|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 THE EFFECTS OF AFFECTIVE RESPONSES TO SPORTS MEDIA-SOURCE CONTEXT ON ADVERTISING EVALUATIONS By MICHAEL JOHN CLAYTON A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Michael Clayton
3 This document is dedicated to my family. Their continued confidence in my abilities and dreams provided the peace of mind to help me through al l the challenges the last three years have presented.
4 ACKNOWLEDGMENTS I would like express m y sin cere thanks and appreciation to those individuals who provided me the inspiration and tools to complete this journey. I would like to thank my chair and experimental design mentor, Dr. Michae l Weigold. His confidence in my ability and willingness to let me walk on my own allowed me the opportunity to grow as a researcher. In addition to his guidance, he ha s had a profound impact on the sight s I have set for myself as a scholar, an educator, and a me ntor for future students. I would also like to thank Dr. Debbie Treise for her dedication to myself and all her doctoral candidates. Her tireless effort provided me invalu able advice and guidance throughout my doctoral studies and career search. I will al ways be amazed and inspired by her passion and dedication to the field, her students, and to life. I would like to thank Dr. Jon Morris for al lowing me to get my feet wet as a teaching assistant, and being patient as I adapted to these new responsibilities. I would also like to thank Dr. Kyriaki Kaplanidou for her enthusiasm for my research topic and constant encouragement throughout the dissertation and job s earch process. Lastly, I am eter nally grateful to all of my professors and peers who contributed to my in tellectual and personal growth throughout this process.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........7LIST OF FIGURES.........................................................................................................................8ABSTRACT.....................................................................................................................................9 CHAP TER 1 OVERVIEW...........................................................................................................................10Theoretical underpinnings of this research............................................................................. 13Research Question and its Importance................................................................................... 142 LITERATURE REVIEW.......................................................................................................16Program Induced Affect (PIA)............................................................................................... 16Dimensions of Mood..............................................................................................................19Dimensions of Emotion.......................................................................................................... 21Theoretical Frameworks......................................................................................................... 23ELM Central vs. Peripheral Processing........................................................................ 23Mood Congruency and Consistency Effect models........................................................24Sports Involvement.................................................................................................................26Spectators v. Fans............................................................................................................28Fan Motivations...............................................................................................................30Team Identification as an Antecedent of Involvement.................................................... 31Psychological Models...................................................................................................... 33Models and Scales in Relati on to TV Viewing of Sports................................................35Program Liking (Content Involvement).......................................................................... 35Advertising Effectiveness.......................................................................................................37Ad Recall.........................................................................................................................37Attitude Toward the Ad (Aad) and Attitude toward the Brand (Ab).............................. 37Conceptualization of Constructs for this Study............................................................... 38Hypotheses......................................................................................................................403 METHODOLOGY................................................................................................................. 42Research Design.....................................................................................................................42Sample and Team Selection.................................................................................................... 43Product Category and Brand Selection................................................................................... 46Experimental Procedure......................................................................................................... .47Measurement.................................................................................................................... .......49Pre-exposure Questionnaire............................................................................................. 49Post-Exposure Questionnaire.......................................................................................... 49
6 4 RESULTS...............................................................................................................................525 DISCUSSION AND CONCLUSIONS.................................................................................. 61Affective Responses Created by Sports Programming...........................................................61Attitude findings.............................................................................................................. .......62Limitations.................................................................................................................... ..........63Future Research......................................................................................................................65Conclusions.............................................................................................................................67APPENDIX A INFORMED CONSENT DISCLO SURES AND QUESTIONNAIRE ................................. 68B ACRONYM SUMMARY...................................................................................................... 75REFERENCES..............................................................................................................................76BIOGRAPHICAL SKETCH.........................................................................................................85
7 LIST OF TABLES Table page 2-1 Motivations identified in sport spectating..........................................................................314-1 One-Way ANOVA results for involvement by condition.............................................. 534-3 One-Way ANOVA results for pleasure by condition.................................................... 554-4 One-Way ANOVA results for arousal by condition...................................................... 564-5 One-Way ANOVA results for dominance by condition................................................ 564-6 One-Way ANOVA results for Aad by condition........................................................... 574-7 One-Way ANOVA results for Ab by condition............................................................. 584-8 Summary of Hypothesis Findings...................................................................................... 60
8 LIST OF FIGURES Figure page 2-1 Proposed Model of Fan Identification on PIA and Ad Effectiveness................................39
9 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE EFFECTS OF AFFECTIVE RESPONSES TO SPORTS MEDIA-SOURCE CONTEXT ON ADVERTISING EVALUATIONS By Michael John Clayton May 2009 Chair: Michael Weigold Major: Mass Communication This research contributes to the theoretical knowledge within the field of PIA (programinduced affect) and has practical implications for sports marketers and advertisers. An experiment was conducted to explore the ability of sports to create bi polar affect responses among highly identified fans of co mpeting teams. The experiment supported previous research in sports marketing regarding the power of sports to create strong affective responses. The research failed to identify any link between these affective responses a nd attitude formation towards advertisements and brands present in this media.
10 CHAPTER 1 OVERVIEW Sport broadcasts have been popular advertisi ng vehicles for advertisers for years, but recen tly these programs have grown in popularity due to the lack of time shifting that occurs with this programming; and consequently fewe r consumers are using Digital Video Recorders (DVRs) to skip commercials during these broa dcasts (Atkinson 2007). This increased demand has led to an increase in CPMs (cost per thousan d) for sports programs and sponsorships, as well as greater difficulty in securing these media properties, due to the large upfront commitments to become a sponsor of many leagues, tournaments and series, including NASCAR, the NFL, Olympics, and major college athletics. Mintel (2007) reports that spor ts advertising spending surpassed $27 billion in 2006. Advertising revenue for ESPN alone exceeded $1.1 billion for 2005, nearly 40% more than the closest cable competitor, and a 26.4% increase from the previous year (Advertising Age Supplement 2/28/2005). Anheuser-Busch, one of the nations largest advertisers, recently dedicated over $200 m illion a year to national TV sports (Heistand 2004). According to IEG, marketers spent $13.4 billion on sport sponsorships in 2006, up from $6.4 billion in 2001 (Mintel 2007). While consumers interest in sports varies, Mintel consumer research reports that there are 94.4 million Casua l sports fans, 36 million Serious sports fans, and 13.4 million Obsessed sports fans. Despite these large numbers, sport marketing still offers a form of narrowcasting due to th e characteristics shared by fans of different sports (Burnett, Mennon, and Smart 1993). While sports like profe ssional football have very broad audiences, less mainstream sports (i.e. bass fishing and tr actor pulling) have very homogeneous groups of fans which are highly sought afte r by certain brands. Clearly, spor t as a media vehicle represents a huge market that deserves the attention of scholars and practitioners alike. Scholars have identified sport as a natural fit for sponsorship due to the strong imagery and broad appeal across countries and social classes (Gwinner and Swanson 2001). Others have
11 suggested that advertisers can positively leverage the exciteme nt and emotional attachment viewers have with sports, in order to increa se effectiveness (Copeland, Frisby and McCarville 1996). Gas stations with monitors at their pumps have recently begun showing highlights from the weeks top-viewed games, hoping to leverage the emotional response and create additional sales (Mintel, 2007). According to Roese and Maniar (1997), for many people, sports spectating represents one of the most pass ionate and intense of human endeavors, utterly dominating affect and cognition for short periods of time (p. 1245). As advertisers and scholars, its important for us to understand the unique challenges and opportun ities presented by sports as media vehicle. The dearth of research on sports as a media vehicle, specifically as it relates to the program-induced affect (PIA) of this genre of media on advert ising evaluations, is surprising given the magnitude of the investments being made in the realm of sport sponsorships and corresponding advertising expenditures. PIA resear ch suggests that advertisements tend to be more effective when consumers have more positiv e attitudes toward the context in which the ad is presented. While most viewers are expected to exhibit positive l iking of the sport programming they view, little is known regard ing the events outcome on PIA, among highly identified fans. While not all viewers have a vest ed interest in every sporting event they watch, there are a large number of viewers who have a st rong emotional tie to specific teams either as a fan, fantasy sport participant, or in a legal or illegal gambling cont ext. The very nature of most sporting events is that you will have fans and ga mblers on both sides of the fence, and therefore mixed emotions with the outcome of every game, and perhaps every play. As stated by Madrigal (2003), sporting events are uniqu e in that one teams achieveme nts come at the expense of a competitor (p. 25). Whereas a drama may have a decidedly happy or positive tone that is consistent among the majority of viewers, fans moods and emotions to ward sports programming
12 are likely to experience wider, bipolar disparities based on the play of each team on the field and the final outcome. Sloan (1979) was among the first to study thes e dual emotions in spectators, caused by wins and losses. Using football, basketball, and boxing spectators this study found that positive emotions (happy, satisfied, pleased) were create d when an individuals favorite team won and negative emotions (angry, frustrated, sad) wh en their team lost. Gantz and Wenner (1995) expanded on this research to include the te levised sport viewing experience and found that while watching, fans were more emotionally involved, feeling nervous and, depending on the flow of action on the field, either happy or angry with the performances they watched (p. 70). Studies involving PIA are largely absent in the realm of sports. De Pelsmacker, Geuens, and Anckaert (2002) de fine media context as the characteristics of the content of the medium in which an ad is inserted (e.g., articles in a magazine, spots in a television program), as they are perceived by the persons who are exposed to them (p. 49). Research on media context has increased as more researchers have recognized contextual factors as key moderators of advertising effectiven ess. The most common conceptualizations of advertising effectiveness have included ad reca ll, attitude toward th e ad (Aad) and attitude toward the brand (Ab) (Aylesworth and MacK enzie 1998; Coulter 1998; Faseur and Geuens 2006; Norris and Colman 1993; Pa velchak, Antil and Munch 1988). A rare study involving sport context effects on advertising was conducted by Hill (2005), which examined the effects of sport context on pull-through advert isements (branding on-screen during a program instead of during a commercial br eak). The results of this study indicated that Aad scores were elevated for evaluations of pull-through advert ising when exciting incidents were happening on screen, as opposed to dead (s toppage in play) and action (on-going play but no imminent scoring) sequences. The same st udy also concluded that recall was significantly
13 higher in the action condition than either the exciting or dead conditions. Unfortunately, these research results appear limited to pull-thro ugh advertisements, but it reaffirms the affective components involved in sports, and their potential affects on advert ising effectiveness. Theoretical underpinnings of this research Several theories in consumer behavior a nd mass communication will be examined in this research, most notably mood congruence theo ry, consistency effect theory, and moodmaintenance repair theory. In addition this paper will examine attitude towa rd the ad (Aad) as a theoretical mediator of attitude toward the brand (Ab), and measur e of advertising effectiveness. Due to the unique nature of sport this research will expand on existing PI A research to treat both forms of affective responses, moods and emotions. Sport fan involvement will be thoroughly reviewed in order to aid in the conceptualizat ion of fans. Lastly, the Elaboration Likelihood Model (ELM) will be reviewed to understand how fan and non-fans may process sport broadcasts and commercials differently. Kamins, Marks and Skinner (1991) explained Mood Congruency theory in terms of PIA and advertising: Mood Congruency theory would imply that embedding either upbeat or depressing commercials in the context of a happy program would lead to more positive cognitions about the commercials, to the ads being rated as more effective, and to higher purchase intention compared to embedding them in a sad program. (p. 3) In the same article the authors introduced a negative-state relief model labeled the Consistency Effect result. In this model the authors posit that whe n sad mood states are matched by stimuli, the stimuli will be perc eived as more positive than inconsistent stimuli would be (p. 4). The results of th eir study supported the Consistency Effect model that a happy commercial is more posi tively evaluated in the context of a happy program, whereas a sad commerci al excelled in the context of a sad program. This model
14 is slightly different from mood-maintenance mo dels which posit that people are generally motivated to maintain positive moods a nd repair negative moods (Martin 2003). Sport content provides an interesting ch allenge to existing schemes as it is expected that sport actually produces bipolar affective responses within a single television program, newspaper article, etc., among fans of the winning and losing teams, when there is a victor and loser, or cont ent involving positive or negative news. Thus, advertisers can not apply Consistency Effect to this programming by matching the tone of their advertisements with that of the context since the latter will actually create pronounced, bipolar affective responses am ong the most involved viewers. So, the question remains of whether or not Mood Congruency holds tr ue to sports programming and whether fans of the losing team have negative affective responses which are detrimental to advertisers as represented by diminished advertising evaluations among this segment. Research Question and its Importance This paper seeks to further understand: a) the role of m edia context in creating affective responses among highly involved fa ns, specifically the creation of bipolar affect based on the outcome of the sport contest; and b) the effect these affective responses have on advertising effectiveness, specifically Aad a nd Ab, among highly identified fans. Given the magnitude and continued grow th of spending in sports programming, advertisers need to better unders tand the risks and consequences of this spending in different sports, among different types of viewers, fans and non-fans. If, in fact, sport content has the ability to create negative affective states among fa ns, and if those affective states are detrimental to the performance of advertisi ng, then marketers and media planne rs would have to reevaluate
15 the way in which advertisers use sport content. Of course future effects research would need to be conducted to determine how fans of different sports differ, as suggested by Sloan (1979): Just as there may be different theories for different sports, there may be different fans for different sports. People may vary in the source of their attraction to sports and needs they seek to satisfy by playing or watching. (p. 256) Possible implications for marketers may include reconsideri ng which sports to sponsor, markets to target, and the best timing for commercials to air in live broadcasts, such as the beginning of the game when the outcome is still in doubt and both sets of fans are experiencing optimism and pleasure. The following four chapters will address this topic in greater detail, starting with the second chapter which will provide a thorough literature review of exta nt research in the areas of PIA theories, moods and emotions, attitude formati on relevant to advertising, as well as models and scales used to study fans and fan behavior. At the conclusion of the second chapter I will introduce the hypotheses formed from this litera ture review, which will be addressed by this study. The third chapter will detail the experimental design of the current study including the procedure, sampling, stimulus, and manipulations. The fourth chapter will detail the results of the study and the statistical conf irmation or disconfirmation of the stated hypotheses. The fifth and final chapter will provide an overview of the findings of this research, will acknowledge limitations of the study, and will offer conclusions and recommendations regarding future research on this topic.
