1 TEAM IDENTIFICATION AND COMPETITION OUTCOME ALTER ATTENTIONAL BREADTH AND FLEXIBILITY: AFFECTIVE AND BEHAVIORAL CONSEQUENCES By MELANIE BLYTH MOUSSEAU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF F LORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Melanie Blyth Mousseau
3 To all of the shoulders Ive leaned onThank you! Without your strong and constant support this would never have been possible.
4 ACKNOWLEDGMENTS I would first like to thank my mentor and committee chair, Dr. Christopher Janelle, who has always allowed me to pursue the topics that fascinate me, while challenging me to achieve my fullest academic potential. Through his dedication and commitment, I have been able to grow exponentially as a writer, teacher, and scientist. I would also like to thank my committee members, Dr. James Cauraugh, Dr. Kyriaki Kaplanidou, and Dr. Tracy Linderholm for their ti me and insightful feedback throughout all of the stages of my dissertation. Additionally, John van den Boogaards flexibility and dependability during the data collection phase was invaluable. I also want to extend my appreciation to the entire Psychology Department at the University of Mary Washington. The research and academic opportunities that I was afforded while an undergraduate student provided the solid foundation from which my entire graduate education was built. More personally, I would like to recognize the unconditional support and constant encouragement I received from my parents, Helene and John, brother, Jonathan, grandmothers, Helen and Jen, the spirit of my grandfathers, Charlie and John, and all of my friends, near and far. Finally, eve ry step of this journey has been shared with my future husband, Dr. Derek Mann, and I want to thank him for being the bright light at the end of every tunnel. He made the countless hours at Starbucks, the frustrating bouts of writers block, and the minor (and major) setbacks manageable, and sometimes even fun. Now we are both Phinally Done!
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 8 LIST OF FIGURES .............................................................................................................................. 9 ABSTRACT ........................................................................................................................................ 11 CH A P T E R 1 INTRODUCTION ....................................................................................................................... 13 Correlates of Sport Fandom ........................................................................................................ 14 Understanding the Highly Identified Fan .................................................................................. 16 Motivational Orientation ............................................................................................................. 18 Limitations of Previous Work .................................................................................................... 19 Statement of the Purpose ............................................................................................................ 20 The Current Study ....................................................................................................................... 21 2 REVIEW OF LITERATURE ..................................................................................................... 25 Identifying and Classifying Sports Fans .................................................................................... 25 Psychological Correlates of Sport Fandom ............................................................................... 27 BIRGing, CORFing, and blasting ....................................................................................... 28 Identification and Perception .............................................................................................. 34 Perceptions of Self ............................................................................................................... 38 Emotional Experience of Fans ............................................................................................ 43 Physiological Fluctuations and Fan Identification ............................................................ 47 The Connection Between Team Identification and Psychological Health .............................. 54 Attentional Bias ........................................................................................................................... 62 Affective Disorders .............................................................................................................. 65 Chemi cal Use and Dependence .......................................................................................... 66 Attentional Breadth and Flexibility ............................................................................................ 68 Limitations in the Literature ....................................................................................................... 71 Conclusion ................................................................................................................................... 73 3 METHODS .................................................................................................................................. 75 Participants .................................................................................................................................. 75 Materials ...................................................................................................................................... 75 Self Report Measures .......................................................................................................... 75 Demographic questionnaire ......................................................................................... 75 Sport Spectator Identification Scale ............................................................................ 75 Sport Fandom Questionnaire ....................................................................................... 76
6 State Trait Anxiety Inventory ...................................................................................... 76 Positive and Negative A ffect Schedule ....................................................................... 76 Behavioral Inhibition System / Behavioral Activation System Scales ..................... 77 Beck Depression Inventory ......................................................................................... 77 Movement Imagery Questionnaire Revised ............................................................... 78 Visual analog s cales ..................................................................................................... 78 Experimental Stimuli and Task ........................................................................................... 79 Emotional imagery scripts ........................................................................................... 79 Stroop task .................................................................................................................... 80 Global local target detection task ................................................................................ 80 Procedure ..................................................................................................................................... 82 4 RESULTS .................................................................................................................................... 86 Psychological Variables .............................................................................................................. 87 Changes in Emotion .................................................................................................................... 87 Changes in Attention ................................................................................................................... 89 Stroop Task .......................................................................................................................... 89 Global Local Target Detection Task .................................................................................. 91 Supplemental Explor atory Analyses .......................................................................................... 96 Temporal Effects .................................................................................................................. 97 Stroop task performance across trial blocks ............................................................... 97 Global local task performance across trial blocks ..................................................... 99 Ang er .................................................................................................................................. 100 Anger reactivity and Stroop task performance ......................................................... 101 Anger reactivity and global local target detection task performance ..................... 101 5 DISCUSSION ............................................................................................................................ 104 Review of the Findings ............................................................................................................. 104 Psychological Variables .................................................................................................... 104 Emotion Induction ............................................................................................................. 105 Changes in Attention ......................................................................................................... 108 Stroop task .................................................................................................................. 108 Global local target detection task .............................................................................. 111 Strengths, Limitations, and Directions for Future Research .................................................. 112 Implications and Conclusions ................................................................................................... 116 APPENDIX A DEMOGRAPHIC QUESTIONNAIRE ................................................................................... 119 B SAMPLE IMAGERY SCRIPTS .............................................................................................. 120 C INFORMED CONSENT .......................................................................................................... 121
7 LIST OF REFERENCES ................................................................................................................. 123 BIOGRAPHICAL SKETCH ........................................................................................................... 133
8 LIST OF TABLES Table page 4 1 Descriptive statistics and results of univariate analysis of variance of self report measures. Asterisks (*) denote significant di fferences ( p < .05) between groups. ............ 86
9 LIST OF FIGURES Figure page 2 1 The Team Identification -Social Psychological Health Model (adapted from Wann, 2006b). .................................................................................................................................... 60 3 1 Visual analog scales. .............................................................................................................. 79 3 2 Experimental design. .............................................................................................................. 83 4 1 VAS ratings by condition across participants. ..................................................................... 89 4 2 Identification by condition VAS sadness ratings. ................................................................ 89 4 3 Stroop task response times on incongruent trials. ................................................................ 90 4 4 Stroop task performance marginal means by condition accounting for SFQ. .................... 91 4 5 Global and local target detection response times by condition. .......................................... 92 4 6 Global and local target detection response times by identification. .................................... 93 4 7 Condition by identification global local target detection task performance. ..................... 93 4 8 Global local task performance marginal means by condition accounting for SFQ. Asterisks (*) denote significant differences ( p < .008) between conditions. ..................... 94 4 9 Local task performance marginal means by identification accounting for number of years a fan. Asterisks (*) denote significant differences ( p < .05) between groups. ......... 94 4 10 Global local task performance marginal means by identification accounting for state level of anger. Asterisks (* and **) denote significant di fferences ( p < .05) between groups. ..................................................................................................................................... 95 4 11 Global local task performance marginal means by identification accounting for state level of sadness. Asterisks (* and **) denote significant differences ( p < .05) between groups. ...................................................................................................................... 96 4 12 Global task performance marginal means by identification accounting for general negative affect. Asterisks (*) denote significant differences ( p < .05) between groups. ... 96 4 13 High and low identified fans Stroop task response times during the neutral condition across trial blocks. .................................................................................................................. 98 4 14 High and low identified fans Stroop task response times during the unpleasant condition across trial blocks. ................................................................................................. 98 4 15 Response times by trial block to global and local stimuli across affective conditions. ..... 99
10 4 16 High and low identified fans response times by trial block to global stimuli across affective conditions. ............................................................................................................. 100 4 17 Anger group by condition Stroop task response times. ..................................................... 102 4 18 Anger group by condition global target detection task performance ............................... 1 02 4 19 Anger group by condition local target detection task performance .................................. 103 5 1 The propo sed SEM model demonstrates that attentional breadth and flexibility will vary as a function of mood state .......................................................................................... 115
11 Abstract of Disser tation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TEAM IDENTIFICATION AND COMPETITION OUTCOME ALTER ATTENTIONAL BREADTH AND FLEXIBILITY: AF FECTIVE AND BEHAVIORAL CONSEQUENCES By Melanie Blyth Mousseau August 2009 Chair: Christopher M. Janelle Major: Health and Human Performance The considerable consumption and importance of sport in the lives of Americans underscores the necessity to furth er understand the mechanisms and consequences of sport fandom and team identification. This study integrates what is known about emotion, motivation, and attention with current understanding of sport fandom to identify the affective and behavioral conseque nces of team identification. I sought to determine how pleasant and unpleasant fan experiences influence attention and performance on tasks that are known to rely on appropriate attentional allocation. I hypothesized that the emotions evoked from positive and negative fan experiences would lead to differences in motivational orientation, thereby promoting changes in attentional breadth and flexibility. Imagery scripts were used to conjure images that resulted in pleasant and unpleasant emotional reactions among high and low identified fans. Performance differences in global local target detection and Stroop tasks followed emotional imagery when other team relevant variables were statistically covaried. Specifically, faster response times to incongruent S troop stimuli followed exposure to unpleasant imagery scripts compared with performance after a neutral, rest condition, suggesting that attentional flexibility was enhanced by the affective condition as a function of the general level of fandom. When acco unting for the number of years as a fan of the
12 target team, as well as state anger and sadness reactivity, highly identified fans exhibited superior performance on both global and local target detection tasks compared to their less identified counterparts. Together, these findings suggest that highly identified fans experience a domain specific increase in arousal when presented with team relevant content (positive or negative), which in turn can facilitate performance. The affective and behavioral correlates of team identification therefore appear to be highly contextual, but are strongly tied to other related constructs such as general sport fandom, the number of years having been a fan, and general affective reactivity.
