as to their personal need to show the
traits to others. Five traits were rated, on average, above the scale midpoint; subjects had
the highest levels of motivation to show themselves as confident, fun-loving, responsible, intelligent, and successfiil to other people.
Also examined were the correlations among self-description, judged importance, and need-to-show for each of the traits. Based upon the literature review, we would expect subjects' responses on these three items to be positively correlated for any trait.
People are assumed to believe they possess those traits they seek to display for others, and they are motivated to display traits which are associated with social rewards. 100
Table 4-8 Mean Ratines of Self Bv Identity Traits
"Self-descriDtion"- Trait M SD fun-loving 6.17^ 0 94 intelligent 6 05* W.OO responsible 5.83' 1.22 successful 5.78"'' 0.97
confident 5.65*' 1 33 60"^ cool 5 1 98 44*' attractive 5 1 97
artistic 4.69'' 1 89
materialistic 4.16'' 1 61
"Judged imoortance' Trait M confident 6.43' 0 95 fun-loving 6 37' 0 RO
responsible 6.34' 1.01 intelligent 6.30' 1.01
successful 5 QR** 1.101 1 8 5.46*^ attractive 1 ,jy59 03*^ cool 5 1 .o8^J artistic 4.57'' 1 materialistic 2.16* 1 35 "Need-to-show": Trait M confident 5.88' 1 1'^ 73'*' fun-loving 5 1 4R 5.70''' responsible 1 S9 intelligent 5.69"' 1.49 successful 5.44" 1.55 attractive 4.82*= 1.66 cool 4.61"= 1.84 artistic 4.12" 1.86 materialistic 1.99' 1.19 Entries (in the same column) with different superscripts are significantly different at the .05 level by Scheffe procedures. 101 Table 4-9 shows the correlation (Pearson's r) among the three items for each of the nine traits tested in the first questioimaire. Table 4-9 Correlations Among Self-Description. Importance, and Need-to-show By Identity Traits "successfijl" : describes me important to show describes me important to show .36* I need to show .14 .39"' "fun-loying" : describes me important to show describes me important to show .54* I need to show .18* .35" "responsible" : describes me important to show describes me important to show .32* I need to show .30* .43^ "artistic": describes me important to show describes me important to show .69* I need to show .60* .71' "materialistic" : describes me important to show describes me important to show .37* I need to show .38* .50" "intelligent" : describes me important to show describes me important to show .57* I need to show .26* .35" 102 Table 4-9--Continued "cool": ucacriDCb me important to show describes me important to show .46* 1 need to show .34* .71* "confident": ucdciiucb me imDonant to snow describes me important to show .23* I need to show .10 .48* "attractive": describes me important to show describes me important to show .27* I need to show .04 .55* Note: Entries marked with (*) are significant at the 0.05 level. This pattern of correlations strongly suggests that subjects hold consistent views about the identity traits they possess, the traits that are most important for display to the outside world, and the traits they feel the most motivation to display to the outside world. Of the 27 correlations shown in Table 4-9, all were positive and 24 were statistically significant. The three exceptions concerned the relationship between "describes me" and "I need to show" for three traits-successfiil, confident, and attractive. Though there is no information in the present study to explain these exceptions, one possible explanation is that subjects believed that possession of those three identity traits is relatively easy for others to see in social situations. In such a case, people would feel little motivation to provide evidence of possessing the trait. Other traits listed (such as being artistic, intelligent, or responsible) might be perceived as more difficult to convey-thus, even 103 when one is confident of possessing the trait, there is still motivation to display evidence of the trait to others. Priming Manipulation Subjects were randomly assigned to either a primed or unprimed condition and, as described above, subjects in the two conditions did not differ significantly in the number of brands correctly recalled from the movie clip. Similarly, tests of brand recognition for each of the six brands in the movie clip revealed no significant difference in recognition between the primed and unprimed groups. However, this general motivational prime may have served to differentiate the two experimental groups in other ways, so an investigation was necessary. Should the two groups differ substantially on the dependent measures of attitude toward the brands and purchase intentions, separate analyses would be conducted on the Model. If the groups were largely the same on these measures, they might be combined for purposes of testing hypotheses. The effect of a strong manipulation should be seen in primed subjects' generally more positive attitude toward brands in the film clip and their intention to purchase those brands. Table 4-10 shows the mean scores, by priming condition, for the combined two- item brand attitude (Ab) and purchase intention (PI) scales. The priming manipulation produced a significant difference in brand attitude or purchase intention only in the case of Saab; primed subjects held a more positive attitude toward Saab and a stronger purchase intention than did non-primed subjects. No other differences in Ab or PI were seen for the other ten brands, so it would appear the priming manipulation produced only very minimal differences between the groups. Since a 1 104 Table 4-10 Scores on Ab and PI Scales For Primed and Unprimed Groups Primed Mean Unprimed Mean F Sig. of F (#) Ford Ab 8.13 8.56 0.75 .389 1 AA (#) Ford PI /A't 7.77 0.40 .529 (#) Saab Ab 12.11 11.14 4.35 .039* (#) Saab PI 9.92 4.11 .044* Chevrolet Ab 7.84 7.72 0.05 .817 Chevrolet PI O.J I 0.42 0.02 .879 Audi Ab 9.07 8.16 3.31 .071 Audi PI 0.32 1.01 .317 (#) BMW Ab 13.40 13.58 0.82 .367 (#) BMW PI iz.Zo 12.26 0.01 .939 (#) Snickers Ab 11.79 11.58 0.19 .665 I 1 T) (#) Snickers PI 1 i.jZ 10.83 0.77 .381 Milky Way Ab 9.29 8.79 0.75 .387 Milky Way PI o.Oj /.9b 1.02 .313 (#) Rolling Rock Ab 9.95 9.69 0.21 .645 Q 1Q (#) Rolling Rock PI y.jy o.9j 0.53 .469 Budweiser Ab 9.29 9.04 yj.mC\ 1 A ./II71 Budweiser PI 9.17 8.64 0.50 .481 (#) Diet Coke Ab 8.31 8.01 0.19 .667 Diet (#) Coke PI 7.57 7.27 0.16 .686 Diet Pepsi Ab 5.80 6.06 0.17 .678 Diet Pepsi PI 5.16 5.65 0.54 .464 Note: Brands marked with (#) were shown, mentioned, and/or used in the movie clip. difference in the dependent measures according to experimental group was seen in only one of the brands tested, and since there was no difference in brand recall or brand 105 recognition according to experimental group, the two groups were assumed not materially different and were combined for purposes of testing the hypotheses. Effects of Prior Viewing of the Movie Eighty-nine subjects (58%) reported having seen Reality Bites prior to this study. Several comparisons were made between subjects who had seen the movie before and those who were seeing the movie for the first time. The first comparisons involved subjects' reported attention to and enjoyment of the movie clip. Since those who saw the movie previously presumably did so out of choice (they enjoy the actors, genre, or some other attribute of the film), they might be expected to show more interest in the movie clip. Prior viewers reported finding the movie clip more interesting (M = 6.49) and attention-holding (M = 6.56) than did those who had not seen the = movie (means 6.08 and 6.16, respectively; both p's < .04). On the measures of liking the movie and finding the movie believable, however, there was no significant difference between the groups. Despite the differences in interest and attention, there was no significant difference in the number of brands recalled by those who had seen the movie (M = 4.37) and those who had not (M = 3.89). Further, there was no difference between the two groups in recognition of any of the 1 1 brands. 106 Tests of Hypotheses The hypotheses advanced in the present study concern relationships between variables which are part of an overall model. That overall model is itself believed to describe a set of brand-specific effects possible through brand placement. The examination of hypothesized relationships should start with the general model, followed by more specific links in the model. One early decision in model construction involves the ordering of variables. In some cases, the ordering is plainly dictated by time. For example, in the present study an individual's preexisting evaluation of a brand clearly precedes exposure to a brand placement, which in turn is followed by any change in evaluation precipitated by the brand placement. In other cases, when no such temporal guideline is apparent, theory must be the guide (Reis 1982). The literature review provided guidelines for the order of variables in the present study. The order and direction of relationships among variables allows for the construction of a recursive model, or one which travels only in a single direction. Since the model is recursive, path coefficients may be estimated by regression weights (Reis 1982). In this study, the overall model may be examined with many combinations of brands, identity traits, and characters. In many (if not most) cases, these combinations are not readily comparable to one another. The analysis here involves an examination of various combinations so that a determination may be made as to whether the path models are consistent with hypotheses. The models may be compared on the basis of the overall variance explained; the squared multiple correlation, a statistic that is independent of 107 units of measurement, gives the proportion of a variable's variance that is accounted for by its predictors (Arbuckle 1997). In examining the hypothesized relationships among variables, one can simply calculate whether the paths hypothesized by each model are significant and in the hypothesized direction. Figure 4-1 shows an operational model for examining the hypothesized relationships. Not shown in this model are the error terms associated with each endogenous variable. Those error terms are necessarily included for purposes of (brand) PI or Ab, time 1 (trait) describes me I need to show (trait) using (brand) (brand) PI shows I am or Ab, time 2 (trait) (character) (trait) is important to show unaided (brand) recall Figure 4-1 General Path Model for Brand / Character / Trait Combinations 108 calculation (so that the model is identified), but for purposes of clarity and simplicity are not included in the figures presented here. While an examination of the model across all combinations of (1) brands in the movie, (2) characters who were paired with the brands in the movie, and (3) identity traits would not be a valid or enlightening one, the model must be examined across several comparable