16 CHAPTER 2 LITERATURE REVIEW Sport spons orship has been studied from several areas, most notably from the goodwill, image transfer standpoint where research has fo und a positive transference from sponsors to advertisers and brands (Meenaghan 2001). Lybe rger and McCarthy (2001) studied attitude changes from the 1998 to the 2000 Super Bowls and found a growing apathy toward sponsorship of the Super Bowl, but its unclear if this is unique to the excessive co mmercial nature of this one sporting event, or has more far reaching implications. Subsequent research comparing attitude toward advertising through sport and non -sport advertising found mo re positive attitudes associated with advertising through sport (Pyun 2 006). However, research to date has failed to explore the bipolar nature of moods and emoti ons created by sport content among fans, and said effect on advertising effectiveness. In order to address this complex quest ion, a literature review was completed in each of the following areas: affect (moods and emotions), the central and peripheral routes of pro cessing, extant PIA theories, conceptua lization and operati onalizations of fanship, and lastly advertising effectivene ss measures (ad recall, Aad, and Ab). Program Induced Affect (PIA) Gardner (1985) took the phenomenological appr oach in defining moods as feeling states that are subjectively perceive d by individuals (p. 282). Conve rsely, Gardner distinguished emotions as being more intense, attention-getti ng, and tied to a specifiab le behavior/object. She goes on to state that an indivi dual is more aware of ones em otions than ones mood. This delineation is consistent with Madrigals (2003) de finition of emotions as the direct result of a subsequent evaluation and interp retation of a subsequent evalua tion and interpretation in which an actual state is compared to a desired state (p. 26). An interesting elemen t of this definition is the choice of language which includes a largel y cognitive element. The relationship between
17 cognition and affect becomes even more pronounced as we later examine the prevalence of cognition in self report measures for affect. Common definitions of the term affect tend to include emotions, feelings, drives, as well as moods (Batra and Ray 1986; Gardner 19 85). For a review of the confusion between affect and emotions see Holbrook and OSh aughnessy (1984). Advertising and consumer behavior research also frequently use the term emotional response. Stout and Leckenby (1986) defined emotional response as a response to so me psychologically important event, real or imagined, past or anticipated a nd state that an emotional respons e exhibits valenced feelings occurring as reactions to self-relevant events (p. 36). Th e authors go on to state that these emotional responses vary depending on the indi viduals ability to make progressively selfrelevant connections to the specific event, person or situation (p. 36). In addition, to emotional response we frequently see the term affective response. The attachment of the word response to affect seems to direct the constr uct further away from mo ods and more directly towards categorization with emo tions, due to the attribution of the response to a specific object. This relevance will be discussed later on in this chapter in regards to the involvement an individual has with a specific sp orting team or individual athlete, as well as sport content. While most program-induced mood research focu ses solely on feeli ngs, there is some evidence that emotions are a critical element of sports programming consumption. McKinley Jr. (2000) stated that: Some researchers have found that fervent fans become so tied to their teams that they experience hormonal surges and other physiological changes while watching games, much as athletes do. The self-esteem of some male and female fans also rises and falls with a games outcome with losses affecting their optimism about everything from getting a date to winning at darts. (p.A1) Pavelchak et al. (1988) studied the rela tionship among program context, emotional experience, and ad recall in a study of Supe r Bowl XX viewers from the winning and losing
18 cities, as well as a neutral ci ty. The study found that those viewers in the winning and losing cities demonstrated lower ad re call than those in the neutral ci ty. The study accounted for arousal and pleasure separately, and found th at arousal affected ad recall, while pleasure had no affect. Thus, viewers in each Super Bowl city experien ced high arousal, but while the pleasure varied based on the outcome of the game, it was posited th at the arousal of high levels of emotion in both winning and losing cities led to the decline in ad recall, thus dem onstrating the Intensity theory. The Intensity theory pos its that high arousal levels le ad to less peripheral processing. Newell, Henderson, and Wu (2001) conducted a sim ilar study as Pavelchak et al. (1988), and were unable to replicate the results using S uper Bowl XXXIV. While the authors hypothesized that fans of the winning and losi ng teams would recall fewer ads due to the arousal created by the game, their results failed to support this hypothesis, despite demonstrati ng that arousal levels were significantly elevated for fans of both teams. One explanation for this could be that no two sporting events are exactly alike. Tobar (2006) studied the effects of s port fandom, utilizing the Sport Fandom Questionnaire (SFQ), on viewers of Super Bowl XL, and found that higher levels of sport fandom and greater enjoyment of advertisements were signi ficantly related to purchase intentions. Mood and enjoyment were also incl uded to understand the influence of affect on purchase intentions, but these affective states we re never directly linked with preference for a particular team, merely overa ll ratings as a sport fan. From a non-sports context Gorn, Pham, and Sin (2001) examined the interplay between valence (pleasant or unp leasant) and arousal (low or high) by dissecting existing mood states (manipulated independently) and affective t ones of advertising. Th is study supported an excitation transfer hypothesis where consumers a ffective states polarize ad evaluations in the direction of the ads affectiv e tone in high versus low arous al scenarios. The affect-as-
19 information framework actually posits that consumers knowingly engage in this behavior as they perceive these feelings to contain valuab le judgmental information (Pham et al. 2001, p. 167). Clearly affective states play a critical role in the consump tion of sports programming and need to be accounted for conceptually. Both c onstructs (moods and emotions) are likely to vary based on an individuals affective investment in the event and participants. Hillman found that ardent football fans at the University of Flor ida experienced extreme physiological arousal when they viewed pictures of Gator football st ars making game-winning plays, but responded indifferently to pictures of other athletes and teams (McKinley Jr. 2000, p. A1). Therefore, a study measuring the effectiveness of advertising in a sports cont ext must measure all affective consequences of the programming context, including mood and emotions, among fans and nonfans. Madrigal (2003) examined spectator affect during live sporting events and found several important antecedents and consequences in relation to attribution-dependent emotions, defined as the praiseworthy and blameworthy actions of th e event participants. Antecedents included goal relevance and affective expectations, while cons equences were satisfac tion and perceptions of entertainment value. Dimensions of Mood Mood has traditionally been conceptualiz ed based on two dim ensions: valence and arousal (Russell and Barrett 1999; Shapiro, MacInnis, and Pa rk 2002), though Aylesworth and MacKenzie (1998) argue that mood is defined pr imarily in terms of valence. Valence is commonly defined in terms of pleasantness (posit ive versus negative) and arousal (high versus low) in either physiological te rms, as degree of energization, activation, inner tension, or alertness, or psychological terms, as a state of wakefulness or action prepar ation (Shapiro et al.,
20 2002, p. 16). While Aylesworth and MacKenzie (1998) examined valence and controlled for arousal, Shapiro et al. (2002) found that valence affects schema or data driven processing, and that negative valence stimulates processing of data. Conversely, high arousal was shown to inhibit ad processing. This resear ch supported the findings of Pave lchak et al. (1988) that higher levels of arousal reduced consum ers recall of ads. The conceptu alization of arousal was similar among the two studies, both classifying arousal as either a function of mood or emotion. As will be discussed later with emoti ons, several scholars have challenged the unidimensional construction of arousal and pleasure. Gardner (1985) identified several dimensions of positive moods (e.g., cheeriness, peacefulness, and sexual warmth) as well as multiple negative moods (e.g., anxiety, guilt, and de pression) and postulated that the effects of negative moods are more complex than those of positive moods, due to the heterogeneity of negative mood states. Holbrook and OShaughnessy (1984) identified several key differences that existed among commonly used affective constr ucts in research. These distin ctions included whether they are: active or reactive, specific or general, and acute or chronic. Moods were largely classified as being general, reactive, and acute. In laymans terms moods are short lived reactions to an environment that lack attribution to a single source. For example, after getting in your car in the morning you have a quicker than usual drive to work, hear your favorit e song on the radio, get a good parking spot, and the weather is pristine. S o, for no specific reason you enter the office in a good mood. After easing into your day your boss notif ies you that youre not getting a raise this year. In an instant, your feelings have ch anged, and you know exactly why; which leads to emotions.
21 Dimensions of Emotion Em otions have been a hot topic in consumer behavior research for quite some time. Some of the seminal work by Mehrabian and Russe ll (1974) produced the PAD scheme around the dimensions of pleasure, arous al, and dominance. Havlena and Holbrook (1986) examined Mehrabian-Russells (1974) three dimension PAD pa radigm along with Plut chiks (1980) eight dimension scheme including; fear, anger, joy, sadness, disgust, acceptance, expectancy, and surprise. This study conducted a multi-level analysis comparing reliabilities, convergent validities, and external generalized abilities of each. The rese arch suggested that within the context of consumption experiences, the Plutchik scheme provided no additional information over the simpler PAD schema. Since the PAD scheme has become a widely accepted framework for studying emotional responses based on the thr ee independent, bipolar dimensions (Chebat, Filiatrault, Gelinas-Chebat, and Vanisky 1995; Eroglu, Machleit, and Davis 2003; Havlena and Holbrook 1986). Lang (1980) made a significant c ontribution to the PAD scheme with the introduction of the Self-Assessment Manikin (SAM). This visual self-report instrument has been shown effective in measuring emotional respon ses to advertising (M orris et al. 2002) and successful in cross-cultural research and studies with children (Morris 1995). In addition, Morris et al. (2002) showed the superi ority of visual self-report aff ective measures over cognition in predicting behavior. Scholars have also suppor ted the simplification of the PAD scheme by eliminating the dominance dimension which has been shown to be less significant than pleasure and arousal (Olney, Holbrook, and Batra 1991; Russell, Weiss and Mendelsohn 1989). In researching multiple emotional responses within a single advertising execution, Morris and McMullen (1994) found that levels of arousal and pleasure could be al tered within a single commercial, but no results were found for dominance. The current research is distinct from the latter, in that it seeks to identify different emotio nal responses within a single advertisement, but
22 as a consequence of audience segmentation ba sed on fan identificati on, not based on creative elements of an advertisement. In recent years scholars ha ve questioned the valence-base d approach to feelings and tested multidimensional views of feelings (Bab in, Darden, and Babin 1998; Faseur and Geuens 2006; Mitchell et al. 2001; Raghunathan and Pham 1999). Mitchell et al. st udied context-induced feelings and found that negative constructs such as anger and sadness produced different ad effects when using a neutral ad. Raghunathan and Pham examined anxious, sad, and neutral feelings in relation to decision making regard ing gambling and job selection. The authors findings indicated that anxiety and sadness convey distinct inform ation to the decision-maker and subsequently effects decision making, with anxiety leading to uncertainty reduction and sadness leading decision makers to seek reward replacement. Faseur and Geuens found support for feelings of coziness, excitement, and romance producing different effects on ad evaluations. Lastly Babin, Darden and Babin found that emotions are not always bipolar, meaning that feeling a negative emotion does not preclude the occurrence of a positive emotion. An additional noteworthy study on affectiv e response is Batra and Rays (1986) examination of affect typologies. The authors categorized their findings from nine studies, including their own, into thirteen unique categories: Interest/Expectancy Surprise Disgust/Scorn Skepticism Anger Fear/Anxiety Shame Guilt Sadness Surgency, Elation, Vigor/Activation (SEVA) Deactivation
23 Social Affection Drives Of the thirteen categories identified in th e literature, three positive affective response categories (SEVA, deactivation, and social affecti on) were operationalized. The empirical results of this study were that all three affective responses (ARs) were an tecedents of attitude toward the ad (Aad). Two studies in sports, Pavelchak et al. ( 1988) and Newell et al. (2001), examined the presence of arousal and pleasure among vi ewers of the Super Bowl (XX and XXXIV respectively) in the cities of the participating teams and a neutral city. Neither study distinguished between moods and emotions, but both studies found significantly higher scores of arousal and pleasure among respondents from the participating cities. Neither study examined the impact of the dominance dimension, or th e effects of arousal and pleasure on attitude formation. Unfortunately, both studies measured th ese emotions through self reports the day after the event, thus requiring respondents to recall previously held emotions in a post-evaluation scenario. Theoretical Frameworks ELM Central vs. Peripheral Processing Mood and persuasion literature commonly supports the belief that positive moods decrease central processing and negative m oods increase it (Aylesworth and MacKenzie 1998; Bagozzi, Gopinath and Nyer 1999; Gardner an d Hill 1988; Mackie and Worth 1989). Assuming motivation, ability and opportunity exist, Pe tty and Cacioppos (1981) Elaboration Likelihood model (ELM) has also been applied to explain how positive versus negative mood states lead one to process advertising centrally or periphera lly. Aylesworth and MacK enzie (1998) posit that consumers in a negative mood state need only pr ocess the source of the mood to resolve this state, thus motivating consumers to process a show centrally, but not the ad. Conversely, the
24 authors propose that positive moods fail to trigger a problem-solving mode, therefore leading to the show and ad being processed centrally. The hypothesi s that divergent moods lead to disparate routes of processing could presen t an interesting challenge to a dvertisers if sports content is shown to succeed in creating these bipolar moods. ELM is also relevant to the current resear ch as one examines the arousal and emotions experienced by fans when watching their favor ite team(s). As previously mentioned, the Intensity theory purports the notion that high ar ousal levels lead to less peripheral processing, and several studies have supported this theory within non-sport contex ts (e.g. Mackie & Worth 1989) and within sports (e.g. Pavelchak et al. 1988). Mood Congruency and Consistency Effect models In general, mood congruency m odels have supported the notion that mood states influence evaluations, judgments, and behavi ors in mood congruent directions (for a comprehensive review of 40 studies on mood s ee Gardner 1985; also Goldberg and Gorn 1987; and Kamins et al. 1991) Quite simply, positive moods prime positive memories and negative moods, negative emotions. Another tenant of the mood-congruency hypothesis is that mood states prime the recall of memories sharing a similar aff ective valence (Martin 2003). Expanding mood congruence theory, studies have supported that the context of a specific program creates a mood that transcends the program and advertisement, and effects attitude toward the ad (Aad) and subsequently the brand (Ab) (Lutz, MacK enzie and Belch 1983; Mackenzie and Lutz 1989; Coulter 1998). Early research in the field was limited solely to the mood of the media context, and suggested that positive program-induced fee lings led to more positive evaluations of Aad (Goldberg and Gorn 1987). Further research expanded on mood congruence by studying the relationship between various context produced moods and commercial messaging and context mood. Specifically, research on medi a-source effects evolved to include greater understanding of
25 Consistency Effects between context mood and a dvertisement mood (De Pelsmacker et al. 2003; Goldberg and Gorn 1987; Kamins et al. 1991). Research in this area generally supports the concept of happy commercials being most effectiv e in the context of ha ppy programs, and sad commercials being most effective in the contex t of sad programming, both of which have been linked to elevating ad and bra nd evaluations. To date, no studie s were found that address the creation of bipolar affective respon ses within a single media context. Scholars in cognitive psyc hology and consumer research agree that the two core dimensions of congruency include relevancy and expectancy (Dahlen 2005; Goodman 1980; Heckler and Childers 1992). Heck ler and Childers (1992, p. 477) define relevancy as material pertaining directly to the meaning of the them e and reflects how information contained in the stimulus contributes to or detracts from the cl ear identification of the theme or primary message being communicated and expectancy as the degree to which an item or piece of information falls into some predetermined pa ttern or structure evoked by the th eme. Specifically as it results to advertising in a sports cont ext, relevancy may include adver tisers using sports imagery to appeal to the interests and tastes of viewers, while expectancy c ould relate to the appropriateness of the products being advertised (i.e., viewers of football may expect to see ads for athletic apparel, beer and sports drinks but may not expect to see ads for diapers or feminine hygiene products). While these determinants of congruenc y are likely to hold rele vance to sport-context programming and advertising, most scholars in the field of media cont ext have studied mood congruency solely between the media-source a nd the mood of the advertisement (i.e., happy media-context with happy ads). The moods produ ced by a sport program are predicted to be bipolar and therefore more difficult for adver tisers to hold congruent with the mood of the advertisement. So, while the mood of the ad may have relevancy and expectancy to the program,