13 CHAPTER 1 INTRODUCTION Two hundred million adults in the United States consider themselves sports fans (King, 2005). As an index of sports popularity, retail sales of sport merchandise exceeded $13 billion in the United States and Canada in 2006 (Lefton, 2007). In a recent survey avid sports fans reported rearranging their daily schedules to listen to, watch, or attend a sporting event, and nearly 40% of respondents reported choosing sporting events in lieu of other professional and personal obligations (Wann, Friedman, McHale, & Jaffe, 2003). Fans are not merely consumers of sport, they are psychologically invested in sports, teams, and athletes (Wann, Melnick, Russell, & Pease, 2001). The label sport fan is often an important and central component of self -concept among highl y identified fans. Highly identified fans possess a greater sense of belonging to and attachment with a particular team. Being a supporter of their team is a crucial element of who fans are (i.e., their identity), and the team is often perceived as an ex tension of themselves (Hirt, Zillmann, Erickson, & Kennedy, 1992). This strong identification among sports fans has been likened to the identification people possess in regards to their gender, nationality, and ethnicity. For less identified fans, the conc ept of team supporter is much less pertinent. These individuals are much less psychologically invested in and impacted by the teams performance. While great variability exists in the factors that contribute to adopting the role of a fan (e.g., spending ti me with other people/family, appreciation of the aesthetic elements of sport, the experience of eustress, escape from their lives/responsibilities, enhancement of self -esteem, simple enjoyment as a pastime), all fans vicariously experience the thrill of vi ctory and the agony of defeat (Wann et al., 2001).
14 Intense emotional reactions often accompany high levels of fan identification. In a study of 1,442 avid sports fans, 44% of the fans reported that they needed at least one day to emotionally recover from a loss, and approximately 64% of respondents noted that they had cried after their team lost a game (Wann et al., 2003). Participants also reported being similarly effected by positive outcomes, with 45% requiring at least one day to come down from the e xcitement and 68% reported having cried after a victory. This snapshot of fandom extends beyond merely highlighting the considerable consumption and importance of sport in the lives of Americans, it also illustrates the profound influence that sport can h ave on the lives of those who classify themselves as sports fans. Correlates of Sport Fandom Team identification has been repeatedly shown to be a key factor in predicting sport related behavior, including event attendance and purchasing team licensed merc handise (Kwon & Armstrong, 2002; Laverie & Arnett, 2000; Wakefield & Wann, 2006). Moreover, identification is significantly correlated with a variety of other, non -sport related variables. Team performance has been shown to have a significant influence on fans general outlook on life, well -being, overall life satisfaction (e.g., Schwarz, Strack, Kommer, & Wagner, 1987), perception of fearful future events (e.g., Schweitzer, Zillmann, Weaver, & Luttrell, 1992), and estimates of their own abilities (e.g., Hirt et al., 1992). For example, Schwarz et al. (1987) reported more positive evaluations of global well -being among fans after their national soccer team won, but less favorable evaluations following a tie. Schweitzer and colleagues (1992) documented incr eased enjoyment and happiness among fans of a winning team and decreased enjoyment and increased disappointment among the losing teams fans. Following their teams loss, participants also reported a greater perceived likelihood of war compared to those re ported by the fans of the winning team. Highly identified fans also interpret their teams successes and failures personally
15 (Hirt et al., 1992). Thus, not only do fans increase their expectations for future team success following a victory while reducing their expectations after a loss, highly identified fans also show a similar pattern of biased estimates of their own cognitive abilities and social skills. Research has clearly shown that sports fans are impacted emotionally by sport and these effects extend beyond the sporting context and into the daily lives of fans, implying that being a highly identified sports fan consists of more than simply rooting for and supporting ones team. For many fans, identification with a team is central to their self -co ncept. Team performance, however, is often inconsistent, and poor performance can promote a decrease in ones self and public image. People therefore engage in various self -presentation and enhancement strategies. Specifically, Cialdini, Borden, Thorn, Wal ker, Freeman, and Sloan (1976) found that people tend to publicly emphasize their connection with successful others (regardless of the strength of their connectedness) and bask in reflected glory (BIRG) to increase their public image. Grounded in Heiders (1958) balance theory, Cialdini et al. (1976) proposed that BIRGing occurs because people evaluate connected parties similarly in order to maintain cognitive harmony. As a result, public association with a successful team is perceived to promote positive evaluations (i.e., enhanced public image). Conversely, Snyder, Higgins, and Stucky (1983) suggested that people will also engage in the protective behavior of decreasing the association between oneself and unsuccessful others by cutting off reflected failure (CORF). This tactic is not viable for everyone, however. Highly identified fans are unable and/or unwilling to distance themselves from their team when they are unsuccessful (i.e., CORF) because their affiliation is a significant component of their personal identity (Wann & Branscombe, 1990), yet they have the natural desire to maintain a positive self image. Therefore, to cope with negative vicarious experiences while preserving their public
16 image and self -esteem, highly identified fans will engage in blasting, a defensive behavior where an individual publicly devalues those with whom they are negatively associated (i.e., the opposition) (Branscombe & Wann, 1992; Wann, 1993). Wann (1993) explained that blasting can extend beyond the verbal degradation of relevant out -groups (e.g., opposing teams fans) and include physically aggressive behaviors (i.e., spectator violence) on the part of highly identified fans. Although both CORFing and blasting are psychological strategies used by fans to protect their self -concept in the face of team failure, they are inherently different. When CORFing, the individual disconnects him/herself from the target group (i.e., exhibiting avoidant behavior), yet with blasting, the individual actively engages in the proce ss of identity control. These divergent patterns of behavior have also been linked with the qualitatively different emotions that fans experience in the face of failure ( Crisp, Heuston, Farr, & Turner, 2007). Although both high and low identified fans rep ort negative emotions following their teams loss, Crisp and colleagues (2007) have shown that highly identified fans feel angry, but not sad when their identity is threatened by a vicarious failure. Low identified fans in contrast, report sadness but not anger in response to the same experience. Together these findings suggest that the way a fan affectively appraises the situation is critical in determining how they will react behaviorally (i.e., engage/aggress or withdraw) and that their level of identif ication moderates their emotional experience [i.e., the higher the level of identification, the greater the level of affective engagement ( Hillman, Cuthbert, Bradley, & Lang, 2004; Hillman, Cuthbert, Cauraugh, Schupp, Bradley, & Lang, 2000) ]. Understanding the Highly Identified Fan The emotional experiences and behavioral correlates found among highly identified sport fans are similar to those of individuals with various psychopathologies (e.g., affective disorders,
17 substance dependence). The first parall el that can be drawn between these two groups is how their daily lives are altered as a result of their identification/pathology. As mentioned, it is not uncommon for an avid sports fan to skip work, school, or important social engagements in favor of watc hing or attending a competition (St. John, 2004; Wann et al., 2003). Because the self concept of such highly identified fans is rooted in the team and its performance (Wann et al., 2001), sport consumption has a tendency to take precedence over other life events, resulting in preoccupation with the team and excessive consumption of team related content (e.g., radio, internet, television), ultimately disrupting interpersonal and professional relationships (Quirk, 1997; St. John, 2004). Likewise, addicts will compulsively engage in a specific behavior (e.g., drug use, alcohol consumption) at the expense of other personal responsibilities, while phobics and individuals with chronic anxiety will alter their pattern of behavior and their surrounding environment i n order to avoid potential exposure to threatening stimuli (Williams, Watts, MacLeod, & Mathews, 1997). For each of these populations, there is an intense emotional connection and reactivity to a particular target (e.g., sports team, narcotics, alcohol, s piders, social situations). While by no means are these groups identical, they do share similar affective reactions to specific stimuli and demonstrate comparable behaviors, suggesting that these similarities may be motivated by similar underlying factors. Motivational orientation has recently gained much attention from researchers (e.g., Coombes, Cauraugh, & Janelle, 2007; Derryberry & Reed, 1998; Gable & HarmonJones, in press; Harmon Jones, Lueck, Fearn, & Harmon Jones, 2006) seeking to better understan d the association between emotion, behavior, and psychological problems such as affective disorders and addictions. Similarly, the distinct emotional reactions and behavioral tendencies among fans
18 may also be explained through gaining understanding of the motivational orientation of highly identified fans. Motivational Orientation Given that attention drives subsequent behavior, of particular interest here are the catalysts that direct and moderate both the breadth and direction of attention. Motivational orientation can have an impact on both perceptual (i.e., the allocation of resources to central and peripheral environmental stimuli) and conceptual (i.e., the activation of cognitive representations) attention, as well as attentional flexibility (Derryberr y & Tucker, 1994; Frster et al., 2006). Approach behavioral tendencies (i.e., engagement) have typically been associated with a broader scope of attention and greater attentional flexibility (i.e., the ability to adaptably shift focus among cognitive ope rations using executive control [p. 70, Friedman & Frster, 2005]), while avoidance (i.e., withdrawal) tendencies restrict the scope of attention (Friedman & Frster, 2005; Frster et al., 2006). Researchers initially suggested that this behavioral dichot omy mirrored affective valence, with unpleasant/negative affect being exclusively linked with an avoidance orientation and pleasant/positive affect with an approach orientation. However more recent research (e.g., Coombes et al., 2007; Davidson, 1995; Harm onJones et al., 2006) suggests that such a pairing is inaccurate, as it is the motivational orientation, not simply affective valence, which influences behavior. Of the core emotions, anger exemplifies an unpleasant affective state that promotes an approach motivational orientation rather than the typical avoidance pattern seen with other unpleasant emotions such as sadness (HarmonJones et al., 2006). Gable and Harmon Jones (in press) suggest that the magnitude of the ensuing approach motivation is critical in moderating the breadth of attention. Specifically, positive affect that is relatively low in approach (i.e., nongoal relevant emotions such as joy and amusement) has been tied to a broader scope of attention, as there is perception of a stable, saf e environment. Reduced