combinations so that its predictive power (and that of the individual paths) may be assessed. As a starting point, we could examine the combination of a brand which showed significant change in Ab and/or PI, a character who was individually and exclusively associated with that brand, and identity traits associated with that character. Consider the match of the "Michael" character with Saab. Michael drove a Saab in the film clip, and no other character was shown with the Saab brand. Saab was generally well-remembered fi-om the clip; there was also a substantial change in Ab and PI for Saab after subjects saw the movie clip. Further, subjects saw in Michael a mix of positive and negative traits including success, responsibility, and materialism. In short, there is a significant impact on the brand and a great deal of variance to be explained. One combination to examine is that of Michael, Saab, and how to show oneself to be successfiil. The results of these analyses are shown in Figures 4-2 and 4-3. In this combination, where a character judged as best exemplifying a positive trait was exclusively paired with the brand, the model is strong. The predictors in the model in total account for 58% of the variability in attitude toward Saab and 55% of the variability in purchase intention for Saab after subjects saw the movie clip. Moreover, in both 109 Saab Ab, time 1 Notes: Nonsignificant paths removed. Model x^= 18.37,p<.02. Squared multiple correlation of Saab Ab = .58. Squared multiple correlation of "using Saab shows I am successful" = .21. Figure 4-2 Path Analysis Results for Saab / Michael / Successful Combination on Ab 110 Saab PI, time 1 successflil describes me I need to show success — — .14 using Saab Saab PI, » .41 shows I am time 2 .21 successful .98 Michael is successflil success IS ' important -.31 to show unaided Saab recall Notes: Nonsignificant paths removed. Model x^= 16.29,p<.04. Squared multiple correlation of Saab PI = .55. Squared multiple correlation of "using Saab shows I am successful" = .21, Figure 4-3 Path Analysis Results for Saab / Michael / Successful Combination on PI . Ill models of the variability 21% of the brand's perceived power in displaying the trait is explained. The first hypothesis (HI) stated that the more beneficial an identity trait is judged to be, the greater the individual's motivation to claim and display that trait. In this analysis, the importance assigned to showing success to others ("success is important to show") is strongly and positively related to the motivation to display success ("1 need to show success"). This result supports HI The second hypothesis (H2) posited that the greater one's motivation to claim and display a trait, the more utility assigned a brand in conveying that trait to others. In this case, the need to show success was significantly and positively linked to the belief that driving a Saab will demonstrate success to others. H2 is supported. The third hypothesis (H3) posited that the more closely a character is associated with a desired trait, the more useful a relevant brand (i.e. a brand used by that character) will be judged in conveying that trait. In this case, the degree to which Michael is viewed as successful ("Michael is successful") is significantly and positively linked to the belief that driving a Saab will show success to others ("using Saab shows I am successful"). H3 is supported. The fourth hypothesis set (H4a and H4b) stated that the utility assigned to a brand in conveying a desired trait is positively related to the desire for that brand. In this case, there was a very strong positive relationship between subjects' belief that driving a Saab would show others they are successful ("using Saab shows I am successful") and both Ab and PI for Saab after watching the movie clip; these results support H4. 112 While the model proved enlightening for this particular combination, a more complete analysis should include this character / brand pairing across other traits--both traits not so closely associated with this character and more negative traits that were associated with this character. For the former, an analysis was made of the model applied to Michael, Saab, and the power to display the trait of being fun-loving; Michael was judged to be the least flin-loving character among the five in the movie clip. Figures 4-4 and 4-5 show the results of this combination. In this case, the character and brand have not changed but the relationship to the trait has; the character was not judged to have the "fun-loving" trait. The model's overall predictive power dropped, and several path coefficients changed dramatically. Perhaps most telling, the path coefficients representing H4 dropped fi-om very high values (.91 and .98) to nonsignificance. The perceived ability of Saab to convey a fun-loving quality was not a predictor of Ab or PI after subjects saw the movie clip. Note two other particular changes in path coefficients. First, recall of the brand became a much more important predictor of purchase intention than was the case with the prior combination. Where no new identity-relevant information was offered in the movie clip, simple recall of the brand's appearance assumes a more important role. Second, the path fi-om "need to show I am fim-loving" to "using Saab shows I am fun-loving" (describing H2) is no longer significant when the character is not judged as having the trait of being fun-loving. . 113 Saab Ab, time 1 Nonsignificant paths removed. Model x^ = 3.60, NS. Squared multiple correlation of Saab Ab = .5 1 Squared multiple correlation of "using Saab shows I am fun-loving" = .22. Figure 4-4 Path Analvsis Results for Saab / Michael / Fun-loving Combination on Ab 114 Saab PI, time 1 Note: Nonsignificant paths removed. Model x^=3.71,NS. Squared multiple correlation of Saab PI = .46. Squared multiple correlation of "using Saab shows I am fiin-loving" = .18. Figure 4-5 Path Analysis Results for Saab / Michael / Fun-loving Combination on PI 115 The model can also be examined when a more negative trait—materialism-is strongly associated with the character. Michael was judged to be the most materialistic of the characters. Figures 4-6 and 4-7 show the results of combining Michael, materialism, and Saab. Notes: Nonsignificant paths removed. Model x^=5.63,NS. Squared multiple correlation of Saab Ab = .52. Squared multiple correlation of "using Saab shows I am materialistic" = .02. Figure 4-6 Path Analysis Results for Saab / Michael / Materialistic Combination on Ab 116 Saab PI, time 1 ..56 materialistic describes me 1 need to shov ' materialistic using Saab Saab PI, .80 shows I am time 2 materialistic Michael is materialistic materialistic is important to show unaided Saab recall Notes: Nonsignificant paths removed. Model x^ = 4.32,NS. Squared multiple correlation of Saab PI = .46. Squared multiple correlation of "using Saab shows I am materialistic" = .01. Figure 4-7 Path Analysis Results for Saab / Michael / Materialistic Combination on PI In this case, where a character was associated with a more negative trait, the model is again less explanatory than was the case involving a strong association with a positive trait (refer to Figures 4-2 and 4-3). The H3 link (from "Michael is materialistic' 117 to "using Saab shows I am materialistic") and the H2 link (from "1 need to show I am materialistic" to "using Saab shows I am materialistic") are no longer significant. In sum, when a particular trait is not associated with the relevant character, or when the trait associated with the character is a more negative one, the model is not as explanatory. The preceding analyses took one character-to-brand link and compared the Model across different traits. For a further exploration of the robustness of the Model's explanation, the next analyses begin with a group of traits. These traits differ according to the level of motivation subjects reported for displaying the trait. As an example, subjects placed "responsible" in the set of the traits they felt the greatest need to show others (see Table 4-10). Lelaina and Michael were both rated as the most responsible of the movie characters, while Troy was rated least responsible. Troy was exclusively paired with the Snickers brand in the movie clip. Does the Model explain the significant change in purchase intention for Snickers in terms of the Troy-to- responsible coimection? Figures 4-8 and 4-9 show the results of the path analyses. This combination is in many ways comparable to the Saab/Michael/fun-loving combination (Figures 4-4 and 4-5). In both cases, a desired trait is poorly represented by a character; in both cases, the H4 paths from "using the brand shows I have the trait" to Ab and PI are weak or nonexistent (significant in only one of four analyses). Contrast this result with one from a Troy-to-Snickers link that also includes a trait that Troy is strongly perceived to have. Troy was rated highly on being cool, so we would expect analyses of the Snickers/Troy/cool combination to both show more overall . ; !, 118 Snickers Ab, time 1 recall Notes: Nonsignificant paths removed. Model x^=8.87, NS. Squared multiple correlation of Snickers Ab = .60. Squared multiple correlation of "using Snickers shows I am responsible" = .07. Figure 4-8 Path Analysis Results for Snickers / Troy / Responsible Combination on Ab 119 Snickers PI, time 1 :75" responsible describes me recall Notes: Nonsignificant paths removed. Model 10.50, NS. Squared multiple correlation of Snickers PI = .63. Squared multiple correlation of "using Snickers shows 1 am responsible" = .06. Figure 4-9 Path Analysis Results for Snickers / Troy / Responsible Combination on PI 120 predictive power and show stronger links from "using Snickers shows 1 am cool" to Ab and PI for Snickers. Figures 4-10 and 4-1 1 show the results of these path analyses. unaided Snickers recall Notes: Nonsignificant paths removed. Model x^=3.97,NS. Squared multiple correlation of Snickers Ab = .65. Squared multiple correlation of "using Snickers shows I am cool" = .24. Figure 4-10 Path Analysis Results for Snickers / Troy / Cool Combination on Ab .J 1 » 121 Snickers PI, time 1 cool describes me I need to show cool — —» .08 using Snicker 3 Snickers PI, » .86 shows I am time 2 .33> cool .54 Troy is cool cool is important to show unaided Snickers recall Notes: Nonsignificant paths removed. Model x^ = 2.46, NS. Squared multiple correlation of Snickers PI = .69. Squared multiple correlation of "using Snickers shows I am cool" - .26. Figure 4-1 Path Analysis Results for Snickers / Troy / Cool Combination on PI 122 Even though subjects did not generally feel a high level of motivation to display the trait of "cool," more variance is explained in this combination than in the previous one (squared multiple correlations of .65 versus .60 for Ab and .69 versus .63 for PI). A stronger character-to-trait link appears vital for enhancing explanatory power. The H2 link was mildly positive in the Troy/cool combination (.08), as subjects made some association of the need to show themselves to be cool with the belief that eating a Snickers bar will demonstrate the quality of cool to others. In the Troy/responsible combination, however (Figures 4-8 and 4-9), the H2 link was negative (-.16). The H3 link is more positive when a stronger trait association is made to the character; the coefficient is .33 in Figures 4-10 and 4-11, but only .12 in Figures 4-8 and 4-9. The H4a link from "using Snickers shows I am cool" to Ab (.54) is much stronger than the comparable link for being "responsible" (.24). The H4b link from "using Snickers shows I am cool" to PI (.54) is much stronger than was the comparable link in Figure 4-9 (nonsignificant). So, the combination of a strong character-to-trait link, but with a trait not highly desired for display to others, still reflected support for HI, H2, H3, and H4. A similar character-to-trait link involves Lelaina and attractiveness. Subjects did not rate highly their motivation to show themselves as attractive, but they rated Lelaina as significantly the most attractive character. Lelaina was seen driving a BMW (given to her by her father) in the movie, and was the only character driving a BMW. Figures 4-12 and 4-13 show analyses of the combination of BMW, Lelaina, and attractiveness. The overall variance in Ab and PI explained fell to .38 and .44, respectively, and the pattern of significant paths changed somewhat from the Snickers/Troy/cool combination. 