26 the judgment of the mood of the media-source ma y witness greater variance than in previous media context research. Mood-regulation motivation may have a role in the processing of advertising in a sports context if the mood created by the program is ne gative and the viewer wishes to improve their mood. Chang (2006) recently found that in some situations attentive e xposure to entertaining advertising can reduce negative mood evoked by the media. So, perhaps fans experiencing negative emotions will look toward the adver tising to repair their current mood states. While many elements of commercials have been studied in order to predict effectiveness, one of the simplest classifications is of inf ormational or emotional appeals. While this dichotomy between cognition and affect is a useful simplification tool, it fails to account for the wider array of messaging approaches used by ad vertisers. Message cont ext research and mood congruency theory have expanded to study the various dimensions of program moods and commercial moods to better predict effectiveness. Sports Involvement Im agine two individuals sitting in their re spective homes watching a live sporting event on television. One sits nearly still, showing few physiological signs of arousal or pleasure, the other shows visible signs of nervousness and anger, including occasional verbal outbursts directed at an unresponsive televi sion set. Clearly, one would a ssume that these two individuals differ in terms of their respective involvement with the drama unfolding on the screen. We know from the Petty and Cacioppo Elaboration Likelih ood Model (ELM) that involvement effects the manner in which we process information, and subse quent studies have detailed the importance of involvement because of its latent capacity to in fluence attitudes toward an activity or product, and behavior in regards to said activity or product (e.g. Arora 1985). These studies led to a robust stream of research on involvement, in cluding program induced affect and leisure
27 involvement in recreation and tourism contexts, the latter of which spawned more focused studies on sport fan involvement. Shank and Beasley (1998) defined sports involvement as being the perceived interest in and personal importance of sports to an individua l (p. 436). This definition was born from work by Zaichkowsky (1985), which sought to develop a scale to measure involvement based on the definition of a persons perceived relevance of the object based on inherent needs, values, and interests (p 342). Havitz and Dimanche (1990) identified three invol vement distinctions common in involvement literature: enduring v. situational, emoti onal v. rational, and personal v. impersonal Enduring involvement, while continuously present, may still experience dramatic shifts over time. While an alumnus of George Mason University may have always considered himself a fan of the basketball team and watche d the games and standings with interest, that individuals involvement with the team ma y have reached a crescendo during the schools 2006 run to the Final Four. Other alumnus who neve r followed the basketball team as students or alumnus, likely experien ced a high level of situational involvement with the basketball team during the 2006 tournament, only to be repressed until the next time the school makes a big splash on the national scene. As expresse d by Havitz and Dimanche (1990) enduring involvement is still extremely unstable over the c ourse of a year (in-season and off-season), as well as the anticipation, experience, and recollection of a given ev ent or game. What categorizes involvement for fans is that enthusiasm and interest for the team is ongoing. The next types of involvement, being emotional and rational, also have fairly obvious implications to fans of a participant versus fa ns of specific sport content. While golf fans may enjoy watching tournaments on TV, they may not have been as emotionally touched by the 2000 Sony Open tournament as a Paul Azinger fan whom is more familiar with his entire career and his battle against non-Hodgkin lymphoma. Lastly, there is personal involvement. In the case of
28 Azinger, we may find that someone who has pers onally battled this disease, or lost someone close to them, may be more profoundly affected while watching Azinger achieve his first victory after beating the disease. The mo st highly identified fans are e xpected to demonstrate enduring, emotional, and personal involvement. Laurent and Kapferer (1985) found, involvement could not be measured through a single index, but instead must account for antecedent co nditions of involvement: perceived importance of product or situation, sign value, pleasure value, and risk (divided into two subcomponents). This finding led the authors to determine that the term involvement alone was too imprecise without further explanation, and subsequently led to several distinctions of sport fan involvement and identification over the last 20 years. Trail, Anderson and Fink (2000) presented a theoretical model to account for sport spectator consumption behavior. Of the six factor s identified, the first tw o hold great significance to the current study, these being, mo tives and levels of identificat ion. These factors were posited to function sequentially and to lead to expectan cies, confirmation or disconfirmation, self-esteem responses, and the affective state of the individual, which would in fluence future fan behavior. In the forthcoming pages motives and identification will be examined, but first, a delineation between fans and spectators is required. Spectators v. Fans While involvem ent may manifest itself in terms of program-lik ing of a genre of programming, the involvement, or affinity towards a specific team also needs to be considered as a variable effecting arousal and valence of emotions. Sherif a nd Hovland (1961) introduced the concept of involvement in the field of psychol ogy through their work on social judgment theory. Nearly two decades later Petty and Cacioppos ELM dealt with involvement from a decision making process viewpoint. Recently, sport management literature has addressed sport fan
29 involvement more in terms of a psychological st ate, typically multidimensional in nature, and consisting of affective, behavioral, and cognitive dimensions. Havitz and Dimanche (1997) who studied over 50 leisure involvement data sets, defined leisure involvement as an unobservable state of motivation, arousal or inte rest toward a recreati onal activity or associated product. It is evoked by a particular stimulus or situati on and has drive proper ties (p. 246). While involvement is a common construct in consumer behavior literature, sport management literature has sought to delineate that of sports fans and ha s routinely utilized fansh ip as the construct for discussing an audiences affinity to, or involvement with a sport or team. The previously mentioned Mintel report classified fans as Casual, Serious, and Obsessed sports fans. While scholars have sought to explore a similar classification of fanship, no consensus appears in the literat ure regarding the nomenclature of similar groupings for varying levels of fans. Conceptually, Gantz and Wenner (1995) defined fa nship in terms of perceived knowledge about sports, interest in viewing televised sports, a nd amount of televised sports viewed (p. 61). Guttman (1986) defined fans as emotionally committed consumers of sporting events (p. 6). Trail et al. (2003) and Robinson et al. (2005) rec ognized the lack of delineation between the terms spectators a nd fans within the extant lite rature, and developed a scheme to differentiate the two based on motivational factors and points of attachment. As stated by Trail, Anderson and Fink (2000) fans are usually spectators, however not all spectators are fans (p. 157). Gantz and Wenner (1995) actually go as far as suggesting that with fans viewership is likely to be activ e and participatory (p.57). Other adjectives have been used by scholars as modifiers for different levels of fanship, including avid and extreme (Capella 2002), and die-hard and fair-weather (Wann and Branscombe 1990), but no dominant scheme currently exists within the literature. Sutton, McDonal d, and Milne (1997) presented an additional framework of fan identification and distinguished three leve ls of fanship: low
30 identification (social fans), medium identification (focused fans), and high identification (vested fans). While the authors adequately distinguis hed between the motives and behaviors of each group, the authors did not present an operationali zation of the three cons tructs, or specific methods of measurement. Several studies have examined the antecedents and dimension of fan involvement in a sport context (e.g. Funk, Ridinger, Moorman 2004) while others have utilized involvement profiles for classification purposes of fans (e .g. Kerstetter and Kovich 1997). Recently, sport fanship literature tends to focus on either motivations for spectating (Wann 1995; Trail and James 2001), points of attachment with a te am (Heere and James 2007; Kwon, Trail, and Anderson 2005; Wann and Branscombe 1993; Wann and Dolan 2001), a combination of the two (Robinson and Trail 2005; Trail et al. 2003), or psychological models (Funk and James 2001). Fan Motivations Wann (1995) introduced a for mative study in the development of a scale for the application of measuring motivations of spor ts fans. This scale, labeled the Sport Fan Motivational Scale (SFMS), has be en extensively used and tested now for more than a decade and has spawned similar scales including: Miln e and McDonalds (1999) Motivations of Sport Consumers scale, Trail and James (2001) Motiva tion Scale for Sport Consumption (MSSC), and the Funk et al. (2001) Sport Interest Inventory (SII). Due to the innovativeness of the original SFMS study, the scales content validity has been called into question (Trail and James 2001). The original SFMS did report str ong internal reliability for the eight motivations of spectator attendance identified through extant literature: eustress, self-e steem, escape, entertainment, economic, aesthetic, group affiliation, family; th ough eustress and self-esteem loaded on the same factor. The Milne and McDonald, Trail and James, and Funk et al. studies examined nine, twelve, and ten items respectively, though later work on the Funk et al. SII scheme added four
31 additional motives (see Table 2-1 for an overv iew of motives identified and supported through these four models). All four scales measuring motives were developed specifica lly to those of attendees at live sporting events, and sometimes even specific sports. The same can be said for Kerstetter and Kovichs (1997) simpler Invo lvement Profile (IP) which examined five dimensions: importance, pleasure, risk consequen ce, risk probability, and sign value. The four motivational scales are clearly relegated to m easuring attendance motiva tion, and while the IP scale is a better fit with mediated sport content, the two risk dimensions appear to have greater relevance to live sport attendance. Table 2-1. Motivations iden tified in sport spectating Wann (1995) Milne and McDonald (1999) Trail and James (2001) Funk et al. (2001) Funk, Mahony, Ridinger (2002) Sport Fan Motivation Scale (SFMS) Motivations of Sport Consumers scale Motivation Scale for Sport Consumption (MSSC) Sport Interest Inventory (SII) Sport Interest Inventory (SII) revisited eustress risk-taking Achievement drama drama self-esteem stress reduction Knowledge interest in player interest in player escape aggression Aesthetics interest in soccer interest in soccer entertainment affiliation Drama Team identification team identification economic social facilitation Escape socialization socialization aesthetic self-esteem Family aesthetics aesthetics group affiliation competition physical attraction national pride national pride family achievement physical skills excitement excitement skill mastery Social support women's opportunity in sport support women's opportunity in sport aesthetics vicarious achievement vicarious achievement value development role model role model self-actualization wholesome environment entertainment and value bonding with family Team Identification as an Antecedent of Involvement The second significant genre of fan research consists of team identification (or atta chment to the team) literature. Traditionally, sport management scholars have studied identification in two contexts: athletes identificat ion with his or her athl etic role, and a sport fans identification with a team or player (W ann 2002). This study is on ly concerned with the
32 latter. Trail et al. (2000) defined identification as an orientation of the self in regard to other objects including a person or group that results in feelings of sentiments of close attachment (p. 165-166). Wann (1997) defined team identification as the extent that a fan feels psychologically connected to a team (p. 331) and Gwinner and Sw anson (2003) defined the same concept from a social identity perspective as the spectators perceived connectedness to a team and the experience of the teams failings and achieveme nts as ones own (p. 276). This stream of research dates back to early work by Cialdi ni et al. (1976), which is most renowned for contributions to the sport management literature in the field of BIRGing (basking-in-reflected glory), which was a precursor to the forthcoming CORFing (cutting-off-reflected failure) concept (Snyder, Lassegard, and Ford 1986). Utilizing the Sport Spectator Identification Sc ale (SSIS), Wann and Branscombe (1993) found that team identification is an antecedent of involvement, as fans high in identification tend to be more involved and invested in the team. Similar to scale development on motivations of sports involvement, team identification literature has sought to identify the various dimensions of identification, and several studie s have linked these with specta tor motives (Funk, Ridinger, and Moorman 2004; Robinson and Trail 2005; Trail et al. 2003). The orig inal Points of Attachment Index (PAI) developed by Trail et al. (2003) examined seven di fferent points of attachment: player, team, coach, university, community, sport, and level. Their study al so linked these points of attachment with three types of motives and f ound that motives varied by fans and spectators, and whether those individuals identification was associated with a successful or unsuccessful team. Subsequent studies by Kwon, Trail and Anderson (2005) disc ontinued the use of community as this dimension fit less well, presum ably due to the usage of a student sample in many studies. An additional study by Robinson and Trail (2005) examined motives and points of
33 attachment of live sport spectators to intercollegiat e athletics, and f ound that significant differences existed by gender a nd type of sport attended. In addition to the scales previously identified, Heere and James (2007) recently developed a multi-dimensional team identity scal e (Team*ID) based on social identity theory. Their confirmatory factor analys is identified six dimensions of team identity including: public evaluation, private evaluation, inte rconnection of self, sense of interdependence, behavioral involvement, and cognitive awareness. In addition to the sport and fan identification dimensions previously identified, scholars have also identified university identity as a significant mediator of evaluations of a university footba ll team (Dietz-Uhler and Murrell 1999). To date, scholars have successfully linked team identi fication to changes in physiolo gical arousal (Branscombe and Wann 1992), exhibition of physical effects (Sloan 1979), and affect (Madrigal 1995; Wann et al. 1994), among highly identified fans. Gwinner and Swanson (2003) sought to examine the antecedents of team identification in the realm of spectators for a college football t eam, and found that team identification could be predicted by perceived prestige of the university, involvement with the domain of football, and associations with the team/school. Psychological Models A third significant realm of research on involvement of fa ns includes the Psychological Continuum Model (PCM) and Psychological Comm itment to Team (PCT) scale development The PCM introduced by Funk and James (2001), deta ils the parameters in which a relationship between an individual, sport or athlete is mediated. Similar to Hierarchy of Effects Models, the PCM model outlines a four step process beginning with awareness, and then followed by attraction attachment and allegiance. Instead of focusing on beha vioral change, this model focuses on the psychological rela tionship between the fan and th e sport object (e.g. a team or
34 athlete). Within this framework, the previously discussed attachment cons truct is accounted for within a larger conceptual framework. The second scheme, PCT, provided by Mahony, Madrigal, and Howard (2000) relates specifically to the attachment and allegiance dimensions, and is conceptually defined by the authors as a scale to be used in se gmenting sport consumers based on loyalty (p. 15) and to a ssess the strength of an i ndividuals commitment to sport teams (p. 18). The authors further distinguished team loyalty as a two-dimensional construct representing enduring allegiance to a particular team (p. 16). Again, we see enduring as a critical dimension of fanship. Additional sport and leisure studies research distinguished two forms of loyalty, behavioral and attitudinal (Park and Ki m 2000). Attitudinal loyalty is the closer construct to attachment as desc ribed by Park and Kim in the c ontext of recrea tional sport as the process of attaching psychologically to a se lected recreational sport program (p. 198). Here the authors extend the loyalty construct from an organizational commitment model developed by Allen and Meyer (1990), consisting of affec tive, continuance, and normative components. Wann and Pierce (2003) compared the aforem entioned SSIS to the PCT scale, in the context of fan behavior utilizing the Sport Fandom Questionnaire (SFQ). The SFQ, introduced by Wann (2002), represents a measure of ones identification with being a sport fan. Unlike previous identification studies discussed, this construct goes be yond the identifica tion with any single team or player, and instead examines ones self-perceptions as a sp ort fan. Comparing the SSIS and PCT, Wann and Pierce determined that the two measures were highly correlated and appear to assess a similar constr uct, as both scales predicted fa n behaviors as represented by the SFQ measure. The results of this study may imply that loyalty and team identification are in fact similar constructs, at least as operationalized in the two unique unidimensional scales examined. Of these two scales, Wann and Pierce concluded that the SSIS was more strongly related to a general measure of fandom, as represented by the SFQ.