19 attentional broadening, however, is evident following the induction of positive affect that is higher in approach motivation (e.g., desire and enthusiasm). In the latter case, it is necessary to disregard irrelevant stimuli when acti vely approaching a goal (Harmon & Jones, in press). The opportunity/expectancy to act in a manner that promotes goal attainment, therefore, reduces the scope of attention, despite being fostered by a positive, approach emotion. In sum, emotions alter the m anner in which attention is allocated and information is extracted from the environment through the promotion of approach avoidance motivational orientation and these changes are known to underlie behavioral variability. As Hillman and colleagues (2000, 2004) have illustrated, highly identified sports fans exhibit motivated emotional reactivity to team relevant stimuli. Coupling this with our understanding of the relationship between motivational orientation, attention, and behavior, motivational orientat ion is likely a critical core mechanism that drives fan specific behaviors (e.g., BIRGing, CORFing, blasting, spectator violence) and influences the daily functioning of fans. To better understand the nature and consequences of fandom, it is essential to s ystematically examine the affective component of the fan experience, as the emotions evoked from sport spectating may underlie and alter the motivational orientation, attentional breadth and direction, and ultimately the behaviors of this population. Limit ations of Previous Work Although our understanding of sport fandom has advanced over the last two decades, the existing body of literature is limited due to a number of theoretical, methodological, and interpretative shortcomings. First, there has been rel ative inconsistency in the operational definition of fan and the classification of high and low identified fans. Specifically, researchers have used divergent indices (e.g., measures of general fandom and sport/team specific identification, self -classifi cation, consumption behaviors, and study specific questionnaires) to
20 conceptualize their target population. As a result of this variability, limited generalizations can be drawn from existing research, thereby leaving questions concerning the nature, corre lates, and consequences of varying levels of identification unanswered. Next, the majority of what is known about the psychology of sports fans is grounded in correlational evidence. As such, the direction of the relationship between identification and fans thoughts, emotions, and behaviors is uncertain. Moreover, the data extracted from studies exploring the psychology of sports fans has been one -dimensional. That is, researchers have either taken a subjective approach using self -report assessments of thoughts, feelings, and behaviors (e.g., Schwarz et al., 1987; Schweitzer et al., 1992; Wann et al., 2003) or have employed more covert methods, monitoring physiological fluctuations in response to sports cues (e.g., Hillman et al., 2000, 2004). Although e ach approach has yielded valuable information pertaining to the correlates and consequences of fandom, an incomplete perspective of fandom exists. In particular, direct behavioral consequences associated with sport fandom have been largely ignored. More sy stematic exploration of the cognitive, behavioral, and affective factors associated with team identification in controlled environments can provide insight into how fans process information, and specifically, how sport relevant stimuli influence cognitive processing, attention, daily functioning, and factors associated with mental health issues. Statement of the Purpose Nearly 67% of the United States population characterize themselves as sports fans (King, 2005). Understanding how fan identification imp acts general well -being and functioning is therefore of broad importance. To date, our understanding of the psychology of sports fans is limited. While there are countless anecdotal reports about sports fanatics, there remains a lack of empirical underst anding of this group due to the combination of relatively minimal research being carried out on this topic and the extant literature being hindered by theoretical and
21 methodological shortcomings. Therefore, the current study systematically explored the cog nitive, behavioral, and affective factors associated with team identification in a controlled environment, with the purpose of identifying whether team identification and competition outcome alter attentional breadth and flexibility as indicated by perform ance on global -local target detection and Stroop tasks. The Current Study The considerable consumption and importance of sport in the lives of Americans underscores the necessity to further understand the nature and consequences of sport fandom and team id entification, including delineation of the functional costs and benefits of high level of identification. I designed the present study to test the central hypothesis that the emotions evoked from positive and negative fan experiences lead to attentional ch anges (i.e., broadening/ narrowing of attention) derived from differences in motivational orientation which are reflected in performance changes. Perceptual and conceptual attentional alterations were assessed through performance on global local target det ection (Frster et al., 2006; Navon, 1977) and Stroop tasks (Friedman & Frster, 2005; Stroop, 1935), respectively. The global local target detection task assesses attentional scope (i.e., the breadth of attention) and the Stroop task measures attentional flexibility (i.e., the ability to shift attention). Emotional imagery was used to induce pleasant (e.g., happy, elated, excited) and unpleasant (e.g., sad, angry, frustrated) emotional states among participants. In the pleasant conditions, participants li stened to narratives that guided them through their most pleasant memories generated as a University of Florida football fan. While it was hypothesized that these imagery scripts would promote an increase in positive affect in both high and low identified fans, it was expected that the magnitude would differ, with the highly identified fans experiencing the largest increase in positive emotions. The resulting positive affect was expected to promote an
22 approach orientation, as research has shown both high and low identified fans exhibit a tendency to BIRG (Cialdini et al., 1976). Approach orientation was predicted to promote greater attentional flexibility thereby increasing the ability to shift between salient content while broadening the scope of attention (both conceptual and perceptual). These changes were predicted to enhance Stroop task performance as well as detection of global targets (Friedman & Frster, 2005; Frster et al., 2006). Moreover, approach orientation magnitude was manipulated and an expe ctancy to act was induced by providing fans with the opportunity to draft a brief statement about what it means to be a fan. Although pleasant/positive emotions and an approach orientation would be evoked through this manipulation, it is hypothesized that attentional scope would be less broad and flexible due to the propensity to ignore irrelevant stimuli, with the most notable changes expected among highly identified fans as they are more invested in the team and its performance compared to their less iden tified counterparts (Gable & Harmon Jones, in press). Therefore, enhanced performance would not be evident on the Stroop and global target detection tasks, as approach motivated positive affect prompts a decreased broadening of attention. In the unpleasant conditions, participants listened to narratives that guided them through their most unpleasant memories generated as a University of Florida football fan. It was hypothesized that the unpleasant imagery scripts would increase negative affect among fans, with highly identified fans reporting the largest increase in anger (i.e., approach), and less identified fans reporting a significant increase in sadness (i.e., avoidance) (Crisp et al., 2007). Because high and low identified fans have been shown to exper ience qualitatively different emotions and exhibit variable behavior when their identity is threatened (Branscombe & Wann, 1992; Crisp et al., 2007; Wann, 1993), I hypothesized that the two groups would have disparate underlying motivational tendencies. Hi ghly identified fans were expected to manifest an
23 approach orientation because of their propensity to experience anger and blast in the face of failure/identity threat. Again, this approach orientation was hypothesized to improve performance on the Stroop task and the detection of global targets on the global local target detection task due to the broadening of attentional scope and greater attentional flexibility. In contrast, research (e.g., Branscombe & Wann, 1992; Crisp et al., 2007; Wann 1993) has show n that less identified fans typically report sadness when their team fails, and engage in CORFing. It was therefore hypothesized that less identified fans would engage in an avoidance orientation, prompting decreased attentional flexibility and scope. Thes e changes would be evident in decreased performance on the Stroop task (i.e., reduced attentional flexibility) and detection of global targets in the global -local target detection task, yet increased performance on the recognition of local targets in the s ame task (i.e., narrowed scope of attention). Finally, although high and low identified fans were expected to react differently to emotionally evocative stimuli, these differences were hypothesized to be specific to varying levels of identification and not to differences on measures of personal psychological well -being (e.g., depression, anxiety, general affect) (Wann, 1994; Wann, Dunham, Byrd, & Keenan, 2004). Therefore, I hypothesized that differences would not emerge between high and low identified fans on measures of trait anxiety, depressive symptomology, trait approachavoidance tendencies, and general experience of positive and negative affect (i.e., personal psychological well -being). In sum, this study integrates what is known about emotion, motivation, and attention with our current understanding of sport fandom to fulfill my objective of identifying the functional consequences of team identification. While highly identified fans may possess psychologically healthy profiles, I predicted that their emotional reactivity and ensuing behaviors in sport specific
24 situations would manifest in a manner similar to those demonstrated by individuals with affective disorders. Attention is inherent in all aspects of daily functioning, as it drives how we extrac t and subsequently process, store, retrieve, and apply information. By gaining quantitative insight in to the affective significance of sport fandom we will be able to better understand the functional consequences of identification and develop an awareness of the personal, social, and behavioral effects that are tied to team performance.