123 BMW Ab, time 1 Notes: Nonsignificant paths removed. Model x^ = 4.61,NS. Squared multiple correlation of BMW Ab = .38. Squared multiple correlation of "using BMW shows I am attractive" = .15. Figure 4-12 Path Analysis Results for BMW / Lelaina / Attractive Combination on Ab 124 BMW PI, time 1 Notes: Nonsignificant paths removed. Model x^=8.94, NS. Squared multiple correlation of BMW PI = .44. Squared multiple correlation of "using BMW shows I am attractive" = .14. Figure 4-13 Path Analysis Results for BMW / Lelaina / Attractive Combination on PI The HI, H2, H3, and H4a links are significant and positive, though the H4a link is fairly weak. In this case, however, the H4b link is no longer significant. The breakdown of the explanatory model at the H4 link in this particular case may relate to how the trait 125 of attractiveness is perceived. As mentioned, showdng oneself to be attractive was not a strong motivation for the subjects in this study. Further, subjects may consider attractiveness—especially if it is conceived as physical attractiveness—as not easily conveyed through product use (especially a product such as a car). Finally, as a contrast, a trait which was rated highly on the "need to show" is considered. The trait with the highest mean rating on "need to show" was confidence, and the mean score for Michael on confidence was the highest among the characters (though the score was not significantly higher than that for Lelaina at the .05 level). Figures 4-14 and 4-15 show the combination of Saab, Michael, and showing confidence. As display of the trait is more highly desired, the squared multiple correlation rises and the paths described by H4 are again significant. The HI, H2, and H3 links are positive and significant as well. 126 Saab Ab, time 1 .34 X confident describes me .10 Saab Ab, .28 time 2 confident is important to show unaided Saab recall Notes: Nonsignificant paths removed. Model x^= 18.90,p<.02. Squared multiple correlation of Saab Ab = .55. Squared multiple correlation of "using Saab shows I am confident" = .13. Figure 4-14 Path Analysis Results for Saab / Michael / Confident Combination on Ab 127 Saab PI, time 1 Notes: Nonsignificant paths removed. Model x^= 19.83,p<.02. Squared multiple correlation of Saab PI = .52. Squared multiple correlation of "using Saab shows I am confident" = .13. Figure 4-15 Path Analysis Results for Saab / Michael / Confident Combination on PI 128 ' Summary of Results This study was conceived around the idea that the effects of brand placement could not be adequately explained by pre-existing brand attitudes and/or memory for the brand placement; given the depth of audiences' relations with media characters, a model incorporating audiences' processing and use of identity-relevant information is needed. Further, the use of identity-relevant information in the brand placement was hypothesized to have several related links: where traits are viewed as important to display for others, individuals feel more motivation to display those traits; the motivation to display specific traits is positively related to the perceived value of a relevant brand in displaying the trait; the more a media character (who is paired with a brand) is judged to have a trait, the greater the perceived value of the brand in displaying the trait; and the greater the perceived value of the brand in displaying the trait, the greater the desire for the brand. All of these hypothesized links were supported to some degree in the combinations analyzed above. Further, the impact of brand placement, in the forms of enhanced brand attitudes and purchase intentions, was robust across expensive and inexpensive brands. Evidence for the four hypotheses varied according to the character-to-trait combination studied. In all seven combinations, HI (the link from importance given to display of a trait to motivation to display that trait) was supported, as the link was * positive and significant. There was only mixed support for the H2 link (from the need to display a trait to the perceived value of the brand in conveying the trait). In the seven combinations discussed above, the H2 link was positive and significant in only three cases. In six of the seven combinations, H3 (the link between a character possessing a 129 trait and the value of a paired brand in conveying the trait to others) was supported. Only in the combination of Saab, Michael, and materialism (Figures 4-6 and 4-7)~where a negative trait was involved~was the H3 link not significant. In six of the seven combinations, the H4a link (from the value of a paired brand in conveying a trait to others to Ab) was supported; the combination of Saab, Michael, and fun-loving (where the character was not perceived to have the trait) was the lone exception. In the combinations of Saab, Michael, and fun-loving (Figure 4-5), Troy, Snickers, and responsible (Figure 4- 9), and BMW, Lelaina, and attractiveness (Figure 4-13) the H4b link was not significant. Two of those combinations also involved poor character-to-trait matches, so they are understandable from the conceptual model. Overall, the analysis of different combinations of brands, traits, and characters provides support for HI, H3, and H4, with mixed results for H2. Across combinations, the path analyses explained as little as 38% of the variance in Ab and 44% of the variance in PI for the brand following the movie clip; the analyses also explained up to 65% of the variance in Ab and 69% of the variance in PI. Generally, less variance was explained in cases of a negative trait or where the character was not perceived to have the relevant positive trait (though the BMW/Lelaina/attractive combination was an exception). More variance in Ab and PI was explained when the character is judged to have the relevant positive frait. There was also a substantial range of variance explained in the brands' perceived utility in conveying traits. Among the Ab models, the lowest multiple squared correlation was .02 (the variance explained in "using Saab shows I am materialistic") and the highest 130 figure was .24 (in "using Snickers shows I am cool"). Among the PI models, the lowest variance explained was .01 (in "using Saab shows I am materialistic") and the most variance explained was .26 (in "using Snickers shows 1 am cool"). Again, more variance in brands' utility in conveying traits to others was explained when the character was judged to have the relevant, positive trait. Thus, particular placement executions can affect not only Ab and PI, but also judgments of the brand's power in helping convey desired traits to others. In the most typical cases of brand placement, where easily available brands are paired with characters who possess one or more desirable traits, the Model explains much of the variation in subsequent desire for the brand. Importantly, the conceptual approach taken in this study—relating the effectiveness of placement to its utility in helping audiences display desired identity traits—provides incremental explanation over and above brand recall and preexisting purchase intentions. CHAPTER 5 CONCLUSIONS This dissertation investigated audience effects of brand placements, using a framework suggested by impression management research. Some important advances to the existing brand placement literature were made in the process. Not only were the changes in brand attitudes and purchase intentions consistently strong for brands appearing in the movie clip, they were also largely confined to those brands which did appear (and not to competing brands which did not appear). Importantly, a mechanism related to personal identity display was found to explain brand placement impact to a far greater extent than does memory alone--a mechanism involving individuals' desire to appropriate brands which reflected positive traits shown by the program characters. Three of the four hypotheses underlying the proposed mechanism were consistently supported, while a fourth was supported only in situations involving desired traits which were strongly associated with a relevant movie character. One surprising result in the present study concerned the impact of affordability on desire for the brand, especially the component of desire for the brand related to purchase intention. While more positive brand attitudes might result from brand placement for almost any type of product, we might expect that significant increases in purchase intention would not occur for high-priced brands. Audience members might like a brand more after seeing it in a program, but would still consider cost in gauging their likelihood 131 132 of purchase. However, the significant increases in purchase intentions occurred both in very low-priced brands (such as Snickers and Diet Coke) and in very high-priced brands (such as BMW and Saab). Why the significant increases in both brand attitudes and purchase intentions, even for high-priced brands? One possible explanation is that, while the brands are expensive, the context surrounding their inclusion in the movie may have made the brands seem relatively attainable for the audience. The expensive brands (BMW and Saab) were shown being used by young adults who were only a few years older than the subjects in the study. Perhaps subjects saw, through the experiences of these young professionals, a greater possibility of purchasing these brands themselves. Had the characters in the movie been older, or otherwise less like the subjects in the study, the pattern of results might have been different. '• , ' i Limitations of the Study . ' . ; \ The results of this study are encouraging, but they remain a first step in understanding an under-researched brand communication tool. Several issues of reliability and validity may limit the conclusions to be drawn from this study. Reliability and Validity The issues of reliability and validity must be addressed before any conclusions are drawn from this study. Reliability is the degree to which individuals' scores on some test yield consistent results over repeated administrations or alternate forms (Cook and • . . ' . . . 