35 Models and Scales in Relation to TV Viewing of Sports The lone use of a scale for television viewing was introduced by Gantz and W enner (1995) and created a fanship index based on an individuals: interest in watching TV sports in general, perceived knowledge of ones favorite sport, exposure to sports programming on the weekends, and exposure to televised sporting events on weekdays. While this study was not focused on scale development, the authors di d introduce a scale for sport fanship based on television sport viewing experiences, which reported a Cronbachs alpha of .74. The fanship index moved beyond existing unidimensional appro aches by accounting for affective, cognitive, and behavioral components. Freque ncy distributions were then us ed to identify respondents in the outer two quartiles, labeled fans and non-fa ns. Other interesting c ontributions from this research included the link between fanship and affective responses to a sport exposure, which determined that fans were more emotionally involved and more invested in the television viewing experience. Program Liking (Content Involvement) Murry, Lastovicka, and Singh (1992) used Gardne rs definition of m oods in a subsequent definition of program liking as a summary evalua tion of the experience of viewing a television program. This definition disti nguishes program liking from affective responses. Thus, a preference for watching sports pr ogramming would indicate a high li king for this material, but would indicate little regarding the affective in vestment in the program, or outcome of such behavior. The authors distinguish program-elic ited feelings from program liking by recognizing these feelings as temporary affective states that are subjectively percei ved by an individual (p. 442). Based on this definition, program-elicited fee lings can be construed as either moods or emotions. So, while two baseball fans watching a Yankees-Red Sox game may both respond consistently regarding their lik ing for viewing MLB baseball on television, the outcome of the
36 events on the field, is posited to move their program-elicited feelings differently and either subtly effecting each viewers mood, or producing mo re noticeable emotional outbursts depending on each individuals affective investment in the two teams competing. Murry and Dacin (1996) elaborated on the re lationship between program-elicited feelings and program liking. The authors postulated that program-elicited positive emotions directly influence program liking, while negative emotions trigger cognitive processing which seeks to interpret how the program may threaten the viewers well-b eing (p. 440). While on the surface one may not associate the viewing of sports, re gardless of the outcome, with a threat to the viewers well being, the common pr actice of gambling on sports may be an exception. Social identity theorists may also argue that when a fans favorite team loses that individuals wellbeing is threatened. Hirt et al (1992) found that afte r a team loss, highly identified fans felt worse about themselves and their own abiliti es. Gwinner and Swanson (2003) described this relationship in that fans highly involved with a team can be extremely lo yal, holding a particular team as central to their identity where team succe ss and failure is interpreted as personal success or failure (p. 277). Murry, Lastovicka, and Singh (1992) concluded th at the liking of a program was a better predictor of an advertisements success than the negative or positive emotion created by the program. Thus, they concluded th at program liking was a better predictor of success than PIA. Schumann (1986) also found that product evaluations (of advertised products) were influenced by program liking. Applying these fi ndings to the topic of sports context media we may find that the liking for sport media should be more instrume ntal in ad effectiveness than program-induced affect, and the outcome of play may have little effect on attitude formation.
37 Advertising Effectiveness While ad recall has been examined as a cons equence of program creat ed affective states, this study seeks to introduce the effect of these affective states on attit ude formation, specifically Aad and Ab as core measures of ad effectiveness, and thus seeks to extend the knowledge within this field of study. Ad Recall Ad recall has been the m ost frequently studi ed dependent variable of sport programming created affective responses (Pav elchak et al. 1988; Newell et al. 2001). Both studies sought to support an intensity arousal theory utilizing Super Bowl viewersh ip in three separate cities (winning team city, losing team city, and neutral city). The tw o studies found conflicting results regarding ad recall, with Pavelchak et al. finding higher recall scores for viewers with no team preference, while Newell et al found no diminished ad recall am ong fans from the participating cities. Both studies however did find support for their initial hypothesis regarding elevated levels of arousal and pleasure for view ers with a team preference. While ad recall is a noteworthy consequence of pleasure and arousal to be studied, this measure remains a front-line measure of advertisi ng effectiveness and fails to facilitate a real understanding of attitudinal effects of fanship. Th is viewpoint is similar to that expressed by Meenaghan (2001) in regards to th e evolution of sport sponsorship research that focuses on event sponsorship recall. Attitude Toward the Ad (Aad) and Attitude to ward the Brand (Ab) While the attitude toward the ad (Aad) construct has received near consensus regarding its mediating role in influencing attitude toward the brand (Ab) (Batra and Ray 1986; Lutz et al. 1983; Mitchell and Olson 1981) and a dvertising effectiveness, scholar s have differed in the core dimensions of the construct (Olney, Holb rook and Batra 1991). While early studies
38 operationalized Aad as a unidimensional global a ffect measure (Lutz et al. 1983; Mitchell and Olson 1981), numerous studies since, have atte mpted to isolate the cognitive or affective elements of a subjects evaluation (e.g. Bruner II 1998). Most commonly, researchers have examined Aad as a multidimensional construct (Batra and Ray 1983; Holbrook 1978; Muehling 1987) and recent scale development has sought to isolate the affective and cognitive dimensions of advertising evaluations. Conceptualization of Cons tructs fo r this Study The literature reviewed thus far le d to the following conceptualizations: Affective responses Due to the uniqueness of sport as a media vehicle, conceptualization must include all affective components of sport-pr ogramming induced feelings, including mood and emotions. Whereas other media contexts may have more subtle effects on moods, the effects of sport spectating is known to produce greater arousal which can be directly attributed to the results of the events on the field of play. Fanship as involvement Within this study there are two objects which will be studied regarding an individuals involvement: involvement with a par ticipant (team or athlete) and involvement with a piece of sport content. Concep tually, this study shall define fanship in terms of fanship toward specific spor t content (Fan Content) and fanship toward specific sport participants (Fan Participant). The term partic ipant was selected to account for fanship toward athletes in individual, as well as team sports. When speaking of Fan Participant, this term will represent an individuals fan or fanship, and the construct of Fan Content shall be similar in nature to program liking, but with a more enduring conceptualization, some what similar to the domain involvement construct identified by Gwinner and Swanson ( 2003). The distinction between content and participant is of critical importance as one delineates between the multidimensionality of the
39 mediators of affective responses. While the term program liking seems to be a natural fit with television programs, Fan Content shall represent all mediated communications (e.g. magazine or newspaper article, or website/webpage) an indivi dual consumes in relation to the object on their fanaticism. Figure 2-1Proposed Model of Fan Iden tification on PIA and Ad Effectiveness The inclusion of mood and emotions as mediators of Aad seems like a logical progression, but other researches who have tried to identify the mediating role of moods have failed in identifying any mediation. Hirst et al (1992) found significant results of fanship and event outcome on moods, team es timates (how the team would do in the future), and self estimates (how the respondent would perform on skills test). Significantly, the authors also found that mood did not mediate the effects of out come on subjects estimates. A second study by Snyder, Lassegard and Ford (1986) added self-esteem as a new variable; and while affected by outcome, the authors found that mood did not si gnificantly predict self-esteem over and above the independent variables. The authors also atte mpted to find a mediating role for mood in the context of CORFing, but their re sults indicated that the postulated mood mediational model did not fit the data.
40 While there are questions about the mediating effects of moods (and in this study emotions) on Aad in a PIA study, the role of program context is more accepted. Mackenzie and Lutz (1989) posited that program context is an antecedent of Aad, and is likely to mediate program context influences on Ab. Hypotheses Whereas m ost media creates affective responses that are fairly consistent and valence congruent among most viewers, sp orts have a uniquely opposed a udience which is expected to experience bipolar affective outcomes depending on an individuals team affiliation and the strength of that relationship (For a review of anteceden ts of spectator affect during a live sporting event see Madrigal 2003). H1 Level of fanship will moderate mood change s when exposed to team specific media regarding their preferred team. H1a The greater the fan involvement, the greater the negative moods following exposure to a losing team effort. H1b The greater the fan involvement, the greater the positive moods following exposure to a winning team effort. H2 Among highly identified fans of particip ating teams, sporting events will create significantly different emotions among fans of the winning and losing teams (or positive versus negative situations). H2a Pleasure will be significantly higher for highly identified fans of winning teams, versus highly identified fans of losing teams. H2b Arousal will be significantly higher for hi ghly identified fans of winning and losing teams, compared to the control group of individuals with low levels of fanship. H2c Dominance will be significantly higher for highly identified fans of winning and losing teams, compared to the control group of individuals with low levels of fanship. H3 Affective states created by the sport programming will affect advertisement evaluations in this media, with Aad and Ab being directly correlated to positive and negative affective states created by the sport media, among highly identified fans. H3a Highly identified fans of winning teams will w itness elevated levels of Aad as compared to highly identified fans of the losing team and non-fans.
41 H3b Highly identified fans of losing teams will report lower Aad scores than highly identified fans of the winning team and non-fans. H3c Highly identified fans of winning teams will witness elevated levels of Ab as compared to highly identified fans of the losing team and non-fans. H3d Highly identified fans of losing teams will re port lower Ab scores than highly identified fans of the winning team and non-fans.
42 CHAPTER 3 METHODOLOGY Research Design While Pavelchak et al. (1988), Newell et al. (2001) and Tobar (2006) utilized consum er selected real-world settings for exposure to various Super Bowls due to the high viewership of these live events, a ll three studies failed to account for opportunity and ability factors which may have inhibited respondents from centrally processing the advertisements. A national consumption study of Super Bowls XXXII and XXXIV by Lyberger and McCarthy (2001) found that 91% of respondents did not plan on watching the Super Bo wl by themselves. Since Super Bowl viewing, especially in Super Bowl cities, is often categorized by parties and social gettogethers, the presence of add itional distractions may have likely inhibited the respondents opportunity and ability to proce ss the advertisin g. While these environments had unquestionable external validity strengths, an experimental design is preferred for this study in order to control for all stimuli and to measure mood and emotions closer to the point of stimulation and not in post-evaluation scenarios like the previous studies. Unfortunately the very nature of sport makes it difficult to recreate in a lab environmen t. As Madrigal (2003) detailed, sporting events are a type of live, unscripted pe rformance consumed by spectators either in person or via media (p. 25). An additional limitation of previous work by Pavelchak et al. (1988) and Newell et al. (2001) was the nature of the timing delayed asse ssments, used within these evaluations, which required subjects to respond with stored evaluations, and therefore may have been biased by mood states at the point of retrie val, or the ability to accurately retrieve previously experienced moods. For this study, an experimental design with three conditions was employed with sport outcomes (win/loss/neutral or control) as the experimental conditions. Teams winning and losing
43 were chosen to present positive and negative outcomes, but these are by no means the only sport outcomes which results in positive and negative affective responses. All respondents began the experiment by co mpleting a brief questi onnaire containing self-report scales for team identification with th e two college basketball te ams expected to have low fanship among the study sample, due to their lack of geographic proximity to the sample. In addition, global mood evaluations we re collected at this time to reflect mood states upon entering the experiment. Following the completion of the first brief survey, students in the control group were asked to watch a brief clip of a basketball game between the two college basketball teams they just answered questions about re garding their personal identificati on as fans or non-fans of those specific teams. The nearly eight minutes of ba sketball footage was immediately followed by a single commercial for a newly released cell phone. Students in the non-control groups were provid ed additional instructions after completing the first survey. These details included a stipulati on that the participant w ould receive three extra credit points if the team they were randomly assigned to scored more points in the forthcoming clip. Conversely, if their team was outscored th ey would only receive one extra credit point. Assignment to each of the two conditions (w in and loss) was determined by randomly distributing playing cards among a ll participants present from a full and shuffled deck. Those receiving red cards were assigne d to the win condition while t hose receiving black cards were assigned to the loss condition. Sample and Team Selection This study included 298 m ale and female unde rgraduate students enrolled in three different sections of advertising courses at a large Division I south eastern university. The decision to use college students was supported by the subject matter being studied, college
44 sports, which typically has great relevance among this audience. Several previous studies have indicated that the most highly identified fans are those supporting colleg e athletics (Goldstein and Arms 1971; Schurr, Ruble, and Ellen 1985; Zillman et al. 1979), thus undergraduate students at a large public university were deemed an appropriate sample for studying fan behavior. According to Calder, Phillips and Tybout (1981) the ideal theory fals ification procedure, however, is to employ maximally homogeneous respondents. This entails sampling from groups of individuals that are similar on dimensions lik ely to influence the variables of theoretical interest (p. (199-200). In this re search, the homogeneity in the sample of students was expected to be high, thus providing a stable sample for measurement of attitudes towards a sporting event and commercial presence. The most important si milarities included their length of familiarity with cell phones and brand experiences. Since this study is specifically concerned with the affective re sponses and attitude formation of highly identified fans compared to non-fans, it was necessary to compare fans and non-fans under positive and negative situations. In order to accomplish this, subjects were randomly assigned to one of three experimental conditions: (a) high-fanship and positive content; (b) high-fanship and negative c ontent; and (c) control (non-fans, or low-fanship, likely to view the stimuli as either positive or negative). Tr aditionally, fanship could be measured in terms of identification with a specific te am. The current study utilized th is approach, but attempted to manipulate temporary involvement with one of the two teams participati ng in order to create random assignment of involvement as a substitute for actual fanship. Research participants were recruited from tw o different advertising courses. Participants for the win and loss conditions were recruited from two sections of a large introductory course consisting of various majors from several colleges within the university. These participants were recruited during their class and told that if they signed up for an extra credit project they would
45 receive the standard two point s of extra credit commonly offe red by their instructor. These individuals then signed up for one of several ti me slots available for th is experiment. Subjects from this class then entered a single classroom at various dates and tim es within the following ten days. Regardless of which day and time the part icipants chose they were treated to the same experimental procedures. Within every scenari o, participants were being randomly assigned to either the win or loss condition based on a playing card handed to them upon entering the classroom where the experiment occurred. The control group was specifically held at a separate time for several notable reasons. First, it was critical for participants to believe that their extra credit points would vary depending on the outcome of the game. This manipulation wa s critical in creating the involvement needed to replicate fan behavior. If control group subjects were interspersed with the two test conditions, the subjects likely would have had a harder time accepting the multi-faceted classification system and varying levels of ex tra credit. While this may not ha ve lead to hypothesis guessing it may have led subjects to doubt that so many di fferent extra credit levels would actually be offered to a single class. In addi tion, the IRB required that students have the ability to opt-in to a study worth two extra points if th ey were uncomfortable with the gamble of one or three extra credit points. If a group of stude nts within the same experiment were going to be receiving two points they may have asked to be included in this sample, which would have eliminated the randomness of the assignment. The entire control group was taken from a second advertising class within the college who were all gathered together for class and told that they could receive two extra credit points for participation in the study. Since theses subjec ts had already completed the course where the other subjects were collected from, it decrea sed the likelihood that any discussion occurred where an individual might have discussed their assignment to rooting for a specific team based
46 on playing cards. No playing cards were involved in this class as all subjects were included within the same condition. Product Category and Brand Selection An addition al element of the experiment that was carefully considered was the selection of the advertisement to be tested. In order to increase internal validity, a commercial was required that featured a product of relevance to the sample. The commercial chosen for this experiment was a relatively new commerci al for the Samsung Behold smartphone. This commercial was chosen as cell phones represent a pr oduct category that is relevant to the audience. Academic studies have found that over 97% of college students own a cell phone (Gemmill and Peterson 2006) and recent industry reports indicate that this percentage has grown to 99% in the most recent quarter (IBM 2009). A recent Ball State University study also found that 27% of college students owned smartphones, which was 8% higher the national average (DN Online 2009), thus making smart phones a relevant product for the sample used in this study. The commercial was also newly released, whic h was preferred for this study, as it was less likely that partic ipants had already been overexposed to the stimuli, thus allowing the exposure within this experiment to more accura tely measure initial attitudes formed in the context of the current study, instead of prev iously formed attitudes created under previous immeasurable circumstances. The brand selected al so represents a second tier brand within the cell phone market (not one of th e top three manufacturers of thes e products) and thus was less likely to be owned, and therefore less likely to possess the well entrenched attitudes common among products consumers have greater experience with, and therefore had the opportunity to witness variance in attitude formation within the experiment. A more established brand/product (i.e. the Apple iPhone), would have had more strongly held engendered attitudes which would
47 have been more difficult to alter within a labor atory experiment measuring attitudes toward a brand following a specific exposure. Experimental Procedure W ith the permission of instructors, the researcher recruited students through multiple advertising courses. The students were promised two extra credit poin ts for participation. Students then signed a sheet indicating a specific tim e and date to participat e in the study. In the test conditions students entered a classroom and were handed a playing card from a randomly shuffled standard deck of car ds. All test groups were conducte d in the same classroom to maintain internal validity. At the designated start time a note was poste d on the door indicating that the experiment had begun and that late arrivals would not be ad mitted and would need to signup for an alternate date and time. At this time the researcher passe d out a brief questionnaire including pre-exposure questions (see following section for detail on specific items) and an IRB informed consent disclosure which was signed and dated. Once all pre-exposure questions had been comple ted the researcher read new instructions approved by an IRB committee. The new instructions were: Thanks. You will now be shown the beginning of a previously recorded basketball game between Memphis and the University of Ma ssachusetts (UMASS). When you entered the classroom you were randomly given a playi ng card. If you received a BLACK playing card you have randomly been assigned Memphis. If you received a RED playing card you have randomly been assigned the University of Massachusetts (UMASS). In order to make this research more interesting your professor has ag reed to let me adjust his standard 2 point extra credit format. Instead of rewarding ever yone 2 extra credit points for participating in this study, you will earn either 1 or 3 extra credit points based on which team scores more points during this portion of th e game you are about to watch. So if the University of Massachusetts (UMASS) outscores Memphis, red card holders will earn 3 extra credit points while black card holders will earn one extra credit po int. Conversely, if Memphis outscores the University of Massachusetts (UMASS), black card holders will earn 3 point while red card holders will earn one point. At the conclusion of the study today you will need to turn your survey in with your pl aying card to receive extra credit points.