25 CHAPTER 2 REVIEW OF LITERATURE Sport is a multibillion dollar business (Rascher, 2005) that has impacted contemporary society, with approximately 200 million adult Americans identifying themselves as sports fans (King, 2005). These numbers continue to increase as sports become more accessible to consumers through avenues such as team and league websites, fantasy leagues, all -sport television networks and radio stations, as well as mobile telephone services that provide the most up -to -the -minute sporting news and statistics. Recognizing the widespread prevalence of sport fandom, researchers have taken a keen interest in understanding the psychological correlates of this s ocial phenomenon. This literature is summarized and critically reviewed below. Moreover, a conceptual link is made between the cognitive hallmarks of clinical disorders and the explicit actions of highly identified fans. Identifying and Classifying Sports Fans Both in the popular media and scientific reports, the terms fan and spectator are often used interchangeably. However the two terms represent orthogonal classifications. Spectators are consumers of sport (i.e., those who watch competitions live o r otherwise), while fans (abbreviation for fanatics) are individuals who actively follow and are interested in sport, a team, and/or an athlete (Wann, Melnick, Russell, & Pease, 2001). There are, however, countless individuals who encapsulate elements of both spectators and fans (i.e., have an inherent interest in sport and attend competitions), yet vary in their level of psychological investment. That is, the degree to which one is interested in a sport, team, or athlete ranges across a continuum. At one end of the continuum are individuals who possess a distinct psychological connection to a team. For these highly identified fans being a supporter of their team is a crucial element of who they are (i.e., their identity), and the team is often perceived as an extension of themselves. This
26 strong identification with a sports team has been likened to the identification people possess in regards to their gender, nationality, and ethnicity. In contrast, for less identified fans, the concept of team supporter is much less pertinent. These individuals are much less psychologically invested in and impacted by the teams performance (Wann et al., 2001). Although sports fan research was prevalent prior to 1993, Wann and Branscombes (1993) Sport Spectator Identification Scale (SSIS) was the first measure of team identification to be validated and published for widespread use. The SSIS is a 7 item, 8 point Likert -scale inventory designed to assess individual allegiance or identification with a sports team (Wan n & Branscombe, 1993, p. 3). In initial validation studies, highly identified fans ( M = 7.15) displayed greater involvement and investment in their team, more positive attributions regarding their teams performance, were more optimistic in their predictio ns for the teams future performance, perceived other fans of the same team as special, and reported that it was more important that their friends also be fans compared to those participants classified as moderately ( M = 5.88) or low ( M = 3.55) identified fans. Wann and Branscombe (1993) concluded that the SSIS is a useful and predictive tool for assessing fans levels of identification with sports teams. Following its validation, the SSIS has become one of the primary tools used in sport fan identifica tion research. It is probable, however, that there are individuals who can be classified as highly identified general sports fans who do not identify with a specific team or player (Wann, 2002). Given that the SSIS is limited to measuring only identification with a specific team (or athlete) and does not quantify an individuals level of identification as a general sport fan (i.e., simply enjoys the sport itself), Wann (2002) developed the Sport Fandom Questionnaire (SFQ). The SFQ is a psychometrically va lid five item self -report questionnaire designed to assess an individuals self -
27 perception of being a sport fan and the degree to which one identifies him/herself as a fan of sports in general. Three independent investigations employing diverse sampling te chniques and populations revealed that the SFQ is normally distributed, internally consistent, possesses concurrent and predictive validity, and strong test retest reliability. Additionally, males reported consistently higher SFQ scores than females, howev er no relationship was evident between SFQ scores and age. The SSIS and SFQ have provided researchers with the ability to classify sports enthusiasts and as a result have further enabled researchers to more precisely explore the behavioral and psychologica l correlates associated with varying degrees of fandom and identification. Psychological Correlates of Sport Fandom Identification with a sports team has repeatedly been shown to be a key factor in predicting sport related behavior, including event atten dance and the purchasing of team licensed merchandise (Kwon & Armstrong, 2002; Laverie & Arnett, 2000; Wakefield & Wann, 2006). A survey of 1,442 self identified die hard (i.e., watch most games and follow up on scores and sports news several times a da y) and avid (i.e., watch many games and follow up on scores and sports news once a day) sports fans revealed that 70.2% of respondents reported rearranging their professional (i.e., work and school) schedules to listen, watch, or attend a sporting even t, and 39.2% had skipped work/school to accommodate a sporting event (Wann, Friedman, McHale, & Jaffe, 2003). The emotional and psychological impact of sports fans affiliation is often manifested behaviorally. That is, both the covert thoughts and overt a ctions of fans can be the result of their appraisal of their connectedness to and affiliation with a sport, team, and/or athlete.
28 BIRGing, CORFing, and blasting Peoples desire to be positively evaluated by others often moderates their behavior. Cialdini, Borden, Thorn, Walker, Freeman, and Sloan (1976) provided the first empirical evidence to support the notion that people use their affiliation with successful others to increase their public image. In the first of this seminal series of studies, Cialdini e t al. (1976) observed the apparel choices of students at seven major football universities each Monday of the season, with the hypothesis that more students would display the schools insignia following a victory than after a loss or a tie. Data confirmed the anticipated trend that students would wear university apparel to show their affiliation with a victorious team. In their second investigation, Cialdini et al. (1976) explored students terminology when referring to a recent university football game Results supported the hypothesis that participants would emphasize the association between themselves and a successful team through the use of the pronoun we to describe a recent victory (i.e., we won) and non -we terminology to distance themselves from a non-victorious team (i.e., they lost). This effect was magnified after the students received personal failure feedback regarding their performance on a questionnaire; students tendency to associate themselves (i.e., use the pronoun we) with a successful team was exaggerated when their self image was endangered. In their final experiment, Cialdini et al. (1976) again found that students used self referenced terminology when describing a victory twice as frequently than when discussing a non -vic tory. This tendency was amplified when participants first described a game when the universitys football team was not victorious (i.e., lost or tied) and then spoke about a game when the team won. Cialdini et al. proposed that initially talking about a no n -victory led participants to feel that their perceived prestige was reduced and therefore were more likely to
29 attempt to restore their image by increasing their public affiliation with the team when it was successful. In sum, Cialdini et al. (1976) found empirical evidence to support the notion that individuals will publicly emphasize their connection with successful others and bask in reflected glory (BIRG), despite not producing the success themselves. Grounded in Heiders (1958) balance theory, Ciald ini et al. (1976) proposed that BIRGing occurs because people want to maintain cognitive harmony and therefore evaluate connected parties similarly. As a result, public association with a successful team is perceived to promote positive evaluations (i.e. enhanced public image). Moreover, people have an increased tendency to BIRG when their esteem is threatened (either by personal failure or affiliation with a negative source), as a means of maintaining or increasing ones public image. However, despite a cknowledging the existence of this distinct behavioral tendency, these initial investigations did not monitor affiliation tendencies during periods when the teams did not compete. According to Cialdini et al. (1976), theoretically inherent in the concept of BIRGing is the act of distancing oneself from unsuccessful others. Snyder, Higgins, and Stucky (1983) believed that this tactic was distinct, and labeled the protective action of decreasing the association between oneself and unsuccessful others as cutting off reflected failure (CORF). Snyder, Lassegard, and Ford (1986) assessed both BIRGing and CORFing tendencies of participants who received success, failure, and no performance feedback concerning their groups performance on a problem -solving task. In addition to completing the task, participants were informed that their group would be making a presentation to a panel of judges and were asked to what degree they would prefer not to participate in the presentation (i.e., publicly distance themselves from the group), as well as to what degree they wanted to be a part of the groups public presentation (i.e.,
30 publicly associate themselves with their group). Finally, participants were given the opportunity to take and wear badges that displayed their groups name. CORFing was evident among members of the groups that received failure feedback, as these individuals reported a greater preference for not being included in the group presentation and were less likely than the members of the other groups to take and wear team badges, thereby avoiding additional damage to their self image. In comparison, those who receive success feedback had more interest in being present during the groups presentation than those who received failure feedback. Taken together, this e vidence confirms that people use social distancing strategies (i.e., CORFing and BIRGing) to manipulate (i.e., maintain or increase) their image. Cialdini and Richardson (1980) also explored another image management strategy commonly used when peoples public esteem is damaged. Following an experimental manipulation of public image through failure feedback, participants engaged in one of two image -enhancement strategies: basking or blasting Whereas basking serves to increase ones association with a s uccessful source, blasting publicly devalues those with whom an individual is negatively associated (i.e., opposition). Specifically, participants reported more favorable evaluations of their university and less favorable evaluations of a rival university in an attempt to enhance their public image. Cialdini and Richardson (1980) explained that participants engaged in basking and/or blasting behaviors in an attempt to influence public appraisal of themselves. Because people have the propensity to evaluate positively connected groups similarly and negatively connected groups differently (Heider, 1958), people will emphasize the positive characteristics those which they are amicably associated with, while highlighting the negative attributes of those individuals/groups with which they are negatively affiliated.
31 As highlighted above, highly identified fans affiliation with a team is a significant element of their self -concept. As a result, these individuals have a greater tendency to BIRG and to share in the triumphs experienced by their team, thereby instilling a sense of pride (Dalakas, Madrigal, & Anderson, 2004) and serving an ego -enhancing function. Alternatively, these highly identified fans are unable and/or unwilling to distance themselves from an uns uccessful team (i.e., CORF) (Wann & Branscombe, 1990), yet they have the natural desire to maintain a positive self image. Therefore, to cope with the negative vicarious experience and preserve their public image and self -esteem, research has shown that highly identified fans will engage in blasting (e.g., Branscombe & Wann, 1992; Wann, 1993). Wann (1993) explained that blasting can extend beyond the verbal degradation of relevant out -groups (e.g., opposing teams fans) and include physically aggressiv e behaviors on the part of highly identified fans. As a whole, the success of others with whom one is affiliated has been linked to a variety of public behaviors intended to maintain and/or increase ones identity. With an increased understanding of the strategies people employ to protect their self image and the growing prevalence of aggression among sport spectators, Dimmock and Grove (2005) sought to explore spectator aggression among fans of differing levels of identification. Level of identification (using the SSIS), attitudes towards spectator aggression (both physical and verbal), beliefs about subjective norms of spectator aggression, and perceived behavioral control at sporting competitions were collected from 181 fans of professional soccer, prof essional basketball, and Australian Rules football teams. It was hypothesized that fans with varying levels of identification would differ in their attitudes toward and belief about spectator aggression, as well as on their perceived subjective norms of spectator aggression.