133 ; : .. . Campbell 1979). In this study, correlation coefficients were calculated to assess the internal consistency of some pairs of items purported to measure the same underlying construct; in all cases, the pairs met the criteria for confirming an adequate level of internal consistency. Validity is a larger issue, involving the best available approximation to the truth or falsity about propositions (Cook and Campbell 1979). Four major categories of validity- internal, external, construct, and statistical conclusion validity-deserve attention before conclusions can be drawn from the results of this study. Internal validity refers to our ability to infer that a relationship between two variables is causal; its essence is the accounting for third-variable alternative explanations of presumed two-variable relationships. External validity refers to the ability to infer that a presumed causal relationship can be generalized to and across different types of people, settings, and times. Construct validity concerns the possibility that an operation meant to represent one particular cause or effect can be construed in terms of more than one construct. Statistical conclusion validity concerns drawing correct conclusions about the covariation among variables (Cook and Campbell 1979). Internal Validity Cook and Campbell (1979) specified a number of threats to internal and external validity that must be guarded against. One potential threat to internal validity that is relevant to this study is "ambiguity about the direction of causal influence." This is a particular threat in correlational studies that are cross-sectional, when it is often plausible 134 that A could cause B or that B could cause A. While many of the relationships specified in the present study could be ordered on the basis of either theory or clear temporal patterns, for some relationships the paths might be open to debate. Three issues involved with the administration of the study might have led to validity concerns. The first is the issue of the motivational prime. While the manipulation check showed a statistically significant difference in knowledge of the prime between groups, there are some practical concerns. Subjects in the primed condition were generally quite sure that they had, in fact, been primed; the mean score on confidence that the prime had occurred was very high. Subjects not in the primed condition, on the other hand, did not answer the manipulation check item near the bottom of the scale; their answers were clustered about the center of the scale. So, it is possible that subjects in the primed condition knew they had been primed but many subjects in the unprimed condition simply were not sure. This may have served to mute actual differences between subjects in the two groups. Another area of some concern is that, given a fairly long pair of questionnaires and a limited amount of time for subjects to complete the study, it is possible that some subjects became fatigued and hurried to finish. A briefer questionnaire, or a slightly longer time period for subjects to complete the questionnaires, might lead to an even stronger pattern of results. A third area of concern is the possibility of demand characteristics. Demand characteristics "include all aspects of the experiment which cause the subject to perceive, interpret, and act upon what he believes is expected or desired of him by the 135 experimenter" (Sawyer 1975, p. 20). Subjects may adopt one of several roles: the "good" role, in which one tries to confirm what one believes to be the experimental hypothesis; the "faithful" role, in which one closely follows instructions and is nearly docile in the research setting; the "negative" role, in which one tries to disconfirm a suspected experimental hypothesis; or the "apprehensive" role, in which a subject worries about performance in the research setting. Biases may result from the "good" or "negative" roles. Bias should not occur if subjects assume the "faithful" role. The impact of the "apprehensive" role is the most ambiguous and unpredictable (Sawyer 1 975). The possibility exists that a number of subjects adopted the "good" role, correctly guessing the purposes of the study and aligning their responses accordingly. However, it is unlikely that such demand characteristics drove the results of the present study. The results were not uniform-one brand mentioned in the film clip (Ford) was subsequently rated more negatively on both Ab and PI. Further, if hypothesis-guessing was a major determinant of the experimental results, we should expect to see more evidence of it among subjects in the primed condition. As noted above, there were very few differences in the responses of primed subjects versus unprimed subjects. In future research that employs similar methods, a number of steps could be taken to alleviate concerns about demand characteristics. A cover story could be presented to subjects, which might lessen the possibility of hypothesis-guessing regarding the role of the program stimulus. Initial questions about brands and traits-presented to subjects in the first questionnaire in the present study-could be presented much earlier; separating 136 these questions in time from the subsequent presentation of a program stimulus could lessen the chance of hypothesis-guessing as well. External Validity Issues of external validity are also a particular concern, given the artificiality of the movie viewing situation. Subjects were in a forced viewing situation and saw only a short portion of a feature film (rather than the entire film, as they would in a more natural movie viewing environment). The combination of a short movie clip and a clear testing situation might have inflated the level of attention paid to the movie and thus led to inflated recall percentages and recognition scores. The recall percentages reported in Chapter 4 should not be taken as representative of the typical memory for brands placed inside a movie. It also is not clear that the results can be generalized to other programs or audiences. Subjects represented a convenience sample that was concentrated in a narrow age group and socioeconomic status, so the applicability of the model to other parts of the population remains an open question. Further, it is far from certain whether other types of movies, or even television shows or other media vehicles, would elicit a similar pattern of processing. While the results of the present study are encouraging for our understanding of brand placement effects, it may be the case that these effects are combined to fairly narrow instances. Future research will be needed to establish boundary conditions for the effects seen in this study. 137 Construct Validity and Statistical Conclusion Validity Cook and Campbell (1979) list a number of potential threats to both construct validity and statistical conclusion validity. Among the threats to construct validity are inadequate explication of measures prior to data collection, evaluation apprehension, experimenter expectancies, and restricted generalizability across constructs. In the present study, the measures of Ab and PI were well established from prior studies by other researchers; other measures had been pretested with a different group of subjects. There seems to be, in the present study, an acceptable fit between constructs and the measures designed to tap those constructs. As Cook and Campbell (1979) point out, though, researchers "can use the obtained pattern of data to edit one's thinking about both the cause and effect constructs, and one can suggest, after the fact, other constructs that might fit the data better than those with which the experiment began" (p. 69, emphasis in the original). The process of obtaining the best measures for anticipated constructs is clearly an evolving one. -'^ Among the threats to statistical conclusion validity are low statistical power, low reliability of measures, and random heterogeneity of respondents. In the present study, these threats appear minimal due to acceptable indicators of reliability and the use of within-subject tests. In fiiture work, the use of more multiple-item measures would provide more opportunity to assess constructs and would raise statistical power as well. 138 Conclusions and Implications for Brand Placement Research and Practice This study has extended the brand placement literature by identifying a model of effects that moves beyond memory alone. The framework of impression management was shown fruitful in explaining how brands function as meaningful social badges, as well as how those badges can be adopted from media depictions. The results of this study might also serve to expand our conception of how audiences relate to and emulate elements of media characters. Since the present study employed a convenience sample, a single short movie clip, and a forced-viewing environment, we might ask whether the results could provide practical guidance for advertisers considering the use of brand placement in many media settings. case A can be made that, even with this single and narrow study, some practical knowledge has been gained. Two primary areas of insight come in the measurement of brand placement effectiveness and the specific program contexts in which advertisers might seek to place their brands ^ for maximum impact. ; '. While advertisers have typically used simple recall or recognition measures to evaluate brand placement effectiveness, the present study suggests more market-relevant measures (such as brand attitude or purchase intention) may be valid as well. Further, the context in which the hypothesized model was most predictive (the brand is closely associated with a character believed to exemplify one or more positive traits) is the same context advertisers say they seek in brand placement opportunities. As discussed in Chapter 2, many advertisers maintain specific, active efforts to avoid negative associations with their brands in media programs. Placements are also considered more 139 valuable when they are moved from a