48 If this uncertain outcome makes you uncomfo rtable, or you are uncomfortable with the nature of this experiment you are free to exit the experiment now or at any point and we will be happy to reschedule you into another ex periment worth 2 extra credit points for all participants. Participants were then asked to sign a revi sed IRB informed consent form. There were no instances where an individual was uncomfortable or unwilling to continue. At this point the researcher also instructed the pa rticipants in attendance not to ve rbally express their pleasure or discomfort while watching the video. The video then began to play on a large screen at the front of the classroom. At the conclusion of game footage all subjects were shown the same commercial from the previously identified advertiser. At the conclusion of the video a second questionnaire was distributed containing th e post-exposure questioni ng. All subjects were instructed to remain in their seats afte r completing the second survey. After all subjects had completed the second survey the resear cher read the following statement: Thank you for your participation in this experi ment. The purpose of this experiment was to study the effects of affective st ates created through sports ba sed on affiliation with a team, leading to a positive or negative sports viewi ng experience. In order to accomplish this it was necessary to create a rooting interest in each of the teams participating. In order to be fair to all participants however, everyone w ill receive 3 extra credit points for your class regardless of which team you were randomly assigned. Now that you have learned the true and full purpose of the current study and know about the actual manipulations that took place as part of this study, do you wish to have your data included in this research project? Participants were then asked to si gn a revised IRB form indicating their acknowledgement of the experimental manipulation and their permission to include their results within this study. Again, all pa rticipants agreed to have th eir responses included. Finally, participants were informed that this study w ould be ongoing for several days, and in order to maintain the integrity of the study they were to ld not to discuss the experiment itself with individuals. Subjects were then free to leave the classroom.
49 Measurement Pre-exposure Questionnaire The pre-exposure questionnaire was comprise d of two sections. Th e first recorded the respondents global mood state and was measured through a three-item semantic differential scale utilized by Pham (1996) with a Cronbach alpha of .94. The items included in this global mood scale were I am in a good mood/bad moo d, I am feeling happy/unhappy, and I am pleased annoyed. While several other scales conceptualized their us age as mood scales, the items clearly represented mo re visceral emotions. The second section measured th e individuals team identification (Fan Participant) and leveraged the original Wann and Branscombe (1993) SSIS. This scale consists of seven items, and uses a Likert-scale with end points from one to eight, with the summation of all seven items being used to create an aggregate score of fanship. Previous results suggested that the scale was both reliable, alpha of .91, and unidimensional. While other scales have examined the multidimensional nature of identific ation to include players, the team, coach, etc., this study is purposefully interested in fan identification with a team as the primary object of interest, thus the selection of the SSIS. Post-Exposure Questionnaire After reviewing the game footage and advert isement, respondents were asked to complete a second questionnaire comprised of five sections, program liking (FANCONTENT), advertising effectiveness (Aad a nd Ab), affective states (mood and emotions), involvement, and biographical data. In order to measure program liking in the context of the current study, a modification of existing scales wa s required to measure involvement with the specific sport The scale with the best fit for m odification appeared to be Speed and Thompsons (2000) scale measuring involvement with an event. Their origin al scale consisted of four items and reported a
50 Cronbach alpha of .96. A slight modification of th is scale produced the fo llowing three items to be measured on a seven-point Likert-type scale c onsisting of the following statements: I am a strong supporter of college bask etball, I would want to atte nd college basketball games, and I enjoy following coverage of college basketball. The second section consisted of Aad and Ab scales directed at the advertised brand. Several scales have been developed to measure Aa d, either by affect or cognition, or by media. For the purpose of this study, an attitude toward the ad scale will be borr owed from one created and utilized by Cho, Lee and Tharp (2001). This s cale consists of eight Likert-type statements designed to evaluate overall attitudes regarding a specific advertisement. The actual scale items were not provided in this article, but were detailed by Bruner, Hensel and James (2005). The eight items utilized in this scale are as follows: I like this ad, this ad is entertaining, this ad is useful, this ad is important, this ad is interesti ng, this ad is informative, I would enjoy seeing this ad again, and this ad is good (p. 735). The scale de monstrated high reliability for this study, with an alpha calculated at .87. This study will also utilize the Ab scale used by Cho, Lee and Tharp in their study of ba nner advertising effectiveness. Th is scale is composed of three Likert-type statements assessing overall attitude s regarding the exposed brand. The three items utilized in this scale, and provided by Br uner, Hensel and Jame s are: I like __________, __________ is satisfactory, and ___________ is desira ble. The authors report that the scale possesses an alpha of .92 which indicates strong reliability. While the vast majority of PIA research excludes emotions, this study will measure both moods and emotions due to the intense emotions known to be associated with sports, among highly identified fans. The scale usage for the mood variable can be found in the pre-exposure questionnaire section of this paper. To measur e emotions, the frequently used Mehrabian and Russell (1974) semantic differential scale will be us ed due to the strong validity and reliability of
51 the scale. This scale will allow for pleasure, arousal and dominance to be examined individually, utilizing a six-item, seven-point Likert sc ale to measure each dimension of emotions. Next, a three-item, seven-point Likert scale will be utilized to m easure the participants involvement with the stimuli to check if fan behavior was created through the experiment manipulation. For this measure the study utilized a situational involvement scale originating from the Zaichkowsky (1985) Personal Involvement Inventory (PII). This sc ale has been used in several studies and Cronbach al phas have been reported in th e range of .89 to .99 (Bruner, Hensel, and James 2005). The scale consists of three semantic differential pairs: unimportantimportant, of no concern-of concern to me, and ir relevant/relevant. The survey instrument also contained questions dealing with basic de mographic data including age and gender.
52 CHAPTER 4 RESULTS This chapter reports the statistical tests of the hypotheses and research questions. SP SS 17.0 and Microsoft Excel were used for the data an alysis. The first step wa s to determine if any respondents needed to be removed from the sample due to high levels of actual identification with either of the teams in the tested stimu li. Examining the scale used to measure team identification, this study obtained a Cronbach alpha of .83 and .84 for the usage of the sevenitem scale twice within this study, once for each of teams participating in the study. Of the 298 participants, only three indicated a level of identification over five on an eight-point scale measuring team identification with the two co mpeting teams. Given the small number of individuals indicating some level of fanship with one of the two teams, it was decided to leave these subjects in the analysis. As for the samp le itself, 70.4% of the subjects were female (n=209) and the mean age of the participants was 20.5 years. Means were compared on the variables studied to determine if there were any differences based on gender which may have affected the results of this study. Men and wome n reported no significant differences in pre or post moods, involvement with the stimulus, and re ported levels of pleasur e and arousal. The two groups did report significant differences on the scale used to measure fanship of college basketball, but this variable was seen to have no effect on any of the dependent variables tested. The next step required was a test to measur e the success of the manipulation of the two involvement treatments. Was the random assignment of individuals to a team and the promise of varying levels of extra credit e ffective in creating temporary involvement with the stimuli? In order to answer this question a manipulati on check was conducted utilizing a one-way ANOVA (analysis of variance) in which the three treatment conditions (w in, loss and control) were the independent variables and a three-item involve ment with the game and outcome scale was utilized as the dependent variab le. The three-item involvement s cale utilized reported a Cronbach
53 alpha of .96 in this experiment showing high reliability for this scale. Results showed involvement with the stimuli was significantly hi gher for the win and loss conditions than the control group where no team assignment or promise of additional rewards occurred. The means for the involvement levels among the test c onditions and control group were significantly different, F (2, 292) = 19.527, p < .05 (Table 4-1). Means for the win and loss condition were 4.8 and 4.49 respectively, while the mean for the control groups involvement was a mere 2.36. A post hoc Scheffe analysis confirms a signi ficant delineation be tween the two test conditions and the control group, thus demonstrati ng that during the experi ment the manipulation was successful in producing a significantly higher level of involvement among individuals in the treatment conditions. Thus, it is believed that this manipulation was successful in creating involvement in the stimuli representative of wh at fans experience while watching their teams compete. Therefore, it should be concluded that this manipulation yielded the correct conditions for a test of the previous ly identified hypotheses. Table 4-1. One-Way ANOVA results for involvement by condition Win Loss Control 4.8 (n=130) 4.5 (n=135) 2.4b (n=30) Note: means with different superscripts di ffer at p < .05 using Sc heffe post hoc test. H1 Level of fanship will moderate mood change s when exposed to team specific media regarding their preferred team. H1 theoretically tests the bipolarity of sports as a media vehicle. In order to test these hypotheses the dependent measure of post-mood wa s compared to the pre-mood measure using MANOVA (multivariate analysis of variance) with team performance (win, loss and control) as the independent variables. H1a The greater the fan involvement, the greater the negative moods following exposure to a losing team effort. H1b The greater the fan involvement, the greater the positive moods following exposure to a winning team effort.
54 Utilizing tests of between-s ubjects effects for the thr ee conditions demonstrated no significant differences on the measure of pre-experiment moods, F (2, 293) = 2.422, p > .05. The global mood scale reported a Cronbach alpha of 85 for the pre-exposure usage and .93 for the measure of moods post exposure, both of which i ndicate satisfactory reliability for this scale. Referencing Table 4-2 one can see that the two treatment co nditions were successful in creating significant mood changes among the two treatmen t conditions and control group F(2, 293) = 47.905, p < .01, but as expected the stimuli produ ced no significant mood changes among the control group when compared to their pre-m ood means. For respondents in the win condition, self-reported moods were signifi cantly higher after watc hing the stimuli, conversely, those in the loss condition reported significantly lower moods than they reported prior to beginning the experiment. Thus, both hypotheses were statistically supported. Table 4-2. MANOVA for mood by condition 95% Confidence Interval Dependent Variable Condition Mean Std. Error Lower Bound Upper Bound Win 5.300c .116 5.072 5.528 Loss 5.195c .114 4.971 5.419 premoodavg Control 5.774 .237 5.307 6.241 Win 5.569 c .128 5.317 5.821 Loss 3.901bc .126 3.654 4.148 postmoodavg Control 5.559 .262 5.044 6.075 Note: means with b superscripts differ between groups with a superscripts at p < .05 using Scheffe post hoc test. Means with c superscripts differ within groups from pre to post reporting. H2 Among highly identified fans of particip ating teams, sporting events will create significantly different emotions among fans of the winning and losing teams (or positive versus negative situations). H2a Pleasure will be significantly higher for highly identified fans of winning teams, versus highly identified fans of losing teams. Table 4-3 contains the means and standard deviations for the pleasure measure among the three conditions. A significan ce difference between the thr ee conditions was found using a
55 one-way ANOVA, F (2, 293) = 52.788, p < .05. Pleasure was measured using a 6-item scale consisting of the following semantic differential pairs: happy -unhappy, pleased-annoyed, satisfied-unsatisfied, contented-melancholic, hop eful-despairing, and re laxed-bored. All items were rated on a 7-point Likert scale with higher scores indicati ng positive pleasure responses and lower pleasure scores indicating negative pleasur e responses. Within this study the six-item scale reported a Cronbachs alpha of .9. Table 4-3. One-Way ANOVA results for pleasure by condition Win Loss Control 5.2 (n=130) 3.75b (n=135) 4.94a (n=31) Note: means with different superscripts di ffer at p < .05 using Scheffe post hoc test. A post hoc Scheffe analysis supported hypothesis H2a as respondents in the win condition reported significantly gr eater pleasure than responden ts in the loss condition. In addition, the control group reporte d significantly higher scores on th e pleasure dimensions than the loss condition. Interestingly, there was no signi ficant difference between individuals in the win condition and those in the control group. H2b Arousal will be significantly higher for hi ghly identified fans of winning and losing teams, compared to the control group of individuals with low levels of fanship. Table 4-4 contains the means and standard deviations for the arousal measure among the three conditions. A significan ce difference between the thr ee conditions was found using a one-way ANOVA, F (2, 293) = 6.945, p < .05. Arousal was measured using a 6-item scale consisting of the following semantic differentia l pairs: stimulated-relaxed, excited-calm, frenzied-sluggish, jittery-dull, wide awake-sleepy, and arousedunaroused. All items were rated on a seven-point Likert scale with higher scores indicating higher levels of arousal and lower pleasure scores indicating lower levels of arous al. The six-item scale for arousal reported a Cronbach alpha of .76.