32 The data from Dimmock and Groves (2005) paper did not support the primary hypothesis. Notably, however, perceived behavioral control varied as a function of the fans level of identification. That is, highly identified fans reported that they felt as if they had less control over their behaviors at competitions than both the moderately and low identified fans. Dimmock and Grove (2005) presented two potential explanations for why individuals of varying levels of identification differed i n their perceptions of behavioral control. First, the perceived loss of behavioral control may be attributable to the strong emotions that highly identified fans have been shown to experience as a result of a exposure to sport (e.g., Wann, Dolan, McGeorge, & Allison, 1994). Specifically, they suggested, fans who feel that their identity is threatened (or enhanced) are likely to experience emotions that could lead to a loss of behavioral control (p.43). In the alternative explanation, Dimmock and Grove proposed that highly identified fans are more likely to deindividuate when in the presence of other fans (i.e., other in-group members), resulting in a reduced level of self awareness and social restraint. Although Dimmock and Groves study did not find ev idence to suggest that highly identified fans are more likely to support aggressive behavior, it is important to note that the data was collected in a neutral environment (i.e., not following or during a competition), and therefore may not have effectively assessed feelings or behaviors commonly related to competition environments and game outcome. According to Cialdini and colleagues (1976) original work, both BIRGing and CORFing related behaviors are public acts. Boen, Vanbeselaere, and Feys (2002), how ever, hypothesized that BIRGing and CORFing behaviors are not limited to public displays and may also occur in private as a way for individuals to draw personal connections between themselves and successful/positive others. To test if individuals have a tendency to BIRG and CORF in the
33 absence of an audience, Boen et al., (2002) monitored the number of visitors to Belgian and Dutch soccer teams websites the first work day following a match during the final weeks of the 19992000 championship. It was hyp othesized that because of peoples drive to BIRG and CORF in private, more individuals would visit a teams website following a victory than after a defeat. Individuals expectations were also predicted to moderate this relationship, with more individuals visiting a teams site following an unexpected (vs. expected) victory. Significantly fewer individuals were anticipated to visit the site after a surprising defeat. Finally, it was hypothesized that the margin of victory/defeat would also influence private BIRGing and CORFing. Specifically, a positive correlation was expected between the number of visitors to a website and the margin of victory, whereas a negative correlation was expected between the number of visitors and the margin of defeat. In accord with the primary hypothesis, more individuals visited a teams website following a victory while greater than average activity (i.e., visits to the website) was evident after a win. Less than average activity was noted after a loss. No evidence, however, w as found to support the secondary hypotheses that outcome expectations and margin of victory /defeat would influence the number of visitors to the websites. Boen et al. (2002) successfully extended the seminal work of Cialdini et al. (1976) to reveal that not only do individuals strive to increase associations between successful others and distance themselves from less successful others when in public, but they also engage in similar patterns of behavior when in private. According to Boen et al. (2002), pr ivate BIRGing and CORFing are additional techniques for maintaining a positive self -concept. The tendency for fans of all identification levels to engage in various tactics to maintain and increase their
34 perceived esteem is one example of the psychological impact that fandom can have on an individual. Identification and Perception The self -protective behaviors of fans not only underscore the importance of their affiliation, but also highlight the potential implications of identification. Recognizing that the involvement and attachment of fans is real, strong, and that widespread and the goals, achievements and failures of sports teams are taken very seriously by the fans, perhaps in many cases, as seriously as they are by the sports participants themselves (Sloan, 1979, p.221), researchers have sought to capture how team identification and sport fandom are associated with various psychological variables. For example, Schweitzer, Zillmann, Weaver, and Luttrell (1992) probed whether game outcome and the r esulting emotional states altered judgments regarding the probability of threatening events. Students from two rival universities watched the annual football game between the two schools and completed a series of questionnaires relevant to their experience The surveys assessed their enjoyment of the game, support for each of the teams competing, their mood at halftime and their mood end of the game. Additionally, participants completed the World Events Inventory, which asked participants to estimate the likelihood of a war occurring between Iraq and the United States, the duration of such a war, and the number of causalities that could be expected. As predicted, Schweitzer and colleagues (1992) found that although the two groups of supporters did not dif fer in their ratings of enjoyment at halftime (i.e., when there was only a difference of three points between the teams), there was an increase in enjoyment among the winning teams fans and a decrease in enjoyment ratings among the losing teams fans at t he end of the game. Similarly, at the conclusion of the game, greater disappointment was reported by
35 losing teams fans, whereas greater happiness was reported by winning teams fans. Additionally, following their teams loss, students reported greater es timates of war (i.e., more casualties and longer duration) compared to those reported by the fans of the winning team. Moreover, these post -game estimates of war were greater than the estimates provided by the control participants from the same school five days following the competition. No time of assessment differences were evident for students from the victorious school; participants did not report lower estimates of war after their teams victory when compared to the control assessment. In sum, the glo om (i.e., negative affect) rather then exultation (i.e., positive affect) appears to play a larger role in influencing perceptions and appraisal. Highly identified fans are more likely to attend their teams home and away competitions than non -identif ied fans (Wann & Branscombe, 1993). As a result, Wann, Dolan, McGeorge, and Allison (1994) proposed that highly identified fans associate team/player performance with cheering and other forms of fan support and consequently feel as though they are able to influence the outcome of competitions. Therefore, Wann et al. (1994) hypothesized that a positive relationship would be evident between identification level and perceived ability to influence sporting competitions. They tested the proposed relationship b etween identification and perceived influence using 104 undergraduate students who identified their favorite college basketball, college football, professional basketball, professional football, and professional baseball teams and completed the completed t he SSIS for each team. Participants also identified the number of games they attended for each team, the degree to which they feel that spectators influence the outcome of competitions (i.e., general influence), how much they personally feel they are able influence a competition (i.e., personal influence), and how frequently they exhibit specific behaviors (e.g., yelling) at games. Results revealed varying levels of identification by
36 sport, with individuals reporting greatest identification with college an d professional football teams, followed by college basketball and professional baseball teams. Additionally, differences in identification were noted between males and females, with males reporting greater identification for all sports except college football. There were no other gender based differences on measures of interest (e.g., perceptions of influence). Wann et al.s (1994) data revealed that perceptions of influence (general and personal) varied by sport, with participants reporting that spectator s in general had the largest amount of influence on college basketball and football competitions, and that they personally were most likely try to impact the outcome of college football games. Level of identification was also found to be associated with pe rceptions of control, with more highly identified fans reporting greater perceptions of influence across sports than their less identified counterparts. Fans were only more likely to attend sporting competitions if they felt that they personally were able to influence the competition. Despite this finding, identification was more strongly related to perceptions of influence than was attendance (p. 352). Importantly, perceptions of general influence were higher across all sports than perceptions of person al influence; the highest rating of personal perception of influence was less than the lowest rating of perception of general influence. Because more highly identified fans are more invested and involved in the game (e.g., attempt to personally influence t he outcome), Wann et al. (1994) hypothesized that such fans would be more affected (both positively and negatively) by the outcome. To investigate this question, a second study was conducted. College students completed the SSIS and a mood assessment scale prior to, during halftime, and following, one of three NCAA mens basketball games with varying outcomes (i.e., home team easy win, home team difficult win, and home team loss). Participants were separated into high and low identified groups based on SSI S scores.
37 Data showed that highly identified fans reported the largest increase in positive mood following the difficult win and the largest decrease in positive mood following the loss. However, the high identification group only experienced changes in positive mood that were different from the low identification group after viewing a loss. Additionally, the highly identified fans exhibited both the largest increase in negative mood following a loss, and the largest decrease in negative mood after a diffi cult win, but high and low identified groups only experienced different emotional reactions after a loss. In sum and as expected, fans low in identification experienced less intense negative emotions following a teams loss than highly identified fans, sug gesting that mood states evoked in response to a competition are able to successfully discriminate between high and low identified fans. In a final study, Wann and colleagues (1994) addressed the influence of competition outcome on reported levels of iden tification. Social identity theory (Tajfel & Turner, 1979) suggests that the outcome of competitions should not modulate identification levels. Wann et al. (1994), therefore, hypothesized that self report levels of identification would not fluctuate follow ing a victory or defeat. It was however, hypothesized that fans of traditionally more successful teams would report higher levels of identification. Following four collegiate basketball games, spectators completed the SSIS and indicate which team they su pported in the game they attended. Results confirmed both hypotheses; game outcome was not related to identification level, however fans of more traditionally successful teams reported higher identification levels than fans of less successful teams. In co nclusion, the sequence of studies conducted by Wann and colleagues (1994) revealed that team identification is related to fans perceptions of influence over competition outcome, as well as their affective reactions to competition outcomes.