background spot to be more closely paired with an attractive character, hi short, most advertisers already seek for their brands a close association with characters who have desirable traits. The hypothesized model is thus relevant for evaluating a large percentage of brand placement opportunities. By limiting themselves to simple memory measures of placement effectiveness, however, advertisers may not fully understand the potential communication impact of brand placement. A model such as the one presented here may help advertisers screen placement opportunities to find the rare "home run" occasion. The impression management framework was shown quite usefiil in developing new and expanded views of the growing area of brand placement. As an input into public policy efforts, the framework might be particularly useful as well. In the area of media programs targeted towards children (who are particularly active in learning and establishing identity), the model used here might aid in establishing standards for in- program depictions. In the same way the present framework may aid advertisers who use placement to promote their brands, it might also provide guidance for public policy makers to protect segments of the audience who are particularly vulnerable or who do not recognize promotion attempts. hi considering the applicability of the present framework across many settings and cultures, three other groups of issues arise: the social power of brands in different settings, including perceptions of brands' utility in conveying identity; perceptions as to whether all individual identity traits may be effectively conveyed through brand usage; 140 and the overall role of self-presentation across cultures. All of these issues suggest the framework must be tested in different economic and cultural settings. The first of these three issues~the communication value of brands themselves-- may concern both social norms and the relative development of a market economy in a given culture or setting. As seen in the present study, the power of brand placement relates to consumers' beliefs that brands carry social information, hi the United States, many brands are well-known and well-understood as to their social value; these brands are also generally available to most consumers. If an audience member learns some important piece of social information about a brand, she can usually purchase and/or use the brand, and do so knowing that other people will recognize and decode the social information contained m that brand. In short, she can generally encode desired information about herself in the brand she chooses, and feel confident that others will decode that information in the way she intended. If brands are not well recognized as carriers of social information, their utility in conveying identity is lost. The present model might not be applicable in areas with developing economies and less tradition of extensive brand use and display. Some evidence concerning the second issue-whether all identity traits can be effectively conveyed through brands-was seen in the present study. For instance, the model was less explanatory when the trait of attractiveness was involved than when traits such as success or confidence were involved. People may believe attractiveness is a trait that is easily judged by others, and rather unaffected by brand use. It is also possible that traits involving material success or social membership are easier to convey through brand 141 usage than are more internally-centered traits such as intelligence. Therefore, we should not consider all identity traits of program characters as equally likely to help in brand communication efforts. . final A issue concerns the role of self-presentation across settings. The literature review painted a picture of individuals with an ongoing motivation to express desirable elements of identity to others, and a social audience generally accepting of those displays. Brand placement's effectiveness comes in large part through its role in this ongoing process. But, if the social backdrop were different-especially if the claim and display of identity traits was not socially acceptable or useful-then brand placement would only be effective in communicating non-social brand attributes; placement would be best used to show functional attributes in an entertaining way. We might predict more effectiveness for brand placement in those cultures and settings where it is both common and socially acceptable for individuals to claim and display desired elements of their identities. A number of questions seem worthy of attention; it is hoped this study will frame and stimulate productive research in this important area. APPENDIX A BRAND APPEARANCES IN REALITY BITES time in scene brands 0:00:10-0:01:03 open: graduation speech by Leiaina BMW (dialogue) 0:01:03 -0:03:50 Lelaina, Troy, Vickie, and Sammy: drinking, Diet Coke (in use~ celebrating graduation on rooftop Lelaina) Rolling Rock (in use-all) Coke (dialogue-Troy) 0:03:50 - 0:06: 10 Lelaina at dinner with parents; gets BMW Infiniti (dialogue) BMW (dialogue-Lelaina) Ford (dialogue) 0:06:10 - 0:08:49 Lelaina at her house; Troy leaves a girl's Snickers (in use-Troy) apartment; Vickie records another sexual conquest (#66); Troy reading while at work at newsstand 0:08:50 - 0: 10:50 see set of "Good Morning Grant," where Lelaina works as a production assistant; Lelaina forgets host's coffee, gets scolded 0: 1 0:5 - 0: 1 1 :25 in the Gap, where Vickie 1 is assistant manager Gap (dialogue; logo) 0:1 1:25 - 0:12:52 Lelaina and Vickie in BMW; see Michael for BMW (in use-Lelaina; the first time (on cell phone, driving); two logo); Saab (in use- vehicles are in a fender-bender Michael) 0: 12:53 - 0: 14:35 Lelaina and Michael in Michael's office Coke 0: 14:36 - 0: 16:25 Lelaina and Vickie back at house; Troy moves m 142 143 time in scene brands 0:16:26-0:18:05 see Lelaina's home video clips: Troytalking about his band; house scenes 0:18:06 - 0:18:33 Lelaina looking at her video shots while at work; more problems with boss 0:18:34 - 0:21:15 Troy, Vickie, Sammy drinking, smoking pot, Gap (dialogue) watching TV at house-- tell Lelaina that Vickie is now manager of Gap 0:21:15 - 0:22:18 Lelaina, Troy, Vickie, Sammy take gas card to Pringle's, Diet Coke, convenience store to buy snacks (dance to "My Evian (dialogue). Coke Sharona") 0:22: 19 - 0:23:50 group together at house, drinking; Michael Rolling Rock (in use-all) comes by for date with Lelaina 0:23:50 - 0:3 1 :25 Michael and Lelaina on date, later making out; Snickers (dialogue- Troy sees them, is jealous; Troy and Lelaina Lelaina, about Troy) fight 0:31:26 - 0:33:45 more of Lelaina's video shots: Vickie at Gap; Gap (dialogue-Vickie), Troy talking about his parents; Vickie getting "Quarter-pounder AIDS test w/cheese" (dialogue- Troy), Camel (dialogue- Troy) 0:33:45 - 0:35:48 Lelaina at work: boss insults her, video; Lelaina sabotages his note cards for show 0:35:50 - 0:37:30 Lelaina at house, tells Troy, Vickie, Sammy she Pringle's (in use-Troy), has been fired Nutrasweet (dialogue), .. Gap (dialogue-Vickie, Lelaina), Diet Coke 0:37:30 - 0:40:00 Troy and Lelaina go for a walk, smoke Radio Shack (dialogue- cigarettes; Troy makes advance Troy) 0:40:00 - 0:41:03 more of Lelaina's video shots: Lelaina talks Diet Coke about her parents, family; Vickie talks about her parents 0:41 :03 - 0:42:55 Michael offers to put Lelaina's video on his show 144 time in scene brands 0:42:55 - 0:46: 10 Lelaina has several unsuccessful job interviews; sees Troy at diner 0:46:10 - 0:47:12 Lelaina, while smoking, asks parents for money Diet Coke, Roiling Rock 0:47:13 - 0:48:21 Lelaina looking for fast-food job 0:48:22 - 0:53:45 Lelaina back home, starts self-pitying TV- Cheetos (in use~Lelaina) watching marathon; starts to call "psychic" 900 numbers 0:53:45 - 0:54:25 Lelaina talking to her dad in plant, he won't lend her money 0:54:25 - 0:58:25 Lelaina starts using gas card to raise money; Diet Coke, Rolling Rock back at house later; argument with Troy 0:58:25 - 1 :03:00 Lelaina on walk with Vickie, then at diner together; Vickie worrying about AIDS; Lelaina gets call from Michael about video 1 :03:00 - 1 :05:05 more of Leiaina's video shots: Vickie gets AIDS test result (negative); Sammy struggling with his homosexuality 1 :05:05 - 1 :07:56 at house, Michael comes by; argument with Troy - 1 :07:56 1 : 12:40 Lelaina and Michael at video screening; Pizza Hut (on video) argument 1:12:41 - 1:16:35 Lelaina at home; long talk with Troy; they make Diet Coke (in use- love Lelaina) -1 1 : 16:35 :24:50 next morning, Troy leaves; Troy's band playing Rolling Rock (in use), at club; Lelaina goes to club; Michael shows Diet Coke (dialogue), up; Lelaina leaves Cocoa Puffs (dialogue- Troy) 1:24:50- 1:27:45 Troy leaves town, goes to see dying father, Continental Airlines Lelaina missing him 145 time in scene brands 1 :27:45 - 1 :32:00 Lelaina finds out from Sammy that Troy is in Chicago; she prepares to go; Troy comes back; they profess love 1 :32:00 - 1 :32:45 Lelaina and Troy together APPENDIX B QUESTIONNAIRES FOR MOTION PICTURE STUDY Ouestionnaire One 1. Are you: male female 2. What is your major field of study in school? business administration / accounting mass communication liberal arts other 3. What is your classification in school? freshman sophomore jvmior senior graduate student 4. What is your family's approximate yearly income? less than $20,000 $20,000 - $29,999 $30,000 - $39,999 $40,000 - $49,999 $50,000 - $59,999 $60,000 or more don't know / not sure 146 147 5. In an average month, how many times do you go to movies in theaters? never once per month twice per month three times per month more than three times per month 6. In an average month, how many times do you watch rental movies on a VCR? never once per month twice per month three times per month more than three times per month 7. What are your favorite types of movies (check all that apply)? comedy action/adventure romance mystery biography drama musical/concerts animation/cartoons not sure/none of these apply to me 148 8. Please give your opinion about the following brands . Place a mark in the box that most closely matches your opinions on each brand: pond - bad Chevrolet Audi Snickers Milky Way DUQweiser n n n n n n n Rolling Rock Diet Coke Diet Pepsi BMW Ford Saab I like it I dislike it Chevrolet Audi Snickers Milky Way Budweiser Rolling Rock Diet Coke Diet Pepsi BMW Ford Saab 11 149 / definitely I definitely do not intend to buy it intend to buy it n n1 1 n n n1 1 n n AuHi n n n n1 1 n n n Snickers n n n n n n1 1 n1— Vfillfv AVflv n n n n n1 1 n n n n n1 1 n n1 1 n n Rolling Rock un n U U Ur~l U u1— Diet Coke Diet Pepsi BMW Ford Saab / would buy it I would not buy it ifI could even ifI could Chevrolet Audi Snickers Milky Way Budweiser Rolling Rock Diet Coke Diet Pepsi BMW Ford Saab 150 9. The