56 Table 4-4. One-Way ANOVA results for arousal by condition Win Loss Control 4.0 (n=130) 3.5b (n=135) 3.8ab (n=31) Note: means with different superscripts di ffer at p < .05 using Sc heffe post hoc test. A post hoc Scheffe analysis indicated that there was no significant difference between the fan conditions (win and loss) and the contro l condition. There was however a significant difference between fans in the win and loss c onditions, with individuals in the win condition reporting significantly higher levels of arousal than individuals in the loss condition. These results require us to reject hypothesis 2b. H2c Dominance will be significantly higher for highly identified fans of winning and losing teams, compared to the control group of individuals with low levels of fanship. Table 4-5 contains the mean s and standard deviations for the dominance measure among the three conditions. A significance differe nce between the thr ee conditions was found using a one-way ANOVA, F (2, 293) = 7.082, p < .05. Dominance was measured using a 6-item scale consisting of the following semantic differe ntial pairs: controllin g-controlled, dominantsubmissive, influential-influenced, important-awed, autonomous-guided, and in control-cared for. All items were rated on a 7-point Likert scale with higher sc ores indicating higher levels of dominance and lower scores indicating lower leve ls of dominance, or increased levels of submissiveness. The six-item scale for dominance reported a Cronbach alpha of .75. Table 4-5. One-Way ANOVA results for dominance by condition Win Loss Control 4.4 (n=130) 4.0b (n=135) 4.4ab (n=31) Note: means with different superscripts di ffer at p < .05 using Sc heffe post hoc test. A post hoc Scheffe analysis identifies a sign ificant difference betw een the fan conditions, with subjects in the win condition indicating greater dominance than th eir peers in the loss condition who reported greater submissiveness (the opposite of dominance as detailed by
57 Mehrabian and Russell (1974)) There were no si gnificant differences between fans in the treatment conditions and individuals in the cont rol group. These results re quire us to reject hypothesis 2c. H3 Affective states created by the sport programmi ng will effect advertisement evaluations in this media, with Aad and Ab being directly co rrelated to positive and negative affective states created by the sport media, among highly identified fans. Table 4-6 contains the means and standard deviations for the Aad measure among the three conditions. Though a significance difference between the three conditions was found using a one-way ANOVA, F (2, 291) = 6.924, p < .05, a post hoc Scheffe analysis provides a more detailed look at differences between individual conditions which will be discussed further. The eight-item scale for Aad showed strong reliab ility with a reported Cronbach alpha of .93. Table 4-6. One-Way ANOVA results for Aad by condition Win Loss Control 4.4 (n=130) 4.2a (n=133) 5.1b (n=31) Note: means with different superscripts di ffer at p < .05 using Sc heffe post hoc test. H3a Highly identified fans of winning teams will w itness elevated levels of Aad as compared to highly identified fans of the losing team and non-fans. A post hoc Scheffe analysis fails to suppor t hypothesis 3a, as subjects in the win condition showed no significance difference in rati ng the tested advertisement on the Aad scale compared to the subjects in the loss condition. In addition, the analysis indicates that while there is a significant difference between the win and control conditions, th e difference is not in the direction expected. Respondents in the c ontrol condition reported a significantly higher Aad, p<.05, than the respondents in the win c ondition. Thus, we must reject hypothesis 3a. H3b Highly identified fans of losing teams w ill report lower Aad scores than highly identified fans of the winning team and non-fans. The statistical results of a post hoc Scheffe analysis failed to support hypothesis 3b, as subjects in the loss condition showed no si gnificance difference in rating the tested
58 advertisement on the Aad scale compared to the subjects in the win condition. Similarly to the win condition, the analysis indicates that ther e is a significant differe nce between the loss and control conditions, p< .05, with the difference being in the hypothesized direction. Given the lack of difference between the win and the loss conditions however, one can assume that the differences in Aad results may be more highly related to a factor othe r than the independent variables of wins or losses. H3c Highly identified fans of winning teams will witness elevated levels of Ab as compared to highly identified fans of the losing team and non-fans. Having determined that there were no si gnificant differences between the two test conditions on the Aad measure, it would foretell th at there is likely no significant affect on Ab measures. Table 4-7 contains the means and standard de viations for the Ab measure among the three conditions. The three-item Ab scale showed acceptable reliability with a Cronbach alpha of .91. Similar to the Aad test, a one-way ANOVA fo r Ab scores across identified no significant differences between the test conditions. In addition, there were no si gnificant differences between either of the two test conditions and the control group, F (2, 292) = 2.817, p > .05. Again, a post hoc Scheffe analysis was utilized to provide a more detailed look at differences between individual conditions. This analysis indicates that the greatest difference on the Ab measure was between the loss a nd control conditions, but these scores were not significant, p=.079. Table 4-7. One-Way ANOVA results for Ab by condition Win Loss Control 4.5 (n=130) 4.2 (n=134) 4.8 (n=31) The post hoc Scheffe analysis confirms th at there were no significant differences found on the Ab measure for the win, lo ss and control conditions. Therefor e, we can reject hypothesis 3c as no significant differences were found.
59 H3d Highly identified fans of losing teams will re port lower Ab scores than highly identified fans of the winning team and non-fans. Referring back to Table 4-7 we see that there were no sign ificant differences between the loss condition and the win and cont rol conditions on the Ab measure for the tested ad. Again, we must reject hypothesis H3d as no significant differences were found for individuals in the loss condition.
60 Table 4-8. Summary of Hypothesis Findings Supported Failed to Support H1a The greater the fan involvement, the greater the negative moods following exposure to a losing team effort. X H1b The greater the fan involvement, the greater the positive moods following exposure to a winning team effort. X H2a Pleasure will be significantly hi gher for highly identified fans of winning teams, versus highly identified fans of losing teams. X H2b Arousal will be signifi cantly higher for highly identified fans of winning and losing teams, compared to the control group of individuals with low levels of fanship. X H2c Dominance will be significantly higher for highly identified fans of winning and losing teams, compared to the control group of individuals with low levels of fanship. X H3a Highly identified fans of winning teams will witness elevated levels of Aad as compared to highly identified fans of the losing team and non-fans. X H3b Highly identified fans of losing teams will report lower Aad scores than highly identified fans of the winning team and non-fans. X H3c Highly identified fans of winning teams will witness elevated levels of Ab as compared to highl y identified fans of the losing team and non-fans. X H3d Highly identified fans of losing teams will report lower Ab scores than highly identified fans of the winning team and non-fans. X
61 CHAPTER 5 DISCUSSION AND CONCLUSIONS The purpose of the current research was to be tter understand the affective states created by sports, and their im pact on advertising effectiv eness as measured by attitudes toward ads and brands. This study differentiated itself from previous research by Pavelc hak et al. (1988) and Newell et al. (2001) in th at it expanded advertising effectiven ess from brand recall to attitudes toward specific advertising a nd attitudes toward brands with in sport programming. The extant literature in Aad research has supported the pr emise that moods created by TV programming have a significant effect on adve rtising effectiveness, yet resear ch has failed to address the possible creation of bipolar moods created by sp ort programming, and what effect these bipolar moods may have on advertising ev aluations within a single progr am among different audiences. This issue has significant relevance to marketer s and advertisers if th e creation of negative moods and emotions, in the case of negative event outcomes (i.e. loss es), has a detrimental effect on advertising and sponsorship evaluations. While the unpredictability of sports would prevent an advertiser from avoiding exposure to fans in the latter condition, it wo uld provide caution to the brands, which could manifest itself in lo wer CPMs to reach these audiences due to the volatile nature of moods among highly involved fans. Affective Responses Created by Sports Programming This study corroborates the findings of previous research studies that sports are capable of creating changes in af fective states among high ly involved sports fans. This was demonstrated through the significantly different moods reported among highly i nvolved fans in the win and loss conditions. These globa l mood states moved significantly in the expected directions for fans of the winning and losing teams. In a ddition to significant changes in self reported moods, participants who were in the win c ondition reported significantly higher levels of pleasure when compared to the loss conditi on. Additionally, indi viduals in the win
62 condition reported higher levels of arousal and dominance than their peers in the loss condition. These results vary from those of Newell et al. (2001), which f ound that arousal levels were significantly higher for fans of both the winning and losing teams when compared to nonfans. Pavelchak et al. (1988) found that arousal was a greater predic tor of ad recall than pleasure, but the former study failed to examine the relati onship of these dimensions of emotions to attitudes towards adve rtising and brands. While the link between sport events (wins /losses and positive/negative outcomes) and mood and emotion changes is quite sensible, the bi polarity of these affective responses are still poorly understood and under researched by scholar s. This study was unable to demonstrate a significant relationship between these affective stat es and attitude formation toward advertising; extant research would suggest that variance in evaluations is like ly given the breadth of research showing a correlation between moods and advertising evaluations. Attitude findings While this study identified elevated Aad scores among the control group, representing non-fans, relevant to the test c onditions, the study was unable to conclusively attribute this difference to any of the factors m easured, includi ng; the three dimensions of emotions and mood states. While involvement with the program was significantly lower among the control group, this difference fails to explain the heightened attitudes for th e advertisement given that the involvement failed to correlate with the arousal reported between the test conditions and the control group. Given that no significant differences were found between the two test conditions on the Aad measure, it comes as no surprise that no si gnificant differences were found between these two groups for the Ab measure. The lack of si gnificance between the two test conditions and the
63 control group is also explaina ble due to the moderate, but si gnificant difference between the control group and test conditi ons on the Aad measure. Limitations Due to the pioneering nature of this study, the results presented in this paper have significant lim itations as the protoc ol for this research was desi gned specifically for this study. The first and most notable limitation of this res earch is the creation of temporary involvement through the manipulation of extr a credit. While the author believes this approach was a successful method of creating invol vement with the specific team s in the stimuli being viewed, the audience clearly lacked the long-term orientation and knowle dge of the participants that would be found by conducting a simila r study with actual fans during a live game. A critic could also argue that the affective res ponses created were the result of the incentive and not the actual sport consumption, but for the purpose of this st udy I believe the two were one in the same since the incentive was directly correla ted to the performance of the a ssigned teams. Ideally this study would be conducted utilizing real fans watching a game in real time, but this was unrealistic for this research given the financial limitations of recruiting fans of two co mpeting teams to watch a game live. A benefit of the current research design was the homogeneity of the sample. This same homogeneity would be much more difficult to accomplish using fans of actual teams. For instance, in Chicago, Cubs fans are known for be ing more white-collar and affluent than their cross town White Sox brethren. Cubs fans and White Sox fan are also percei ved as to being fans for different reasons, with Cubs fans being more motivated by the social affiliation and entertainment associated with Wrigley Field and the surrounding neighborhood. So, while real fans may improve the external validity of th e research, there would undoubtedly be additional confounds that the current study was able to elim inate. Again, for theory testing research
64 homogeneity is of paramount importance, t hus this sacrifice wa s indeed prudent, and advantageous to this study. Unfortunately, live broadcasts also force th e researcher to relinquish control of the advertising featured in the stimuli. So, if the advertisements are all familiar ads for top tier brands like Apple, Coca-Cola and Bud Light, it may not provide the best test of attitude formation if attitudes have already been info rmed and are fairly well entrenched in the consumers. The second limitation of the research was the lack of variety in s port stimuli tested. As the literature review di scussed, there are numerous factors th at affect emotions of viewers, beyond mere wins and losses. The ebb and flow of many sporting events also creates emotional swings which this study was unable to replicate and measure. This study relied on a single game that was characterized by a closely fought game until the last couple of minutes in which one team pulled ahead. A last second shot, a monume ntal comeback, an uplifting David vs. Goliath underdog story, or a controversial call by refere es could be additional factors which would influence affective responses besides the final score. An additional limitation of this research was the size of the control group. A larger sample would have made statistics utilizing th is group more stable. The timing of the control sample so close to a holiday resulted in lower than expected turn out which negatively affected attendance for this sample. The decision to use a separate class for the control group could also be criticized for lacking true random assignment, but for the integrity of the study the researcher chose to accept this risk in order to ensure the manipulation of the extra credit points and interest in the program was successfully manipulated. If participants felt that situation was unrealistic and their extra credit would not vary based on the performance of the two participating teams, then would have been less involve d and less likely to experience cha nges in their affective states.
65 While the selection of a second tier brand in the cell phone category was meant to foster a cleaner slate for attitude forma tion, the loyalty held for other br ands within this category may have affected attitudes. For instance, iPhone advocates may have been predisposed not to like a non-Apple product due to their preference for this particular manufacturer Unfortunately, this bias was not accounted for and thus can not be disc redited as a possible cova riate of the attitudes reported in this study. Lastly, since emotions are attributable to a specific source, the author determined it was improper to compare emotions from a pre and post exposure scenario, so we are limited to the variations between the three c onditions at the post-e xposure point only. Thus we are unable to determine how pleasure, arousal and domina nce changed within subjects throughout the experiment. Since emotions were self reported at the conclusion of the stimulus exposure, these results are directly attributable to the emoti ons created by the stimuli. This same set of questioning seemed inappropriate prior to the exposure since the respondents would lack a reference point for evaluation. The primary diffe rentiator between moods and emotions was the attribution of emotions to a sp ecific factor, a factor which wa s absent at the onset of the experiment. The redundancy of the questioning woul d also have lead to fatigue due to the breadth of the scale for these measures. Future Research Due to the m assive investments being made by corporations in this genre of marketing, future research is certainly needed in this ar ena to better understand the bipolarity of affective responses created by sports programming. Specifi cally, how these divergent affective responses influence recall and attitude formation. While this research found no relationship betw een negative affective states and negative attitude formations among fans of losing teams, more conclusive research needs to be conducted
66 in the field to better understand why this is not occurring, when mood congruency theory suggests that we should see an impact on attitude formation. If research does indeed discover that negative emotions are detrimental to marketers in these programs then additional research could be conducted to identify how different sports vary in the creation of affective responses. For instance, are college sports fans more feverish th an professional team s ports fans? Are European soccer fans more feverish than U.S. soccer fans ? How do affective responses differ for individual sports vs. team sports? Perhaps tennis and golf vi ewers are more connected with the sport than specific athletes, thus less likel y to experience the emotional ebbs and flows created by viewing team sports where there is a higher probability of witnessing a win or a loss due to the smaller number of participants. So while a NASCAR fan may favor a specific driver, say Jeff Gordon, they may realize that he is competing against 40 or so drivers every week, thus the probability of him winning is much lower than a basketba ll team thats playing a single opponent. This study also focused on the evaluation of an advertisement plac ed immediately after the conclusion of an event. Additional studies in memory formation would be of value to understand how individuals code advertisements in their memory, specifically in regards to emotions felt at the time of exposure compared to emotions felt at the conclusion of an event. So, when a fan recalls a previously seen advertis ement, do they code th eir attitudes with the emotions felt at the specific moment they saw th e ad, or with the emotions tied to the overall sport viewing experience based on the final outcome of the contest? Future research may also include the testi ng of commercials of various moods within the context of sport programming. For instance, do fans feeling negativ e emotions respond better to happy or sad commercials? Some may argue that this is irrelevant since marketers are unable to predict the outcomes of sporting events, and thus would be unable to targ et different audiences with different creative, but this knowledge would be of significant value in supporting a
67 consistency effects or mood maintenance theory of the role of creative mood tones and PIA in attitude formation. Conclusions The primary objective of this research was to better understand affective states created by sports programming to determine whether or not attitudes toward advertisements and brands differed based on the moods and emotions felt by fans of competing teams. While the passion sports fans have for the programming is unquest ioned, and has largely been believed to be a positive affiliation for advertisers, this research was unable to identify any detrimental effects linked to the negative emotions felt by fans of losing teams. Conversely, no positive rub-off was witnessed among fans of the winning team. Whil e the affective respons es created by sports programming may still have signif icant relevance to advertisers, this research was unable to demonstrate a clear link between mood changes and attitudes toward advertising. Admittedly, not enough is known regarding attitude formation for advertising relevant to the moods and emotions felt at various points in time. These findings fail to support the mood congr uency theory that attitudes toward advertisements will move in a congruent directio n with the mood of the media context. This research was also unable to explore the cons istency effects and m ood maintenance repair theories of PIA research.