38 Perceptions of Self The personal investment and emotional experiences of fans have been associated with changes in the way that these individuals perceive their surroundings. Furthermore, Schwarz, Strack, Kommer, and Wagner (1987) proposed that identification also has a n impact on more self relevant perceptions. Specifically, Schwarz et al. (1987) sought to explore the effects of mood on judgments of satisfaction with life using the outcome of two Soccer World Championship Games (a victory and a tie for the German team) to induce various mood states. A total of 55 German males were interviewed before and after two Soccer World Championship games, one that Germany won and the other Germany tied. Participants reported their general life satisfaction (i.e., global well -bei ng), their satisfaction with their employment and income, and their satisfaction with the level of national prestige. Differences were noted between before and after competition assessments of global well -being as a function of competition outcome. Partici pants reported greater general life satisfaction after the German soccer team won than before the competition, and less favorable evaluations following the teams tie. The temporary fluctuations in mood resulting from the outcome of the soccer match, however, did not alter the participants evaluations of specific elements of their life (i.e., employment and income). In sum, results illustrated that not only can the outcome of a sporting competition alter ones mood, but also that the resulting mood can inf luence ones general perceptions about his/her well -being and satisfaction with life. It has been proposed that as a result of fans high level of identification, their teams successes and failures are interpreted personally. Hirt, Zillmann, Erickson, and Kennedy (1992) tested this assumption by having fans watch a live televised basketball game that featured their schools team or two other universities. Prior to the game, participants completed a brief questionnaire that assessed their interest in and fa nship toward the their universitys basketball
39 team. Following the competition, participants assessed their current mood state, rated the performance of each team and selected players, and made predictions about the teams success next season. Under the gu ise of a second experiment, participants were asked to estimate how well they could perform a variety of tasks, including motor skill, mental skill, social skill, and chance tasks and then complete each task. Results indicated that the games were successfu l in inducing different mood states; participants who watched their team win reported more positive mood states than those in both the loss and control conditions. Additionally, game condition influenced participants expectations for future success of th e team; participants in the win and control conditions predicted better performance by the team next season than the participants in the loss condition. With respect to estimates of their own ability, game condition influenced ratings of mental skill ability. Specifically, in accord with their estimations of team performance, individuals in the win and control conditions estimated that they could solve more anagrams than those in the loss condition. Differences were also noted between those individuals in t he win and loss conditions with respect to their estimations for the social skill task. No differences however were found in participants estimates on either the motor skill or chance task. Although, differences in performance estimates were evident acros s conditions, actual performance did not differ between conditions. Taking into account level of identification (i.e., fanship), Hirt et al. (1992) found differences in mood and estimates of team and personal performance between the win and loss game condi tions only among individuals with high levels of identification. Highly identified fans are not willing/able to distance (i.e., CORF) themselves from an unsuccessful team because being a fan is an important element of their self -concept, and therefore hol ding an allegiance to a
40 team becomes a gamble, with gains accrued from team victories and costs incurred from team defeats (Hirt et al., 1992, p.730). Hirt and colleagues (1992) selected male college students who scored high on the fanship survey to participate in a follow up study that directly addressed the relationship between personal and team performance and subsequent affective reactions. Data indicated that individuals in both the team and personal success conditions experienced similar levels of low negative affect, yet team failure produced greater negative affect than did personal failure. Similarly, individuals in both the team and personal success conditions experienced comparably high levels of positive affect, however individuals in the pe rsonal failure condition reported more positive affect than did the individuals in the team failure condition. Collectively, team failure had a larger impact on participants mood than personal failure. Hirt et al., (1992) did not find differences self -est eem between personal and team outcome conditions when examined independently. However, differences emerged when the data was collapsed across personal and team conditions; participants in the positive outcome and control conditions reported higher self -es teem levels than participants who experienced negative outcomes. Because the participants in the failure conditions consistently reported mood and self esteem scores that were different from the control group, and only one difference was noted between the control and success groups, the authors suggested that the negative effects suffered by failure appear to be greater than the benefits accrued by success (p.733). Outcome also influenced participants estimates of team performance in Hirt et al.s (1992) study. Participants in the positive outcome and control conditions estimated that the target team would perform better than the individuals who experienced negative outcomes. Differences were also noted in participants estimates of their performance on the motor task and social
41 skills. Additionally, participants in the positive outcome conditions predicted better performance compared with individuals in both the control group and negative outcome conditions. Moreover, no differences were noted between gr oups in their estimates of performance on a chance task. Although the groups varied in their perceived abilities, task performance did not vary between groups. In fact, most individuals, including those individuals who were in the negative outcome conditio ns, overestimated how they would perform on the various tasks. Therefore, it appears that experiencing failure reduces ones level of optimism. In sum, Hirt and colleagues (1992) provided empirical evidence to support the contention that the performance of sports teams can prompt cognitions and affective reactions, which resemble the reactions of individuals to personal performance; fans perceive themselves to be less competent after their team losses. According to Hirt et al. (1992, p.725) fanship, lik e any other important social identityconstitutes an affiliation in which a great deal of emotional significance and value are derived from group membership. Furthermore, threats to identities of such importance and prominence are treated as a threat to self (p.725). In essence, team successes and failures are as significant as personal successes and failures in impacting a highly identified fans identity. The extension of social identity theory to understanding sport fans is a natural extension, as many fans base their identity and sense of self in their affiliation with sports teams. While individuals have the desire to maintain a positive social identity, the identity of sports fans becomes threatened whenever their team loses (Dietz -Uhler & Mu rrell, 1999). To explore the reactions of sport fans to game outcomes, Dietz -Uhler and Murrell (1999) assessed whether participants who were highly identified with their university would evaluate their team more favorably (i.e., make group based reactions ) when evaluating the outcome of competitions
42 more so than participants who did not identify strongly with their university. In this longitudinal study, the authors were also interested in how contextual factors (i.e., game outcome, expectations of succes s/failure, and media attention) influenced both the perceptions and reactions of sport fans. To assess fluctuations in appraisals over the course of one football season Dietz -Uhler and Murrel (1999) administered weekly questionnaires to 74 university stude nts that assessed reactions to the previous weekends football game, and their perceptions of the football team (i.e., goodness, successfulness, intelligence, and skillfulness). Additionally, prior to the first game of the season, participants completed the Collective Self -Esteem Scale (Luhtanen & Crocker, 1992), to measure their identification with their university. Scores on the Collective Self Esteem Scale were used to dichotomize participants into distinct groups with either strong or weak university identification. Results revealed that strongly identified participants evaluated the team more favorably as the season progressed and weakly identified participants maintained similar team evaluations over the course of the season. In addition, the unive rsity football team was evaluated more favorably by participants who identified more strongly with the university than those who were weakly identified; strongly identified fans aim to maintain positive social identities, as indicated by their tendency to view the university football team in a more positive light than weakly identified participants, regardless of game outcome. Data also indicated that strongly identified participants rated the team more favorably than weakly identified participants, especi ally when the team was favored to win and when the team received positive media attention. In sum, the results of Dietz -Uhler and Murrells (1999) study indicate that highly identified fans reactions to competition are influenced by a number of factors, i ncluding competition outcome, media attention, and expectations of success. Strongly identified fans
43 evaluated the team more favorably after victories (compared to losses), when the outcome was expected (versus unexpected) and after the team received posit ive media attention. Emotional Experience of Fans Wann and colleagues (2003) study of sport fanatics not only captured the behavioral influence of high identification; it also provided anecdotal insight into the emotions experienced by fans. Participants reported being significantly impacted emotionally by a teams performance, with 44% requiring at least one day to emotionally recover from a loss and 64.3% noting that they had cried after a defeat. Likewise, 45.4% of participants required at least one day to come down from the excitement and 68% reported crying after a victory. These findings highlight the notion that fans readily recognize that they are emotionally captivated by their teams performance, win or lose. In an experimental design, Wann and Branscombe (1992) explored if reading about a teams victory or defeat was able to trigger an emotional reaction among fans. High and low identified fans read one version of a newspaper article that summarized either a victory by or defeat of the targ et basketball team. The article was either written by a loyal supporter of their team (ingroup), a loyal supporter of the opposing team (outgroup), and a bandwagon supporter of target team, or a supporter of the opposition. Data confirmed the hypothes is that the most positive emotions would be elicited among highly identified fans after reading an article that summarized a victory by the target team, which was written by a loyal fan of the team. Moreover, highly identified participants reported the mos t negative emotions after reading an article written by a bandwagon fan of the target team that highlighted the teams defeat. As predicted, there were no differences in the emotions of low identified participants as a function game outcome, the authors affiliation, or the authors team loyalty. Wann and Branscombe
44 (1992) concluded that, level of identification is critical in determining spectators emotional reactions to information concerning their team (p.58). For highly identified fans, identificat ion with a sports team is a primary component of their identity (Wann, Schrader, & Adamson, 1998). Moreover, research has indicated that individuals who are highly identified with specific sports teams experience emotional reactions that are similar to athletes, yet different from less identified spectators (e.g., Gantz & Wenner, 1995; Sloan, 1989; Wann, 1993; Wann, Dolan, McGeorge, & Allison, 1994). Two explanations have been put forth to account for the intense reactions experienced by highly identified fans. First, people are generally motivated to maintain and protect their self -identity (Tajfel, 1981; Tajfel & Turner, 1979). How ones team performs therefore, influences ones perceptions about his/her own self -worth. Such a link between ones identity and team success does not exist for those who do not avidly support or identify with a sports team, as their self concept is not tied to their affiliation with an athletic team (Wann et al., 1998). Second, being a fan is central to highly identified fans sense of self and therefore they are unable and unwilling to distance themselves from their team when the team performs poorly, thereby allowing their self -identity to be threatened (i.e., the self -concept of highly identified fans is reduced when their team is unsuccessful). In contrast, fans with low levels of identification do not have such personally relevant and strong ties to a team, and therefore are able to protect themselves (i.e., their self concept) by distancing themselves from the target t eam. As a result, team performance moderates the self -identity of highly identified fans more so than traditional fans (Wann et al., 1998). Recognizing the significance of competition outcome to fans, Wann et al. (1998) hypothesized that fans would exp erience cognitive and somatic anxiety related to their teams