following section concerns traits, or elements of people's personalities—both others' personalities and your own. Please be as honest as possible in answering these questions. They deal with traits you may have more or less of, as well as how important you believe it is to show these traits to the outside world. First, please indicate how well you think the following adjectives might describe you : describes me well does not describe me successful fun-loving responsible artistic materialistic intelligent cool confident attractive 10. Next, indicate the extent to which you want to show these traits to other people (both people you already know and people you don't): an important quality not an important quality to show others to show others successful ftin-loving responsible artistic materialistic intelligent cool confident attractive 151 1 1 . Finally, are there some traits you wish you could do a better job of showing to other people (both people you already know and people you don't)? For each of the following traits, indicate whether you would like to better convey that trait to others: / need to show I do not need to show this trait to others this trait to others successful fiin-loving responsible artistic materialistic intelligent cool confident attractive 152 Questionnaire Two 1 First, . we would like to get your reactions to the movie you just saw. Please put a mark in the box which best indicates whether you agree with the following statements about the movie: I strongly I strongly agree disagree I liked the movie. The characters were interesting. The story was believable. The movie held my attention. Before I watched the movie, I was asked to look for ways to express my personality to others. 2. The segment you just saw was from the movie "Reality Bites." Had you ever seen this movie before today? If so, about how many times? 3. Did you notice any brand-name products being used in the movie? If so, please list those brand-name products below. Take about one minute. (If you later think of any other products that you saw in the movie, however, do not come back to this section; we only want to know which brand-name products you remembered easily.) PLEASE CONTINUE AND DO NOT TURN BACK TO ANY PREVIOUS PAGES. 153 4. Next, we would like to know whether you think some specific brand-name products appeared in the movie clip. Please check one box for each of the following, depending upon how sure you are that the brand-name product was or was not in the movie clip: was not in not was definitely the movie sure - in the movie Chevrolet Audi Snickers Milky Way Budweiser Rolling Rock Diet Coke Diet Pepsi BMW Ford Saab 5. Next, we would like your impressions of the movie's main characters. Please indicate below which traits describe particular characters well and which do not: trait: "successful" describes that does not describe that character well character well 'Lelaina" (Winona Ryder) Troy" (Ethan Hawke) 'Michael" (Ben Stiller) Vickie" (Janeane Garofalo) Sammy" (Steve Zahn) trait: "responsible" describes that does not describe that character well - character well "Lelaina" (Winona Ryder) "Troy" (Ethan Hawke) "Michael" (Ben Stiller) "Vickie" (Janeane Garofalo) "Sammy" (Steve Zahn) 154 trait: "artistic" describes that does not describe that character well character well Lelaina" (Winona Ryder) Troy" (Ethan Hawke) Michael" (Ben Stiller) Vickie" (Janeane Garofalo) Sammy" (Steve Zahn) trait: "fun-loving" describes that does not describe that character well character well "Lelaina" (Winona Ryder) "Troy" (Ethan Hawke) "Michael" (Ben Stiller) "Vickie" (Janeane Garofalo) "Sammy" (Steve Zahn) trait: "cool" describes that does not describe that character well character well "Lelaina" (Winona Ryder) "Troy" (Ethan Hawke) "Michael" (Ben Stiller) "Vickie" (Janeane Garofalo) "Sammy" (Steve Zahn) trait: "intelligent" describes that does not describe that character well character well "Lelaina" (Winona Ryder) "Troy" (Ethan Hawke) "Michael" (Ben Stiller) "Vickie" (Janeane Garofalo) "Sammy" (Steve Zahn) 155 trait: "confident" describes that does not describe that character well character well "Lelaina" (Winona Ryder) "Troy" (Ethan Hawke) "Michael" (Ben Stiller) "Vickie" (Janeane Garofalo) "Sammy" (Steve Zahn) trait: "attractive" describes that does not describe that character well character well "Lelaina" (Winona Ryder) "Troy" (Ethan Hawke) "Michael" (Ben Stiller) "Vickie" (Janeane Garofalo) "Sammy" (Steve Zahn) trait: "materialistic" describes that does not describe that character well character well "Lelaina" (Winona Ryder) "Troy" (Ethan Hawke) "Michael" (Ben Stiller) "Vickie" (Janeane Garofalo) "Sammy" (Steve Zahn) 156 6. We are also interested in how you think people can learn about others from the brand- name products they use. Listed below are several personality traits; please indicate whether you think that, if others knew you used (or saw you using) these various brand- name products, they would be likely to think you have those traits: If I wanted to show others I am successful: the brand would show the brand would show others I am successful just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab If I wanted to show others I am responsible: the brand would show the brand would show others I am responsible just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab If I wanted to show others I am artistic: the brand would show the brand would show others I am honest just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab 157 If I wanted to show others I am fun-loving: the brand would show the brand would show others I amfun-loving just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab If I wanted to show others I am cool: the brand would show the brand would show others I am artistic just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab If I wanted to show others I am intelligent: the brand would show the brand would show others I am intelligent just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab 158 If I wanted to show others I am confident: the brand would show the brand would show others I am confident just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab If I wanted to show others I am attractive: the brand would show the brand would show others I am square just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab If I wanted to show others I am materialistic: the brand would show the brand would show others I am materialistic just the opposite driving a Ford eating a Snickers bar drinking a Diet Coke driving a BMW drinking a Rolling Rock driving a Saab 159 7. Please give your opinion about the following brands . Place a mark in the box that most closely matches your opinions on each brand: good - - bad Chevrolet Audi Snickers Milky Way Budweiser Rolling Rock Diet Coke Diet Pepsi BMW Ford Saab I like it I dislike it Chevrolet Audi Snickers Milky Way Budweiser Rolling Rock Diet Coke Diet Pepsi BMW Ford Saab t 1 1 1 1 1 160 / definitely I definitely do not intend to buy it intend to buy it v^-ncvroici LJn LJn LJn LJn LJn LJn LJn Audi LJn LJn LJn LJn LJn LJn LJn LJn n n n n n LJn iviiiKy Way LJn LJn LJn LJn LJn nLJ LJn n n n n n n n Duuwcibcr LJ 1 1 LJ LJ LJ ( 1 1 1 1 — — — — I— — Rolling Rock u U u u u u Diet Coke Diet Pepsi BMW Ford Saab / would buy it I would not buy it ifI could even ifI could Chevrolet Audi Snickers Milky Way Budweiser Rolling Rock Diet Coke Diet Pepsi BMW Ford Saab THANK YOU FOR YOUR HELP WITH THIS STUDY! REFERENCES Aaker, David A. and Alexander L. Biel (1993), "Brand Equity and Advertising: An Overview," in Brand Equity & Advertising: Advertising's Role in Building Strong Brands, David A. Aaker and Alexander L. Biel, eds., Hillsdale, NJ: Lawrence Erlbaum, 1-8. Adamson, Deborah (1996), "Ads in Sheep's Clothing," New York Daily News, September 15,Bl-2. Advertising Age (1994), "Let's Make Shows Again" (editorial), October 3, 16. (1993), "Ad Time Giveaways" (editorial), July 12. (1991), "Social Messages Urged in Product Placements," November 25, 28. (1990), "Lawsuit drills Fox, Cato," December 3, 3. Adweek (1992), "Great ldea--Not" (editorial), September 28, 22. Adweek's Marketing Week (1988), "Moviegoer Study Finds 20% Look for Brand Names," August 29. Agins, Teri (1994), "Is It a TV Show? Or Is It Advertising?" The Wall Street Journal^ August 15, Bl. Arbuckle, James L. (1997), Amos Users ' Guide Version 3.6, Chicago, IL: SmallWaters Corporation. Arkin, Robert M. and James A. Shepperd (1990), "Strategic Self-presentation: An Overview," in The Psychology of Tactical Communication, Michael J. Cody and Margaret L. McLaughlin, eds., Clevedon, England: Multilingual Matters Ltd., 175-193. Amdorfer, James B. (1997), "007's Sponsor Blitz," Advertising Age, July 28, 24. Babin, Laurie A. and Sheri Thompson Carder (1996), "Viewers' Recognition of Brands Placed Within a Film," International Journal ofAdvertising, 15, 140-151. 161 162 Backman, Carl W. (1985), "Identity, Self Presentation, and the Resolution of Moral Dilemmas: Towards A Social Psychological Theory of Moral Behavior," in The Selfand Social Life, Barry R. Schlenker, ed., New York, NY: McGraw-Hill, 261- 292. * Baker, Michael J. and Hazel A. Crawford (1995), "Product Placement," unpublished working paper, Department of Marketing, University of Strathclyde, Glasgow, Scotland. Balasubramanian, Siva K. (1994), "Beyond Advertising and Publicity: Hybrid Messages and Public Policy Issues," Jowr^za/ ofAdvertising, 23 (4), 29-46. (1991), "Beyond Advertising and Publicity: The Domain of Hybrid Messages," Report No. 91-131, Cambridge, MA: Marketing Science Institute. Bandura, Albert (1977), Social Learning Theory, Englewood Cliffs, NJ: Prentice-Hall. Basil, Michael D. (1996), "Identification as a Mediator of Celebrity Effects," Journal of Broadcasting and Electronic Media, 40, 478-495. Bearden, William O., Thomas N. Ingram, and Raymond W. LaForge (1995), Marketing: Principles & Perspectives, Chicago, IL: Irwin. Beike, Denise R. and Steven J. Sherman (1994), "Social Inference: Inductions, Deductions, and Analogies," in Handbook ofSocial Cognition, Vol. 1 (2nd ed.), Robert S. Wyer, Jr. and Thomas K. Srull, eds., Hillsdale, NJ: Lawrence Erlbaum, 209-285. Bloom, Paul N., Julie Edell, and Richard Staelin (1995), "Criteria for Assessing Research on the Effects of Marketing Communications," Report No. 94-123, Cambridge, MA: Marketing Science Institute. BoUen, Kenneth A. (1989), Structural Equations with Latent Variables, New York, NY: John Wiley & Sons. Boyce, Susan and Alexander Pollatsek (1992), "Identification of Objects in Scenes: The Role of Scene Background in Object Naming," Journal ofExperimental Psychology: Learning, Memory, and Cognition, 18 (3), 531-543. Brissett, Dennis and Charles Edgley (1990), "The Dramaturgical Perspective," in Life As Theater, Dennis Brissett and Charles Edgley, eds.. New York, NY: Aldine de Gruyter, 1-46. Bunish, Christine (1995), "Creative Casting," Producer (June), 62-71. . 