68 APPENDX A INFORMED CONSENT DISC LOS URES AND QUESTIONNAIRE Purpose of the Study: The purpose of this study is to gain insights into sport viewing among males and females over 18. Expectation of Study Participants: Those choosing to participate in this study you will be asked to vi ew a brief 10 minute video of a previously recorded college basketball game. Prior and post viewing, you will be asked to answer a total of 58 questions regarding your ex perience. The study should take approximately 20 25 minutes to complete. Potential Risks: There are no potential health or stress risk s involved with this study, nor will there be any personal discomfort. Should any participant feel uncomfortable at a ny time during this study, they are free to discontinue their participation with no penalties or questions asked. Compensation and Benefits: Participants will receive no financial compensa tion for their involvement in this study. An average of two extra credit points will be awarded to individuals from participating sections of Elements of Advertising who co mplete the study. There are no ot her expected benefits from participation in this study. Confidentiality: All information collected will remain confidenti al and responses by participants will not be associated with personal names. Participants ar e free to withdraw at any time during the study for any reason. Questions/Contact Information: If you have any questions or comments regardin g the study feel free to contact Michael Clayton or Dr. Sutherland in the Adver tising Department. Both individua ls have offices located in Weimer Hall, or you may contact them by email at email@example.com and firstname.lastname@example.org For questions about your rights as a research particip ant, contact the IRB @ 352-392-0433. Agreement: I have read and understand all of the above info rm ation and agree to participate in the study. I understand that my participation is completely voluntary and I have received a copy of this information. Participant (print name): ______________________________ Sect (a.m. or p.m.) _________ Participant (sign):____________________________________ Date:____________________ Primary Researcher:__________________________________ Date:____________________
69 Color of card (red or black): ______________ Please read each of the ten pairs of words be low carefully and indicate answer by circling the number which best corresponds with your current state. 1. I am in a good mood 1 2 3 4 5 6 7 I am in a bad mood 2. I am annoyed 1 2 3 4 5 6 7 I am feeling pleased 3. I am feeling happy 1 2 3 4 5 6 7 I am feeling unhappy 4. How important to YOU is it that the Memphis basketball team wins? Not Important 1 2 3 4 5 6 7 8 Very Important 5. How important to YOU is it that the University of Massachusetts (UMASS) basketball team wins? Not Important 1 2 3 4 5 6 7 8 Very Important 6. How strongly do YOU see YOURSELF as a fa n of the Memphis basketball team? Not at all a fan 1 2 3 4 5 6 7 8 Very much a fan 7. How strongly do YOU see YOURSELF as a fan of the UMASS basketball team? Not at all a fan 1 2 3 4 5 6 7 8 Very much a fan 8. How strongly do your FRIENDS see YOU as a fan of the Memphis basketball team? Not at all a fan 1 2 3 4 5 6 7 8 Very much a fan 9. How strongly do your FRIENDS see YOU as a fan of the UMASS basketball team? Not at all a fan 1 2 3 4 5 6 7 8 Very much a fan 10. During the season, how closely do you follow the Memphis basketball team via ANY of the following: a) in person or on television, b) on the radio, c) television news or a newspaper, d) online? Never 1 2 3 4 5 6 7 8 Almost every day
70 11. During the season, how closely do you follow the UMASS basketball team via ANY of the following: a) in person or on television, b) on the radio, c) television news or a newspaper, d) online? Never 1 2 3 4 5 6 7 8 Almost every day 12. How important is being a fan of Memphis basketball to YOU? Not important 1 2 3 4 5 6 7 8 Very important 13. How important is being a fan of UMASS basketball to YOU? Not important 1 2 3 4 5 6 7 8 Very important 14. How much do YOU dislike Memphis basketballs greatest rivals? Do not dislike 1 2 3 4 5 6 7 8 Dislike very much 15. How much do YOU dislike UMASS basketballs greatest rivals? Do not dislike 1 2 3 4 5 6 7 8 Dislike very much 16. How often do YOU display the Memphis bask etball teams name or insignia on your automobile, where you live, or on your clothing? Never 1 2 3 4 5 6 7 8 Always 17. How often do YOU display the UMASS basketball teams name or insignia on your automobile, where you live, or on your clothing? Never 1 2 3 4 5 6 7 8 Always Thanks. You will now be shown the beginning of a previously recorded basketball game between Memphis and the University of Massachusetts (UMASS). When you entered the classroom you were randomly given a playing card. If you received a BLACK playing card you have randomly been assigned Memphis. If you received a RED playing card you have randomly been assigned the University of Massachusetts (UMASS). In order to make this research more interesting your professor has agreed to let me adjust his standard 2 point extra credit format. Instead of rewarding ever yone 2 extra credit points for participating in this study, you will earn either 1 or 3 extra credit points based on which team scores more points during this portion of the game you ar e about to watch. So if the University of Massachusetts (UMASS) outscores Memphis, red card holders will earn 3 extra credit points while black card holders will earn on e extra credit point. Conversely, if Memphis outscores the University of Massachusetts (U MASS), black card holders will earn 3 point while red card holders will earn one point. At the conclusion of the study today you will need to turn your survey in with your playi ng card to receive extra credit points.
71 If this uncertain outcome makes you uncomfortable, or you are uncomfortable with the nature of this experiment you are free to exit the experiment now or at any point and we will be happy to reschedule you into another ex periment worth 2 extra credit points for all participants. If you agree to this modification of th e original disclosure please sign below: Participant (sign):____________________________________ Date:_______________ Your name: ___________________________________________ Please answer the following questions regar ding your attitudes towards the Samsung BEHOLD advertisement you just viewed. In dicate your agreement with each of the following statements by circling a number between 1 and 7 which best represents your feelings on the following 7-point scale: Strongly disagree 1 2 3 4 5 6 7 Strongly agree 18. I like this ad 1 2 3 4 5 6 7 19. The ad is entertaining 1 2 3 4 5 6 7 20. The ad is useful 1 2 3 4 5 6 7 21. The ad is important 1 2 3 4 5 6 7
72 22. The ad is interesting 1 2 3 4 5 6 7 23. The ad is informative 1 2 3 4 5 6 7 24. I would enjoy seeing this ad again 1 2 3 4 5 6 7 25. The ad is good 1 2 3 4 5 6 7 26. I like Samsung 1 2 3 4 5 6 7 27. Samsung is satisfactory 1 2 3 4 5 6 7 28. Samsung is desirable 1 2 3 4 5 6 7 Next, you will be presented with 18 pairs of bipolar adjectives. Please circle a number between 1 and 7 based on YOUR current feelings in relation to each pair of words. 29. Relaxed 1 2 3 4 5 6 7 Bored 30. In control 1 2 3 4 5 6 7 Cared for 31. Aroused 1 2 3 4 5 6 7 Unaroused 32. Hopeful 1 2 3 4 5 6 7 Despairing 33. Autonomous 1 2 3 4 5 6 7 Guided 34. Wide awake 1 2 3 4 5 6 7 Sleepy 35. Contented 1 2 3 4 5 6 7 Melancholic 36. Important 1 2 3 4 5 6 7 Awed 37. Jittery 1 2 3 4 5 6 7 Dull
73 38. Satisfied 1 2 3 4 5 6 7 Unsatisfied 39. Influential 1 2 3 4 5 6 7 Influenced 40. Frenzied 1 2 3 4 5 6 7 Sluggish 41. Pleased 1 2 3 4 5 6 7 Annoyed 42. Dominant 1 2 3 4 5 6 7 Submissive 43. Excited 1 2 3 4 5 6 7 Calm 44. Happy 1 2 3 4 5 6 7 Unhappy 45. Controlling 1 2 3 4 5 6 7 Controlled 46. Stimulated 1 2 3 4 5 6 7 Relaxed Please answer the following questions regardi ng your interest in college basketball. 47. I am a strong supporter of college basketball. Strongly disagree 1 2 3 4 5 6 7 Strongly agree 48. I would want to attend college basketball games. Strongly disagree 1 2 3 4 5 6 7 Strongly agree 49. I enjoy following coverage of college basketball. Strongly disagree 1 2 3 4 5 6 7 Strongly agree Please answer the following questions regarding your interest in the particular game and outcome you just observed. 50. Unimportant 1 2 3 4 5 6 7 Important 51. Of no concern 1 2 3 4 5 6 7 Of concern to me 52. Irrelevant 1 2 3 4 5 6 7 Relevant The remaining questions are intended for classification purposes only. 53. What is your age? _______
74 54. What is your gender? Male Female 55. Do you recall having previously viewed the game you just saw video from? YES NO 56. I am in a good mood 1 2 3 4 5 6 7 I am in a bad mood 57. I am annoyed 1 2 3 4 5 6 7 I am feeling pleased 58. I am feeling happy 1 2 3 4 5 6 7 I am feeling unhappy Thank you for your participation in this experiment The purpose of this experiment was to study the effects of affective states created through spor ts based on affiliation with a team, leading to a positive or negative sports viewing experience. In order to accomplish this it was necessary to create a rooting interest in each of the teams partic ipating. In order to be fa ir to all participants however, everyone will receive 3 extra credit poi nts for your class regardless of which team you were randomly assigned. Now that you have learned the true and full pur pose of the current study and know about the actual manipulations that took place as part of this study, do you wish to have your data included in this research project? ____ Yes, _____________________________ (please sign here) ____ No, _____________________________ (please sign here) Please remember that this study will be going on for several days. Please help us maintain the integrity of the study by not discussing the experi ment with your peers. We thank you again for your participation and you are free to leave.
75 APPENDIX B ACRONYM SUMMARY Aad: Attitud e toward the ad Ab: Attitude toward the brand ANOVA: Analysis of variance BIRGing: Basking-In-Reflected Glory CORFing: Cutting-Off-Reflected Failure) CPM: Cost per thousand DVRs: Digital video recorders ELM: Elaboration Likelihood Model IP: Involvement Profile IRB: Institutional Review Board MSSC: Motivation Scale for Sport Consumption PAD: Pleasure, Arousal and Dominance PAI: Points of Attachment Index PCM: Psychological Continuum Model PCT: Psychological Commitment to Team PIA: Program induced affect SAM: Self-Assessment Manikin SEVA: Surgency, Elation, Vigor/Activation SFQ: Sport fandom questionnaire SII: Sport Interest Inventory SSIS: Sport Spectator Identification Scale
76 REFERENCES Allen, Natalie J., and John P. Meyer (1990), The Measurement and Antecedents of Affective, Continuance and Norm ative Commitment to the Organization, Journal of Occupational Psychology, 63(1), 1-18. Arora, Raj (1985), Consumer Involvement: What it offers to advertising strategies, International Journal of Advertising 4(2):119-130. Atkinson, Claire (2007, May 10), I s Crucial Daypart Finally Past Its Prime?: Megabillion TV Jackpot is Evolving into Big Events (Sports, Idol) Vs. Time-Shifted on Demand (Everything Else) AdAge.com, http://adage.com/p rint?article_id=116578 1-4. Ayelsworth, Andrew B., and Scott B MacKenzie (1998), Context Is Key: The Effect of Program-Induced Mood on Thoughts About the Ad, Journal of Advertising, 27 (2), 17. Babin, Barry J., William R. Darden, and Laurie A. Babin (1998), Negative Emotions in Marketing Research: Affect or Artifact? Journal of Business Research, 42 (3), 271. Bagozzi, Richard P., Mahesh Gopinath, and Prashanth U. Nyer (1999), The Role of Emotions in Advertising, Journal of the Academy of Marketing Science 27(2), 184. Batra, Rajeev, and Michael L. Ray (1986), A ffective Responses Mediating Acceptance of Advertising, Journal of Consumer Research, 13 (2), 234. Bruner II, Gordon C. (1998), Standardization & Ju stification: Do Aad Scales Measure Up? Journal of Current Issues & Research in Advertising, 20(1), 1-18. ________________, Paul J. Hensel, and Karen E. James (2005), Marketing Scales Handbook Volume IV: A Compilation of Multi-Item Measur es for Consumer Behavior & Advertising, Mason, Ohio: Thomson Corporation. Burnett, John, Anil Menon, and Denise T. Smar t (1993), Sports Marketing: A New Ball Game with New Rules, Journal of Advertising Research 33(5), 21-35. Calder Bobby J., Lynn W. Phillips and Alice M. Tybout (1981), Designing Research For Application, Journal of Consumer Research 8(2), 197-207. Chang, Chingching (2006), Beating the News Blues: Mood Repair Through Exposure to Advertising, Journal of Communication 56, 198-217. Chebat, Jean-Charles, Pierre Filiatrault, Claire Gelinas-Chebat, Alexander Vanisky (1995), Impact of Waiting Attri bution and Consumers Mood on Perceived Quality, Journal of Business Research 34(3), 191-196. Cho, Chang-Hoan, Jung-Gyo Lee, and Marye Thar p (2001), Different Forced Exposure Levels to Banner Advertisements, Journal of Advertising Research 41(4), 45-56.
77 Cialdini, Robert B., Richard J. Borden, Avril Thorne, Marcus R. Walk er, Stephen Freeman, and Lloyd R. Sloan (1976), Basking in Reflected Glory: Three (Football) Field Studies, Journal of Personality and Social Psychology, 34(3), 366-375. Copeland, Robert, Wendy Frisby, and Ronald E. McCarville (1996), Understanding the sport sponsorship process from a corporate perspective, Journal of Sport Management, 10(1), 32-48. Coulter, Keith S. (1998), The Effects of Aff ective Responses to Media Context on Advertising Evaluations, Journal of Advertising, 27 (4), 41. Dahlen, Micael (2005), The Medium as a Cont extual Cue: Effects of Creative Media Choice, Journal of Advertising, 34(3), 89-98. De Pelsmacker, Patrick, Maggie Geuens, and Pa scal Anckaert (2002), Media Context and Advertising Effectiveness: Th e Role of Context Appreciatio n and Context/Ad Similarity, Journal of Advertising, 31 (2), 49. Dietz-Uhler, Beth, and Audrey Murrell (1999), Examining Fan Reactions to Game outcomes: A Longitudinal Study of Social Identity, Journal of Sport Behavior, 22(1), 15-27. DN Online (2009, March 24), Study: More Studen ts Using Smartphones. Retrieved on April 3, 2009, from http://media.www.bsudailynews.com/media/storage/paper849/news/2009/03/24/ Dnupdate/Study.More.Students.Using.Sm artphones-3680661.shtml Eroglu, Sevgin A., Karen A. Machleit, Lenita M. Davis (2003), Empirical Testing of a Model of Online Store Atmospherics and Shopper Responses, Psychology and Marketing 20(2), 139150. Fact Pack (2005, February 28). Supplement to Advertising Age, 1-31, New York: Crain Communications Inc. Faseur, Tim, and Maggie Geuens (2006), Different Positive Feelings Leading to Different Ad Evaluations, Journal of Advertising, 35(4), 129-142. Funk, Daniel C., and Jeffrey D. James (2001), The Psychological Continuum Model: A Conceptual Framework for Understanding an Individuals Psychological Connection to Sport, Sport Management Review 4(2), 119-150. _____________, Daniel F. Mahony, Makoto Nakazaw a and Sumiko Hirakawa (2001), Development of the Sport Inte rest Inventory (SII): Impli cations for Measuring Unique Consumer Motives at Team Sporting Events, International Journal of sports Marketing & Sponsorship, 3(3), 291-316. _____________, _______________, and Lynn L. Ridinger (2002) Characterizing Consumer Motivation as Individual Difference Factors: Au gmenting the Sport Interest Inventory (SII) to Explain Level of Spectator Support, Sport Marketing Quarterly, 11(1), 33-43.