45 competitions, but highly identified fans would experience more intense competition related anxiety than less identified fans. Additionally, it was hypothesized that the level of anxiety reported would be influenced by the temporal proximity of the competition, with higher levels of anxiety anticipated closer to the competition. The importance of the competition was also proposed to influence fluctuations in anxiety, with anxious responses only being evoked when the target team faced a skilled opponent. Thirty one college aged participants completed the SSIS and a modified Competitive State Anxiety Inverntory 2 (Martens et al., 1990; Competitive State Anxiety Inventory-Fans, CSAI -F) three days, tw elve hours, three hours, and immediately before two target competitions (one important and difficult game and one less important and less difficult game), as well as at halftime during each of the competitions. Using a median -spilt, participants were class ified as possessing high ( M = 6.03) or low ( M = 2.66) levels of team identification. Wann and colleagues (1998) found that across time intervals for both games, all participants reported higher levels of cognitive anxiety than somatic anxiety. As the im portant/difficult competition approached, cognitive anxiety scores increased and ultimately peaked at halftime. Additionally, highly identified fans reported higher levels of cognitive anxiety than low identified fans. A similar pattern of results was fou nd for somatic anxiety; highly identified fans experienced a more striking increase in somatic anxiety compared with less identified fans, although both groups experienced increases in somatic anxiety that peaked at halftime during the more challenging com petition. Cognitive anxiety did not vary across assessment periods for the less important/difficult game, but counter to expectations, somatic anxiety increased as the less difficult/important competition approached and was highest at halftime during the c ompetition. Moreover, highly identified fans experienced a unique and
46 marked increase in somatic anxiety during halftime that was not present among the low identified fans or during the previous four assessment periods. Wann et al. (1998) explained that re gardless of the significance of the competition, highly identified fans become aroused whenever their team is engaged in competition. To further extend the parallel between the emotional experiences of athletes and sport fans, Wann et al. (1998) in a second study assessed if sport spectators, like athletes, were able to recall their competition related anxiety states one week after the competition. Participants completed the CSAI -F immediately prior to a regular season collegiate basketball game between tw o conference rivals, as well as during halftime of the game and were asked to complete the CSAI -F recalling the feelings that they experienced during halftime of the game one week later. SSIS scores were used to classify participants as high or low identif ied fans. Replicating the findings of study one, participants reported greater cognitive anxiety than somatic anxiety and highly identified fans reported both more cognitive and somatic anxiety than their less identified counterparts. Wann and colleague s (1998) also reported that all participants were able to accurately recall the magnitude of the anxiety (both cognitive and somatic) they experienced during halftime of the target game one week earlier. Further exploration of the data, however, indicated that the highly identified fans were not able to recall their anxiety levels with great accuracy, and the low identified fans were in fact able to recall their anxiety levels with more accuracy than those high in identification. Wann et al. (1998) specula ted that the game time anxiety experienced by highly identified fans impeded their ability to encode information about their emotional state or that their lack of recall was due to state dependent learning. In other words, during the game highly identifie d participants experienced elevated anxiety levels, but during the
47 follow up assessment however, they were in a nonanxious state, thereby creating a discrepancy in the emotional states when the CSAI F was completed reducing recall ability. In sum, the Wan n et al. (1998) studies provide evidence to suggest that fans experience anxiety responses that mirror those of athletes, showing an increase in both cognitive and somatic anxiety levels as important competitions approach. Kerr, Wilson, Nakamura, and Sudo (2005) sought to address the emotional experiences of soccer fans as a function of competition outcome. In the study, Japanese soccer fans completed a modified state version of the Tension and Effort Stress Inventory (TESI; Svebak, Ursin, Endresen, Hjelmen & Apter 1991) prior to, at halftime, or following one of two selected professional J -League soccer games. Results indicated that fans of the winning team experienced less anger than supporters of losing team, and that the losing teams fans reported in creases in anger from the pre -game assessment to the halftime and post -game assessments. A similar pattern also emerged for the other negative emotions of humiliation, sullenness, and resentment. The data however was unable to support the hypothesis that fans of the winning team would report an increase in the positive emotions of relaxation and excitement at halftime and following the game. Despite this, the increase in negative emotions among soccer fans in Kerr and colleagues 2005 study parallel the e motional reactions experienced by fans of a losing basketball team revealed in Wann, Dolan, McGeorge, and Allisons (1994) investigation. Physiological Fluctuations and Fan Identification As reviewed above, the affective states and behaviors of fans are of ten related to their level of identification with a sport team. Identification has also been shown to manifest itself in terms of physiological changes. For example, evidence has shown that fans experience increased adrenal gland stimulation (Lawther, 1951 ) and heart rate (Corbin, 1973) when watching a
48 competition, reactions which mirror changes experienced by the athletes themselves (Sloan, 1979). Branscombe and Wann (1992) proposed that both arousal and aggressive behavior can be mediated by identificati on with an individual/group, and therefore sought to provide an empirical link between physiological arousal, social identity, and aggressive behavior. Because of increased personal investment in a team, highly identified fans have a greater tendency to be come vicariously involved in competitions and are more likely to view themselves as involved in the competition, prompting them to experience emotions similar to the participants. Similarly, highly identified fans experience increased threats to their self -esteem and public image when their team performs poorly. Therefore, Branscombe and Wann (1992) hypothesized that perceptions of arousal and pleasantness would be moderated by participants level of identification with a target and the targets performanc e. In Branscombe and Wanns (1992) study, 41 college aged females that had little interest in the sport of boxing, but variable levels of American identification (i.e., frequency of feeling proud to be an American) viewed both a neutral video depicting the construction of zoos and a six -minute video clip of either an American boxer defeating a Russian boxer or a similar clip in which the Russian boxer defeated the American boxer. Blood pressure was measured prior to and following both the viewing of the neutral video and the experimental video as an index of arousal. Additionally, after watching the experimental video clip, participants completed a questionnaire that assessed their feelings about the fight, the competitors, and the outcome of the fight, a s well as Russians and other out -groups (i.e., non -Americans). Data indicated that high and low identified Americans did not differ in their baseline blood pressure (i.e., pre and post viewing the neutral film), however participants who were highly ident ified experienced an
49 increase in both systolic and diastolic blood pressure after viewing the experimental film. Additionally, highly identified fans reported more enjoyment watching the American boxer win than lose; no differences in enjoyment were noted between the conditions for participants low in identification. Moreover, regardless of the video clip watched, the highly identified participants appraised the Russian boxer more negatively than the low identified participants. Branscombe and Wann (1992) c oncluded that identification with a group that a sports figure represents influences physiological processes during exposure to a competition involving that group (p.91). An overwhelming majority of the sport fandom research has focused on the psychologi cal correlates of team identification/fandom. However, Bernhardt, Dabbs, Fielden, and Luttet (1998) sought to identify physiological changes associated with the viewing of a favorite team succeed or fail. Specifically, Berhardt and colleagues (1998) assess ed changes in pre and post competition testosterone levels among sport fans. Testosterone level was selected as an index of physiological modulation because previous research has shown increases in testosterone levels among males following successful perfo rmance on both physical (Booth, Shelley, Mazur, Tharp, & Kittok, 1989; Elias, 1981) and cognitive tasks (McCaul, Gladue, & Joppa, 1992; Mazur, Booth, & Dabbs, 1992). Because the authors proposed that fans who identify with a team will feel its success or failures as their own, giving rise in them to vicarious experiences of success of failure (p.59), it was hypothesized that the testosterone levels of the fans whose team had won would increase more relative to pre -competition levels than those fans whose team lost. Data from Bernhardt and colleagues (1998) two studies (attendants of a college basketball game between two rival schools and fans watching a World Cup soccer match between two rivals on television in two separate taverns) reflected an increase in post -competition testosterone
50 levels among the fans of the victorious teams and a decrease in testosterone among the fans of the losing team. Bernhardt et al. speculated that the alterations in testosterone levels may have widespread implications, possi bly influencing fan related emotions (e.g., enthusiasm, dejection) and behaviors (e.g., violence, aggression), as well as other more general thoughts and behaviors (e.g., group interaction, mood). Because of the high level of investment of sport spectators and their often strong emotional reactions, Hillman, Cuthbert, Cauraugh, Schupp, Bradley, and Lang (2000) adapted an affective picture viewing paradigm and incorporated physiological indices to better understand the biological mechanisms underlying emotio nal responses of identified fans. The authors approached the issue from a biphasic perspective (Lang, 1985), which views emotions as action dispositions that facilitate either approach or withdrawal related behaviors. According to Lang, people are infor mation processors who direct their attention (i.e., selectively attend) towards motivationally salient information (compared to affectively neutral stimuli) and respond along the two dimensions of emotional valence (i.e., pleasant unpleasant) and arousal ( i.e., intensity of response). Fluctuations in psychophysiology (e.g., electromyographic activity, cortical event related potentials, eyeblink magnitude, and heart rate) are reliably coupled with changes in affective valence and arousal ratings of stimuli. Recognizing this association, Hillman and colleagues (2000) integrated psychophysiological indices and self -report measures of affect within a picture -viewing paradigm with sport fans as their population of interest. Hillman et al. (2000) used the SSIS to categorize 40 university students into different levels of identification: low, moderate, and high. Participants viewed two slide sets: one that depicted the target universitys team in competition and either one that included pictures of other university /professional teams or one that included images of various individual sports (e.g.,