163 Busch, Anita M. (1996), "Shoe Flap Trips Up Pic Tie-in," Variety, December 31, 1. Buss, Arnold H. and Stephen R. Briggs (1984), "Drama and the Self in Social Interaction," Journal ofPersonality and Social Psychology, 47 (6), 1 3 1 0- 1 324. Caro, Mark (1996), "Of all the Jim Beam joints...," Chicago Tribune, August 20, CI. Cialdini, Robert B., John F. Finch, and Maralou E. De Nicholas (1990), "Strategic Self- presentation: The Indirect Route," in The Psychology of Tactical Communication, Michael J. Cody and Margaret L. McLaughlin, eds., Clevedon, England: Multilingual Matters Ltd., 194-206. and Maralou E. De Nicholas (1989), "Self-Presentation by Association," Journal ofPersonality and Social Psychology, 57 (4), 626-63 1 Coleridge, Samuel Taylor (1817), Biographia Literaria, Oxford: The Clarendon Press. Colford, Steven W. (1990), "Tobacco Group Ends Paid Placements," Advertising Age, December 17,31. Cook, Thomas D. and Donald T. Campbell (1979), Quasi-Experimentation: Design & Analysis Issuesfor Field Settings, Boston, MA: Houghton Mifflin. Cooley, Charles H. (1909), Social Organization, New York, NY: Charles Scribner's Sons. Crisafulli, Chuck (1995), "It's a Wrap (But Not Plain)," Los Angeles Times, September 3,4. ft Curtis, James (1996), "Taking Pride of Place," Marketing, November 28, 55. Darlin, Damon (1995), "Junior Mints, I'm Gonna Make You a Star," Forbes, November 6, 90-94. Davis, Mark H., Jay G. Hull, Richard D. Young, and Gregory G. Warren (1987), "Emotional Reactions to Dramatic Film Stimuli: The Influence of Cognitive and Emotional Empathy," Journal ofPersonality and Social Psychology, 52 (1), 126- 133. Debevec, Kathleen and Jean B. Romeo (1992), "Self-Referent Processing in Perceptions of Verbal and Visual Commercial Information," Jowr«a/ ofConsumer Psychology, 1 (1), 83-102. Deighton, John, Daniel Romer, and Josh McQueen (1989), "Using Drama to Persuade," Journal ofConsumer Research, 16 (1), 335-343. 164 DeLorme, Denise E., Leonard N. Reid, and Mary R. Zimmer (1994), "Brands in Films: Young Moviegoers' Experiences and Interpretations," paper presented to the 1 994 Conference of the American Academy of Advertising, Tuscon, Arizona. DePaulo, Bella M. (1992), "Nonverbal Behavior and Self-Presentation," Psychological Bulletin, 111 (2), 203-243. Diener, Betty J. (1993), "The Frequency and Context of Alcohol and Tobacco Cues in Daytime Soap Opera Programs: Fall 1986 and Fall \99\r Journal ofPublic Policy & Marketing, 12 (2), 252-260. Dortch, Sharmon (1996), "Going to the Movies," American Demographics, December, 4-7, 22-23. Eisenstein, Paul A. (1997), "Vehicles That Really Perform," WorldTraveler, January, 18-23. * Electronic Media (1991), "Products play central role in TV series," January 21, 40. Elliott, Stuart (1997), "The Spot on the Cutting-Room Floor," The New York Times, February 7, Dl. e (1992a), "Product Placement Is Under New Attack," The New York Times, September 2, D4. (1992b), "Presenting '15 Best Ideas' for Ads That Work Better," The New York Times, January 9, D17. Englis, Basil G., Michael R. Solomon, and Anna Olofsson (1993), "Consumption Imagery in Music Television: A Bi-Cultural Perspective," Journal ofAdvertising, 22 (4), 21-33. e Entertainment Law and Finance (1993), "A Sample Product Placement Agreement," July, 4. Fawcett, Adrienne Ward (1993), "Free TV 'Ad Plugs' Are On the "Risq;' Advertising Age, July 12,21. Fishbein, Martin and Icek Ajzen (1975), Belief Attitude, Intention, and Behavior: An Introduction to Theory and Research, Reading, MA: Addison-Wesley. Forgas, Joseph P. and Stephanie Moylan (1988), "After the Movies: Transient Mood and Social Judgments," Personality and Social Psychology Bulletin, 13 (4), 467-477. 165 Fraser, Jill Andresky (1992), "Keeping Score: From an 'A' for 'X' To an 'F' for 'Bolero'," The New York Times, November 29, H19. Friedman, Monroe (1991), ^ 'Brand' New Language, New York, NY: Greenwood Press. (1986), "Commercial Influences in the Lyrics of Popular American Music of the Postwar Era," Journal ofConsumer Affairs, 20 (2), 193-213. (1985), "The Changing Language of a Consumer Society: Brand Name Usage in Popular American Novels in the Postwar Era," Journal ofConsumer Research, 1 1 (3), 927-938. Friestad, Marian and Peter Wright (1994), "The Persuasion Knowledge Model: How People Cope with Persuasion Attempts," Journal ofConsumer Research, 21 (4), 1-31. Giacalone, Robert A. and Paul Rosenfeld, eds. (1991), Applied Impression Management, Newbury Park, CA: Sage. Goffman, Erving ( 1 959), The Presentation ofSelf in Everyday Life, Garden City, NY: Doubleday. Goldman, Kevin (1995), "Nissan Plots Perilous Pathfinder Campaign," The Wall Street Journal, September 12, B16. Harmetz, Aljean (1983), "Fox to Sell Right To Plug Goods in Films," The New York Times, December 20, CI 9. Hass, R. Glen and Kathleen Grady (1975), "Temporal Delay, Type of Forewarning, and Resistance to Influence," Journal ofExperimental Social Psychology, 1 1, 459- 469. Kazan, Anna Russo, Helene Levens Lipton, and Stanton A. Glantz (1994), "Popular Films Do Not Reflect Current Tobacco Use," American Journal ofPublic Health, 84 (6), 998-1000. Hewitt, John P. (1991), Selfand Society, Needham Heights, MA: AUyn and Bacon. Hoffner, Cynthia and Joanne Cantor (1991), "Perceiving and Responding to Mass Media Characters," in Responding to the Screen, Jennings Bryant and Dolf Zillmarm, eds., Hillsdale, NJ: Lawrence Erlbaum, 63-101. 166 Holbrook, Morris B. and Mark W. Grayson (1986), "The Semiology of Cinematic Consumption: Symbolic Consumer Behavior in 'Out of Africa'," Journal of Consumer Research, 13 (1), 374-381. J. Walter Thompson USA (1989), "Products as Props in Movies and Television," New York, NY: Media Resources and Research. Jacobs, A. J. (1996), "Mr. Rogers' Rotisserie," Entertainment Weekly (December 20), 16-17. James, William (1890), The Principles ofPsychology, Vol. 1, Cambridge, MA: Harvard University Press. Jensen, Jeff (1994), "Reebok Flexes Programming Muscle," Advertising Age, May 23, 12. Johnson, Bradley (1997), "Rosie Disposition for Giveaways," ^c/verrwrng ^ge, March 17, 8. Jones, E. E. and S. Berglas (1978), "Control of Attributions About the Self Through Self- Handicapping Strategies: The Appeal of Alcohol and the Role of Underachievement," Personality and Social Psychology Bulletin, 4, 200-206. Kanner, Bemice (1993), "All the Right Movies," New York, November 29, 20-21. Kardes, Frank R (1994), "Consumer Judgment and Decision Processes," in Handbook of Social Cognition, Vol. 2 (2nd ed.), Robert S. Wyer, Jr. and Thomas K. Srull, eds., Hillsdale, NJ: Lawrence Erlbaum, 399-466. Karrh, James A. (1995a), "Brand Placements in Feature Films: The Practitioners' View," in The Proceedings ofthe 1995 Conference ofthe American Academy of Advertising, Charles S. Madden, ed.,Waco, TX: Hankamer School of Business, Baylor University, 182-188. (1995b), "Informing Audiences About Brand Placements: Effects on Memory and Attributions," presentation to the 1995 Public Policy & Marketing Conference, Atlanta, Georgia. (1994), "Effects of Brand Placements in Motion Pictures," in The Proceedings ofthe 1994 Conference ofthe American Academy ofAdvertising, Karen Whitehill King, ed., Athens, GA: Henry W. Grady College of Journalism and Mass Communication, 90-96. King, Thomas R. (1993), "'Demolition Man' Trades Tacos for Pizza Abroad," The Wall Street Journal, December 2, Bl. 167 Kinnear, Thomas C. and Ann R. Root (1988), "The FTC and Deceptive Advertising in the 1980s: Are Consumers Being Adequately Protected?" Journal ofPublic Policy & Marketing, 7, 40-48. Kuntz, Mary, Joseph Weber, and Heidi Dawley (1996), "The New Hucksterism," Business Week, July 1, 76-84. a Kuntzman, Gersh (1996), "Plugola!" New York Post, September 8, 32-33. Lackey, William Benjamin (1993), "Can Lois Lane Smoke Marlboros? An Examination of the Constitutionality of Regulating Product Placement in Movies," The University ofChicago Legal Forum, 275-292. Lannon, Judie (1993), "Asking the Right Questions: What Do People Do with Advertising?" in Brand Equity & Advertising: Advertising 's Role in Building Strong Brands, David A. Aaker and Alexander L. Biel, eds., Hillsdale, NJ: Lawrence Erlbaum, 163-176. Leary, Mark R. and Robin M. Kowalski (1990), "Impression Management: A Literature Review and Two-Component Model," Psychological Bulletin, 107 (1), 34-47. Levy, Sidney J. (1959), "Symbols For Sale," Harvard Business Review, 37 (4), 1 17-124. ^ Liebenson, Donald (1990), "Look Familiar? Brand Names Stake a Claim in Film," The Chicago Tribune, December 6, 5, 8. o Lynch, John G. and Thomas K. Srull (1982), "Memory and Attentional Factors in Consumer Choice: Concepts and Research Methods," Journal ofConsumer Research, 9 (4), \S-37. c Marks, Peter (1995), "Casting Calls for Towels, Toys, and You-Name-It On the Stage," The New York Times, Mdi\f 2%, WS. McCracken, Grant (1993), "The Value of the Brand: An Anthropological Perspective," in Brand Equity & Advertising: Advertising 's Role in Building Strong Brands, David A. Aaker and Alexander L. Biel, eds., Hillsdale, NJ: Lawrence Erlbaum, 125-139. (1989), "Who Is the Celebrity Endorser? Cultural Foundations of the Endorsement Process," Journal ofConsumer Research, 16(1), 310-321. Messinger, Sheldon E., Harold Sampson, and Robert D. Towne (1990), "Life as Theater: Some Notes on the Dramaturgic Approach to Social Reality," in Life As Theater, Dennis Brissett and Charles Edgley, eds., New York, NY: Aldine de Gruyter, 73- 84. . 