78 ____________, Lynn L. Ridinger, and Anita M. Moorman (2004), Exploring Origins of Involvement: Understanding the Relationship Betw een Consumer Motives and Involvement with Professional Sport Teams, Leisure Sciences 26, 35-61. Gantz, Walter, and Lawrence A. Wenner (1995), Fanship and the Television Sports Viewing Experience, Sociology of Sport Journal, 12, 56-74. Gardner, Meryl P. (1985), Mood States and Consumer Behavior: A Critical Review, Journal of Consumer Research 12 (December), 281-300. ______________, and Ronald P. Hill (1988), Consumers Mood States: Antecedents and Consequences of Experiential versus Informational Strategies for Brand Choice Psychology & Marketing, 5(2) 169-182. Gemmill, Erin, and Michael Peterson (2006) Technology Use among College Students: Implications for Student Affair Professionals, NASPA Journal 43(2), 280-300. Goldberg, Marvin E., and Gerald J. Gorn ( 1987), Happy and Sad Programs: How They Affect Reactions to Commercials, Journal of Consumer Research, 14 (3), 387. Goldstein, Jeffrey H, and Robert L. Arms (1971), Effects of Ob serving Athletic Contests on Hostility, Sociometry, 34 (1), 83-90 Goodman, Gail S. (1980), Picture Memory: How the Action Schema Affects Retention, Cognitive Psychology 12(4), 473-495. Gorn, Gerald, Michel T. Pham, and Leo Y. Sin (2001), When Arousal Influences Ad Evaluation and Valence Does Not (and Vice Versa), Journal of Consumer Psychology, 11 (1), 43. Guttmann, Allen (1986), Sports Spectator s. New York: Columbia University. Gwinner, Kevin, and Scott R. Swanson (2003), A model of fan identification: Antecedents and sponsorship outcomes, Journal of Service Marketing, 17 (3), 275-294. Havitz, Mark E., and Frdric Dimanche ( 1990), Propositions for Guiding the Empirical Testing of the Involvement Construct in the Recreational and Tourist Constructs, Leisure Sciences, 12(2), 179-196. Havitz, Mark E., and Frdric Dimanche (1997), Leisure Involvement Revisited: Conceptual Conundrums and Measurement Advances, Journal of Leisure Research 29(3), 245-278. Havlena, William J, and Morris B. Holbrook (19 86), The Varieties of Consumption Experience: Comparing Two Typologies of Emot ion in Consumer Behavior, Journal of Consumer Research, 13(3), 394-404. Heckler, Susan E, and Terry L. Childers (1992), The Role of Expectancy and Relevancy in Memory for Verbal and Visual Info rmation: What is Incongruency? Journal of Consumer Research, 18(4), 475-492.
79 Heere, Bob, and Jeffrey D. James (2007), Ste pping Outside the Lines: Developing a Multidimensional Team Identity Scale Based on Social Identity Theory, Sport Management Review, 10, 65-91. Hiestand, Michael (2004, October 28), A ds muscling into live TV events, USA Today Sports, Pg. 11c. Hill, Brad (2005), Processing Pull-through Advertisements Screen ed During Sporting Telecasts: Effects of Advertisement Message Speed, Pr ogramme Context, and Repetition Priming, Unpublished Doctoral Dissertation, Griffith University. Hirt, Edward R., Dolf Zillman, Grant A. Erickson, and Chris Kennedy (1992), Costs and Benefits of Allegiance: Changes in Fans' Se lf-Ascribed Competencies After Team Victory Versus Defeat, Journal of Personality and Social Psychology, 63(5), 724-738. Holbrook, Morris B., and John OShaughnessy (1984), The Role of Emotion in Advertising, Psychology and Marketing, 1(2), 45-64. IBM (2009, March 12) IBM and Ma rist Survey Shows U.S. College Students Want Technology Skills to Compete for Jobs. Retrieved Apr il 3, 2009, from http://www.marketwire.com/pressrelease/Ibm-NYSE-IBM-960420.html. Kamins, Michael A., Lawrence J. Marks, and De borah Skinner (1991), T elevision Commercial Evaluation in the Context of Program Induced Mood: Congruency Versus Consistency Effects, Journal of Advertising, 20(2), 1. Kerstetter, Deborah L., and Geor gia M. Kovich (1997), An invol vement Profile of Division I Womens Basketball Spectators, Journal of Sport Management 11, 234-249. Kwon, Harry H., Galen T. Trail, and Dean S. Anderson (2005), Are Multiple Points of Attachment Necessary to Predict Cognitive, A ffective, Conative or Behavioral Loyalty? Sport Management Review, 8(3), 255-270. Lang, Peter J. (1980), Behavioral Treatment and Bio-Behavioral A ssessment: Computer Applications in Technology. In Mental Health Care Delivery Systems, Joseph B. Sidowski, James H. Johnson, and Thomas A. Williams, eds. Norwood, NJ: Ablex. Laurent, Gilles, and Jean-Nol Kapferer (1985 ), Measuring Consumer Involvement Profiles, Journal of Marketing Research, 22(1) 41-53. Lutz, Richard J., Scott B. MacKenzie, and George E. Belch (1983), "Attitude Toward the Ad as a Mediator of Advertising E ffectiveness: Determinants and Consequences, in Advances in Consumer Research, 10, 532-539. Lyberger, Mark R., and Larry McCarthy (2001), An analysis of volume consumption, consumer interest and perceptions of sport sponsors hip as they relate to the Super Bowl, International Journal of Sports Marketing and Sponsorship, 3(4), 429-448.
80 MacKenzie, Scott B., and Richard J. Lutz (1989) An Empirical Examination of the Structural Antecedents of Attitude toward the Ad in an Advertising Pretesting Context, Journal of Marketing, 53(2), 48-65. Mackie, Diane M. and Leila T. Worth (1989), Pro cessing Deficits and the Mediation of Positive Affect in Persuasion, Journal of Personality and Social Psychology, 57(1), 27-40. Madrigal, Robert (2003), Investigating an Evolving Leisure Experience: Antecedents and Consequences of Spectator Affect During a Live Sporting Event, Journal of Leisure Research, 35(1), 23-48. Mahony, Daniel F, Robert Madrigal, and De nnis Howard (2000), Using the Psychological Commitment to Team (PCT) Scale to Segm ent Sport Consumers Based on Loyalty, Sport Marketing Quarterly 9(1), 15-25. Martin, Brett A.S. (2003), The Influence of Gender on Mood Effects in Advertising Psychology in Marketing 20(3), 249-273. McKinley Jr., James C. (2000, August 11), Sports Psychology; It Isn't Just a Game: Clues to Avid Rooting. New York Times, Vol. 149, Issue 51477, pA1. Meenaghan, Tony (2001), Understanding Sponsorship Effects, Psychology & Marketing, 18(2), 95-122. Mehrabian, Albert, and James A. Russell (1974), An Approach to Environmental Psychology, Cambridge: MIT Press. Milne,George R., and Mark A. McDonald (1999), Sport Marketing: Managing the Exchange Process. Sudbury, MA: Jones and Bartlett Publishers. Mintel International Group Limited. ( 2007, July) Sports Enthusiasts, U.S. Mitchell, Andrew A., and Jerry C. Olson ( 1981), Are Product Attri bute Beliefs the only Mediator of Advertising E ffects on Brand Attitude, Journal of Marketing Research, 18(3), 31332. Mitchell, Monique M., Kenneth M. Brown, Meli nda Morris-Villagran, and Paul D. Villagran (2001), The Effects of Anger, Sadness and Ha ppiness on Persuasive Message Processing: A Test of the Negative State Relief Model, Communication Monographs, 68 (4), 347. Morris, Jon D. (1995) Observations: SAM: the Self-Assessment Manikin: an Efficient Crosscultural Measure of Emotional Response Journal of Advertising Research, 35(6), 63-68. __________, and James S. McMullen (1994), Measuring Multiple Emotional Responses to a Single Television Commercial, Advances in Consumer Research, 21(1), 175-180.
81 __________, Chongmoo Woo, James A. Geason, and Jooyoung Kim (2002), The Power of Affect: Predicting Intention, Journal of Advertising Research, 4(3), 7-17. Murry, John P., John L. Lastovicka, and Su rendra N. Singh (1992) Feeling and Liking Responses to Television Programs: An Examination of Two Explanations for Media-Context Effects Journal of Consumer Research, 18(4), 441-451. Muehling, Darrel D. (1987), An Investigati on of Factors Underlying Attitude-Toward Advertising-in-General, Journal of Advertising 16(1), 32-40. Murry, John P., John L. Lastovicka, and Surendra N. Singh (1992), Feeling and Liking Responses to Television Programs: An Examination of Two Explanations for Media-Context Effects Journal of Consumer Research, 18(4), 441-451. ____________, and Peter A. Dacin, Cognitive Modera tors of Negative-Emotion Effects: Implications for Understanding Media Context, Journal of Consumer Research, 22(4), 439-447. Newell, Stephen J, Kenneth V. Henderson, Bo T. Wu (2001), The Effects of Pleasure and Arousal on Recall of Advertisements during the Super Bowl, Psychology and Marketing, 18(11), 1135-1152. Norris, Claire E., Andrew M. Colman (1993) Context Effects on Memory for Television Advertisements, Social Behavior and Personality, 21(4), 279-296. Olney, Thomas J., Morris B. Holbrook, and Raje ev Batra (1991), Consumer Responses to Advertising: The Effects of Ad Content, Emo tions, and Attitudes toward the Ad on Viewing Time, Journal of Consumer Research, 17(4), 440-453. Park, Se-Hyuk, and Yong-Man Kim (2000), Concep tualizing and Measur ing the Attitudinal Loyalty Construct in Recreational Sport Contexts, Journal of Sport Management 14(3), 197207. Pavelcahk, Mark, John H. Antil, James M. Munc h (1988), The Super Bowl: An Investigation into the Relationship among Program Context, Emotional Experience, and Ad Recall, Journal of Consumer Research 15(3), 360-367. Petty, Richard E., and John T. Cacioppo (1981), Attitudes and Persuasion: Classic and Contemporary Approaches, Dubuque, IA: William C. Brown. Pham, Michel Tuan (1996), Cue Representa tion and Selection Effects of Arousal on Persuasion, Journal of Consumer Research, 22(1), 373-387. ________________, Joel B. Cohen, John W. Pracejus, and G. David Hughes (2001), Affect Monitoring and the Primacy of Feelings in Judgment, Journal of Consumer Research, 28(2), 167-188. Plutchik, Robert (1980), Emotion: A Psychoevolutionary Synthesis. New York: Harper & Row.
82 Pyun, Do Young (2006), The Proposed Model of Attitude Toward Advertising Through Sport, Unpublished Doctoral Dissertati on, Florida State University. Raghunathan, Rajagopal, and Michel T. Pham (1999), All Negative Moods Are Not Equal: Motivational Influences of Anxi ety and Sadness on Decision Making, Organizational Behavior and Human Decision Processes, 79 (1), 56. _________________, and Galen T. Trail (2005), Relat ionships Among Spectator Gender, Motives, Points of Attachment, and Sport Preference, Journal of Sport Management 19, 58-80. _________________, _______________, Ronald J. Dick, and Andrew J. Gillentine (2005), Fans vs. Spectators: An Analysis of Those Who Attend Intercollegiate Football Games, Sport Marketing Quarterly, 14(1), 43-53. Roese, Neal J., and Sameep D. Maniar (1997) Perceptions of Purple: Counterfactual and Hindsight Judgments at Northwes tern Wildcat Football Games, Personality and Social Psychology Bulletin 23, 1245-1253 Russell, James A., and Lisa Feldman Barrett ( 1999), Core Affect, Prototypical Emotional Episodes, and Other Things Called Em otion: Dissecting the Elephant, Journal of Personality and Social Psychology, 76(5), 805-819. _____________, Anna Weiss, and Gerald A. Mendels ohn (1989), Affect Grid: A Single Item Scale of Pleasure and Arousal, Journal of Personality and Social Psychology 57(3), 493-502. Schumann, David W. (1986), "Program Impact on Attitude Toward TV Commercials," in Proceedings of the Division of Consumer Psychology Joel G. Saegert, ed., Washington, DC: American Psychological Association, 67-73. Schurr, K. T., Ruble, V. E., & Ellen, A. S. (1985), Myers-Briggs Type Inventory and Demographic Characterist ics of Students Attending a nd not Attending a College Basketball Game Journal of Sport Behavior 8, 181-194. Shank, Mathw D., and Fred M. Beasley (1998), Fan or Fanatic: Refining a Measure of Sports Involvement, Journal of Sport Behavior, 21(4), 435443. Shapiro, Stewart, Deborah J. MacInnis, and C. Whan Park (2002), Understanding ProgramInduced Mood Effects: Decoupling Arousal from Valence, Journal of Advertising, 31 (4), 15 26. Sherif, Muzafer, and Carl I. Hovland (1961), Social Judgment: Assimilation and Contrast Effects in Communication and Attitude Change. New Have, CT: Yale University Press. Sloan, Lloyd R. (1979), "The Function and Impact of Sports for Fans: A Review of Theory and Contemporary Research," In J. H. Goldstein (ed.), Sports, games, and play: Social and psychological viewpoints, 219-262. New York: Erlbaum.
83 Snyder, C.R., Maryanne Lassegard, and Carol E. Ford (1986), Distancing After Group Success and Failure: Basking in Reflected Glor y and Cutting Off Reflected Failure, Journal of Personality and Social Psychology, 51(2), 382-388. Stout, Patricia A., and John D. Leckenby (1986), "Measuring Emotional Response to Advertising," Journal of Advertising 15(4), 35-42. Sutton, William A, Mark A. McDonald, and Geor ge R. Milne (1997), Creating and Fostering Fan Identification in Professional Sports, Sport Marketing Quarterly, 6(1), 15-22. Tobar, David (2006), Affect and Purchase Inten tions of Super Bowl XL Television Spectators: Examining the Influence of Sport Fandom, Age, and Gender, Sport Marketing Quarterly, 15, 243-252. Trail, Galen T., Dean F. Anderson, and Janet S. Fink (2000), A Theoretical Model of Sport Spectator Consumption Behavior, International Journal of Sport Management, 1, 154-180. Trail, Galen T., and Jeffrey D. James (2001), The Motivation Scale fo r Sport Consumption: Assessment of the Scales Psychometric Properties, Journal of Sport Behavior, 24, 108-127. ____________, Matthew J. Robinson, Ronald J. Dick, and Andrew J. Gillentine (2003), Motives and Points of Attachment: Fans Vers us Spectators in Intercollegiate Athletics, Sport Marketing Quarterly, 12(4), 217-227. Wann, Daniel L. (1995), Preliminary valida tion of the Sport Fa n Motivation Scale, The Journal of Sport & Social Issues, 19, 377-396. ______________ (2002), Preliminary Validation of a M easure for Assessing Identification as a Sport Fan: The Sport Fandom Questionnaire, International Journal of Sport Management, 3(2), 103-115. _____________, and Nyla R. Branscombe (1990), Die-H ard and Fair-Weather Fans: Effects of Identification on BIRGing and CORFing Tendencies, Journal of Sport and Social Issues, 14(2), 103-117. _____________, and _______________ (1993), Sports Fans: Measuring Degree of Identification with Their Team, Journal of Sport Psychology, 24(1), 1-17. _____________, and Thomas J. Dolan (2001), Attrib utions of Highly Identified Sports Spectators, The Journal of Social Psychology, 134(6), 783-792. _____________, _________________, Kimberly K. McGeorge, Julie A. Allison (1994), Relationships Between Spectator Identification and Spectators Perceptions of Influence, Spectators Emotions, and Competition Outcome, Journal of Sport & Exercise Psychology, 16(4), 347-364. _____________, and Stephanie Pierce (2003), Measur ing Sport Team Identification and Commitment: An Empirical Comparison of the Sport Spectator Identification Scale and the
84 Psychological Commitment to Team Scale, North American Journal of Psychology, 5(3), 365372. Zaichkowsky, Judith L (1985), Measu ring the involvement construct. Journal of Consumer Research 12, 341-352. Zillmann, Dolf, Bryant J. Jennings, and Barry S. Sapolsky (1979) "The Enjoyment of Watching Sports Contests." In J. H. Goldstein (ed.), Sports, Games, and Play: Social and Psychological Viewpoints, 297-335. New York: Erlbaum.
85 BIOGRAPHICAL SKETCH Michael J. C layton earned his B.A. in marketing from Miami University in 1999 and his M.A. in advertising and P.R. from Michigan State University in 2006. He began his doctoral program in mass communication at the Univer sity of Florida in 2006. His professional experience includes over seven years of account management and account planning work at BBDO and Campbell-Ewald on multiple national cl ients including Chevrolet, Dodge, Michelin, Pearle Vision, GMAC, and Conoco.