168 Miller, Mark Crispin (1990), "End of Story," in Seeing Through Movies, Mark Crispin Miller, ed., New York, NY: Pantheon Books, 186-246. Morgan, Curtis (1993), "Everyone's a Critic," The Miami Herald, February 23, El © Morrison, Margaret (1994), "Behind the Screen: A Qualitative Analysis of the Differences Between In-Home and In-Theater Viewing of Movies," presentation to the annual conference of the Association for Education in Journalism and Mass Communication, Atlanta, Georgia. Muehling, Darrel D. and Michelle McCann (1993), "Attitude Toward the Ad: A Review, Journal of Current Issues and Research in Advertising, 15 (2), 25-58. p and Russell N. Laczniak (1992), "An Examination of Factors Mediating and Moderating Advertising's Effect on Brand Attitude Formation," Journal of Current Issues and Research in Advertising, 14 (1), 23-34. Nebenzahl, Israel D. and Eugene Secunda (1991), "Consumers' Attitudes Toward Product Placement in Movies," International Journal ofAdvertising, 12 (1), 1-11 Newsweek (1983), "Selling It at the Movies," July 4, 46. * Ong, Beng Soo and David Meri (1994), "Should Product Placement in Movies Be Banned?" Journal ofPromotion Management, 2 (3/4), 159-175. r Owen, Anne and James A. Karrh (1996), "Video News Releases: Effects on Viewer Recall and Attitudes," Public Relations Review, 22 (4), 369-378. • Pardun, Carol J. and Kathy Brittain McKee (1996), "What Advertising Agency Media Directors Have to Say About Placing Clients' Products in Motion Pictures," presentation to the annual conference of the Association for Education in Journalism and Mass Communication, St. Louis, Missouri. Pener, Degen (1997), "Sneaky Business," Entertainment Weekly, January 24, 16-17. Petty, Richard E. and John T. Cacioppo (1979), "Effects of Forewarning of Persuasive Intent and Involvement on Cognitive Processes and Persuasion," Personality and Social Psychology Bulletin, 5 (2), 173-176. o Pope, Kyle (1994), "Product Placements Creep Into Video Games," The Wall Street Journal, December 5, Bl. 0 Reed, J. D. (1989), "Plugging Away in Hollywood," Time, January 2, 103. . ^ 169 Reeves, Byron and Seth Geiger (1994), "Designing Experiments That Assess Psychological Responses to Media Messages," in Measuring Psychological Responses to Media Messages, Annie Lang, ed., Hillsdale, NJ: Lawrence Erlbaum, 165-180. Reis, Harry T. (1982), "An Introduction to the Use of Stuctural Equations: Prospects and Problems," in Review ofPersonality and Social Psychology. Vol. 3, Beverly Hills, CA: Sage, 110-132. Risen, James (1989), "The Stars Are Cars," The Los Angeles Times, May 29, IV, 5. Ross, Chuck (1997), "New TV Show Producer Offers Advertiser hiput," Advertising Age January 27, 4. Rothenberg, Randall (1991), "Critics Seek F.T.C. Action on Products as Movie Stars," The New York Times, May 3 1 , D 1 • Sabherwal, Shonall, Jim Pokrywczynski, and Robert Griffin (1994), "Brand Recall for Product Placements in Motion Pictures: A Memory-Based Perspective," paper presented to the annual conference of the Association for Education in Journalism and Mass Communication, Atlanta, Georgia. Sallot, Lynne M. (1993), "The Effects of Motive, Communication Style, and Licensing on the Reputation of Public Relations: An Impression Management Perspective," doctoral dissertation. University of Florida, Gainesville. o Sapolsky, Barry S. and Lance Kirmey (1994), "You Oughta Be In Pictures: Product Placements in the Top-Grossing Films of 1991," in The Proceedings ofthe 1994 Conference ofthe American Academy ofAdvertising, Karen Whitehill King, ed., Athens, GA: Henry W. Grady College of Journalism and Mass Communication, 89. Savan, Leslie (1996), "Your Show of Shills," Time, April 1, p. 70. Sawyer, Alan G. (1975), "Demand Artifacts in Laboratory Experiments in Consumer Research," Journal ofConsumer Research, 1 (2), 20-30. Schlenker, Barry R. (1980), Impression Management, Belmont, CA: Brooks/Cole. and Mark R. Leary (1982), "Audiences' Reactions to Self-Enhancing, Self- Denigrating, and Accurate Self-Presentations," Journal ofExperimental Social Psychology, \%,^9-\0^. and Michael F. Weigold (1992), "Interpersonal Processes Involving Impression Regulation and Management," Annual Review ofPsychology, 43, 133-168. 170 and (1989), "Goals and the Self-Identification Process: Constructing Desired Identities," in Goal Concepts in Personality and Social Psychology, Lawrence A. Pervin, ed., Hillsdale, NJ: Lawence Erlbaum, 243-290. e Shapiro, Michael A. (1994), "Think-Aloud and Thought-List Procedures in Investigating Mental Processes," in Measuring Psychological Responses to Media Messages, Annie Lang, ed., Hillsdale, NJ: Lawrence Erlbaum, 1-14. » Shuptrine, F. Kelley (1975), "On the Validity of Using Students as Subjects in Consumer Behavior Investigations," Journal ofBusiness, 48, 383-390. Skowronski, John J. and Donal E. Carlston (1989), "Negativity and Extremity Biases in Impression Formation: A Review of Explanations," Psychological Bulletin, 105, - 131-142. and (1987), "Social Judgment and Social Memory: The Role of Cue Diagnosticity in Negativity, Positivity, and Extremity Biases," Journal of Personality and Social Psychology, 52 (4), 689-699. Snyder, C. R. (1985), "The Excuse: An Amazing Grace?" in The Selfand Social Life, Barry R. Schlenker, ed., New York, NY: McGraw-Hill, 235-260. Snyder, Mark (1987), Public Appearances, Private Realities, New York, NY: W. H. Freeman and Company. and Kenneth G. DeBono (1985), "Appeals to Image and Claims About Quality: Understanding the Psychology of Advertising," Journal ofPersonality and Social Psychology, 49 (3), 586-597. Snyder, Steven L. (1992), "Movies and Product Placement: Is Hollywood Turning Films into Commercial Speech?" University ofIllinois Law Review, 301. e Solomon, Michael R. (1983), "The Role of Products as Social Stimuli: A Symbolic Interactionism Perspective," Journal 1 ofConsumer Research, 0 ( 1 ), 3 1 9-329. and Basil G. Englis (1994a), "Reality Engineering: Blurring the Boundaries Between Commercial Signification and Popular Culture," Journal ofCurrent Issues and Research in Advertising, 16 (2), 1-17. — and (1994b), "The Big Picture: Product Complementarity and Integrated Coxnm\xmcdii\om,'' Journal ofAdvertising Research, 33 (1), 57-63. 171 Stapel, Diederik A., Willem Koomen, and Joop van der Pligt (1997), "Categories of Category Accessibility: The Impact of Trait Concept versus Exemplar Priming on Person Judgments," Journal ofExperimental Social Psychology, 33, 47-76. Steortz, Eva Marie (1987), "The Cost Efficiency and Communication Effects Associated with Brand Name Exposure Within Motion Pictures," unpublished master's thesis. West Virginia University, Morgantown. Stewart, David W. and Ingrid M. Martin (1994), "Intended and Unintended Consequences of Warning Messages: A Review and Synthesis of Empirical Research," Journal ofPublic Policy & Marketing, 13(1). Swann, William B., Jr., John J. Griffin, Jr., Steven C. Predmore, and Bebe Gaines (1987), "The Cognitive-Affective Crossfire: When Self-Consistency Confronts Self- Enhancement," Journal ofPersonality and Social Psychology, 52 (5), 881-889. Tan, Alexis S. (1986), "Social Learning of Aggression fi-om Television," in Perspectives on Media Effects, Jennings Bryant and Dolf Zillmann, eds., Hillsdale, NJ: Lawrence Erlbaum, 41-55. Tedeschi, James T. and Nancy Norman (1985), "Social Power, Self-Presentation, and the Self," in The Selfand Social Life, Barry R. Schlenker, ed., New York, NY: McGraw-Hill, 293-322. Thomas, Karen (1996a), "Lighting Up: Tobacco Has a Role in Most Movies," USA Today, November 7, 6D. (1996b), "No Waiting to Inhale: Cigarettes Light Up Movies," USA Today, November 7, ID. Thorson, Esther (1990), "Consumer Processing of Advertising," Current Issues and Research in Advertising, Ann Arbor, MI: Graduate School of Business Administration, University of Michigan, 197-230. The Toronto Star (1991), "Centre tries to pull the plug on commercialism," December 16, CI. Turcotte, Samuel (1995), "Gimme a Bud! The Feature Film Product Placement Industry," unpublished master's thesis. The University of Texas at Austin. Turow, Joseph (1978), "Casting for TV Parts: The Anatomy of Social Typing," Journal ofCommunication, 28 (4), 18-24. . 172 Vollmers, Stacy M. and Richard W. Mizerski (1994), "A Review and Investigation into the Effectiveness of Product Placements in Films," in The Proceedings ofthe 1994 Conference ofthe American Academy ofAdvertising, Karen Whitehill King, ed., Athens, GA: Henry W. Grady College of Journalism and Mass Communication, University of Georgia, 97-102. The Wall Street Journal (1997), "Costs of Producing, Promoting Movies Increased 9% in 1996," March 5. Wanta, Wayne and Yu-Wei Hu (1994), "The Effects of Credibility, Reliance, and Exposure on Media Agenda-Setting: A Path Analysis Model," Journalism Quarterly, 7\,90-9S. Warner, Fara (1995), "Why It's Getting Harder to Tell the Shows From the Ads," The Wall Street Journal^ June 1 5, B 1 Wells, William D. and Kendra L. Gale (1995), "Fictional Subjects in Consumer Research," in Advances in Consumer Research, Vol. 22, Frank R. Kardes and Mita Sujan, eds., Provo, UT: Association for Consumer Research, 306-310. (1989), "Lectures and Dramas," in Cognitive and Affective Responses to Advertising, Patricia Cafferata and Alice M. Tybout, eds., Lexington, MA: D. C. Heath and Company, 13-20. Wicklund, Robert A. and Peter M. Gollwitzer (1982), Symbolic Self-Completion, Hillsdale, NJ: Lawrence Erlbaum. Wimmer, Roger D. and Joseph R.Dommick(\997), Mass Media Research: An Introduction, Belmont, CA: Wadsworth. Wired (\994), "Is Advertising Finally Dead?" February, 71. Wooten, Leah M. (1995), "Brand Placement in Movies: The Consumer's Perspective," unpublished master's thesis. College of Journalism and Communications, University of Florida, Gainesville. Wriite Radio Adsong Information Page (1996), accessed from the World Wide Web at http://soho.ios.com/~wriite/adsng.html. BIOGRAPHICAL SKETCH James A. Karrh was bom in Swainsboro, Georgia, in 1963. Karrh completed his undergraduate studies in finance at the University of Florida in Gainesville, Florida in 1984, earned a master's degree in business administration at Duke University in Durham, North Carolina in 1987, and earned a master's degree in marketing at the University of Florida in 1993. Later that year, he began doctoral studies in mass communication at the University of Florida. Karrh was general manager and primary owner of WSKS-FM and WKGQ-AM radio stations for three years, and was an associate with Bankers Trust Company for two years. He has taught advertising and media management at the college and university levels for the past four years. In August 1 996, Karrh began a position as assistant professor in the Roy H. Park School of Communications at Ithaca College in Ithaca, New York. As of August 1 998, we will begin an appointment as assistant professor in the College of Communications at the University of Alabama. He resides with his wonderfial wife, the former Alison Knott, in Tuscaloosa, Alabama. 173 1 it conforms to I certify that I have read this study and that in my opinion acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Michael F. Weigold, Chai^^ Associate Professor of Journalism and Communications it conforms to I certify that I have read this study and that in my opinion acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Maiy^tjin Ferguson Professor of Journalism and Communications it conforms to I certify that I have read this study and that in my opinion acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philpsophy. y Kent Lancaster Professor of Journalism and Communications it conforms to I certify that I have read this study and that in my opinion acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Richard Lutz Professor of Marketing it conforms to I certify that I have read this study and that in my opinion acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. DeBo^ah M. Treise Associate Professor of Journalism and Communications This dissertation was submitted to the Graduate Faculty of the College of Journalism and Communications and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 1998 Ql'-^^^^^-Ji-^ Dean, (Mlege of Journalism andanc Communications Dean, Graduate School