51 downhill skiing). While viewing the slides, heart rate, subjective ratings of pleasantness and arousal, eyeblink responses to startle probes, and event related potentials ( ERPs) to both image onset and startle probes were recorded. Hillman et al. (2000) hypothesized that the images depicting the universitys team would elicit emotional engagement among highly identified fans. Moreover, the highly identified participants we re hypothesized to rate team -relevant pictures as more pleasant and arousing, show more positive cortical slow waves at stimulus onset, produce smaller P3 responses and eyeblinks with less magnitude in response to startle probes, and show greater heart rat e deceleration when viewing team relevant images compared to control images (i.e., images of other teams or individual sports). Hillman et al. (2000) found that highly identified fans rated team relevant images to be more arousing than team irrelevant ima ges, while low and moderately identified fans did not appraise relevant and irrelevant images to differ in arousal. With respect to valence ratings, highly identified fans rated both team relevant and irrelevant images more pleasurable than lower identifie d fans; a finding that may be attributed to the tendency of highly identified fans to be more captivated by sport in general than low identified participants. Additionally, all participants rated team relevant images to be more pleasurable than team irrele vant images. Hillman and colleagues (2000) also noted variability in participants cardiac responses. Highly identified fans experienced greater heart rate deceleration when viewing team relevant images compared to team irrelevant images. In comparison, a reverse pattern of cardiac response was noted for fans low in identification; low identified fans exhibited greater heart rate deceleration when viewing team irrelevant images. The deceleration found among highly identified fans when viewing team relevant images was expected and is consistent with the interpretation of heart -rate change as indexing attentional engagement with a stimulus (p.27),
52 suggesting that highly identified fans exhibit an increased degree of motivational engagement when viewing images that depict their team. Additionally, smaller P3 amplitudes were noted for both high and moderately identified fans when viewing team relevant images, compared to team irrelevant images. Such findings are in accord with the theory of motivated attentio n, which suggests that individuals are primed to react more to motivationally significant events/stimuli compared with neutral events/stimuli (Lang, Bradley, & Cuthbert, 1998). Based on this data, it is probable that high and moderately identified fans devoted more attention to team relevant material, and therefore had fewer cognitive resources available to delegate to the secondary probe, thereby prompting an attenuated P3 response. However, startle blink reflex did not vary across picture types or lev el of identification. In hindsight, the authors noted that this finding was not particularly surprising since the stimuli (both team relevant and irrelevant) included in the study did not vary considerably in valence ratings and therefore were insufficient in modulating the startle responses of participants. In a follow up study by Hillman, Cuthbert, Bradley, and Lang (2004), fans from two rival teams viewed and rated images from six picture categories (i.e., team relevant pleasant and unpleasant, team irr elevant sport, neutral, mutilation, and erotica) while physiological indices of arousal (i.e., ERP, skin conductance) and valence (i.e., startle eye -blink) were monitored. It was hypothesized that team relevant pleasant images (i.e., their team winning) wo uld evoke smaller startle eye -blink responses and higher ratings of pleasantness than team -relevant unpleasant images (i.e., their team losing), yet no differences in arousal were expected between team relevant pleasant and unpleasant images. However, aro usal was anticipated to vary between both team relevant (i.e., pleasant and unpleasant) image categories and team -irrelevant sports images. No physiological or subjective differences were expected to emerge between the groups of rival
53 fans when presented w ith images from the three control categories (i.e., neutral, mutilation, and erotica). As hypothesized, Hillman and colleagues (2004) found increased affective engagement among highly identified sport fans when they were exposed to images depicting their team compared with general sport stimuli. Specifically, fans rated images of their team winning as more pleasant and images of their team losing as more unpleasant than team irrelevant sport images. Both categories of team -relevant images were perceived t o be more arousing than team irrelevant images. Similarly, both categories of team relevant images elicited smaller startle probe -P3 (ERP) peaks than team irrelevant images, suggesting that fans had fewer cognitive/attentional resources available for proc essing the startle probe when viewing emotional sport pictures (i.e., more attention is allocated towards the motivationally salient team content by highly identified fans). Moreover, reduced startle eye blink responses were found during exposure to team r elevant images, compared to team -irrelevant images, as was increased EEG positivity, a trend that suggests participants were processing the emotional stimuli. Skin conductivity levels also successfully differentiated between the relevant and irrelevant im ages, with greater activity when participants viewed team -relevant images, providing an additional index of increased attentional engagement by highly identified fans. In sum, Hillman and colleagues (2000, 2004) provide strong evidence that highly identifi ed sports fans exhibit motivated emotional reactivity to team relevant stimuli, a finding which provides a preliminary mechanistic link between sport fandom and mental health. Collectively, the extant literature confirms that there is much more to being a fan than simply attending sporting events and rooting for ones team. As a result, researchers have
54 recently directed their interests to exploring the implications of the affective and behavioral engagement of fans on their psychological well -being. Th e Connection Between Team Identification and Psychological Health Earlier in the review, the tactics of BIRGing, CORFing, and blasting were highlighted as means by which fans maintain or enhance their self -perceptions so as to improve their psychological health. Recently, researchers have sought to further explore the suggestion that sports fans exhibit more mentally healthy profiles and to identify the factors that contribute to these enhancements. Wann, Inman, Ensor, Gates, and Cladwell (1999) used the Pr ofile of Mood States (POMS; McNair, Lorr, & Droppleman, 1971) to assess the psychological health of sports fans. They hypothesized that highly identified fans would report higher levels of vigor and self -esteem and lower levels of tension, depression, ange r, fatigue, and confusion, a pattern reflective of a more psychologically healthy profile and commonly found among athletes. This iceberg profile was only expected to emerge among highly identified fans of geographically local sports teams, and not among general sports fans or fans of distant teams. In the first of two studies, 74 college students completed a demographic questionnaire that included questions about their gender, age, general level of sport fandom, a modified version of the POMS, and the SS IS using the universitys basketball team as the target. Participants were classified according to the SSIS data as either high or low identified fans. Wann et al. (1999) found evidence to support the contention that highly identified fans possess more psychologically healthy profiles (i.e., iceberg profiles), and proposed that the psychological benefits result from the social camaraderie and feelings of belonging experienced by highly identified fans of local sports teams. From their affiliation with a local team, many fans develop social connections with likeminded fans. To discern if the more psychologically
55 healthy profiles of highly identified fans can be attributed to increased social interactions with other highly identified fans of the same team, Wann and colleagues (1999) compared the mental health profiles of fans that identified with teams that were geographically close and distant. In their second study, Wann et al. (1999) had college students complete the same battery of questionnaires used in the preliminary project, however the target team for the SSIS was a geographically distant team that was located within the same state as the participants university. No relationship was found between level of identification and POMS subscale scores, s uggesting that identification with a geographically distant team does not afford the same psychological benefits associated with being a fan of a geographically close team. The results of both studies led Wann et al. (1999) to conclude that increased ident ification with a local sports team is related to psychological health, but general sports fandom or identification with a non -local team are not. Wann et al. (1999), however, cautioned that personal situations and connections with other fans may moderate t he relationship between identification, team location, and psychological well -being. The psychological benefits of fandom were also assessed among fans of Australian Rules Football by Wann, Dimmock, and Grove (2003). Psychological health was conceptualize d and assessed along two dimensions, personal and social well being. Personal psychological health was measured using the 14 item Perceived Stress Scale (Cohen, Kamarck, & Mermelstein, 1983) and the 10 item Rosenberg Personal Self -Esteem Scale (Rosenberg, 1979). Social psychological well -being was quantified using the 16-item Collective Self -Esteem Scale (Luhtanen & Crocker, 1991) and the 20item UCLA Loneliness Scale (Russell, Peplau, & Cutrona, 1980). The number of sports teams that participants reporte d closely following determined the participants level of general sport fandom and the SSIS assessed team identification for both the local football team
56 and their favorite football team. Data revealed that not all dimensions of psychological well being ar e related to identification with a local sports team. A positive relationship was evident between level of identification with a local sports team and loneliness (an index of social well being with higher scores representing better psychological health). A similar pattern was not evident for collective self -esteem (social well -being), nor was a relationship found between either index of personal psychological health (i.e., personal self -esteem and perceived stress) and local team identification. However, th e hypothesized associations between social psychological health and the location of the target team were confirmed. Although both the fans of local and geographically distant teams reported similar levels of identification, only those fans that identified with a local team reported higher levels of psychological well -being. Specifically, fans of local teams reported less loneliness and higher levels of collective self -esteem compared to fans of more distant teams. No differences in personal psychological well -being were noted between the two groups. The lack of association between personal elements of psychological health was also evident in Wann, Pierce, Padgett, Evans, Krill, and Romays study (2003.) Specifically, no relationship was found between ide ntification (as indicated by scores on the SSIS) and optimism (assessed with the Revised Life Orientation Test; Scheier, Carver, & Bridges, 1994). Therefore, the increased psychological well -being associated with sport team identification can be attribute d to the social associations that result from identifying with a local team, and are not the direct result of ones identification with a sports team (Wann et al., 2003). To clarify, when an individual supports a local team, he/she is surrounded by fans of the same team, and is expected to experience a sense of camaraderie and social connectedness. These factors are believed to be at the root of the psychological benefits associated with team identification. In contrast, being a
57 fan of a relatively distant team does not afford such psychological benefits, as the proximity to significant, team related social support is also distant. In a subsequent study, Wann, Dunham, Byrd, and Keenan (2004) were able to provide novel connections between psychological healt h and identification with a local sports team. Specifically, positive correlations between team identification and Extroversion, Openness, and Conscientiousness subscales of the revised NEO Personality Inventory were noted. However, the predicted negative association between identification and Neuroticism was not supported. Taken together, the findings from these two studies further emphasize that the mental -health benefits drawn from team affiliation are tied to increased social interactions with likemind ed others ; optimism, as well as the components of Neuroticism (e.g., anxiety, depression, hostility, and anger,) are more closely related to personal well being. In a more recent investigation Wann and Pierce (2005) included the modified Satisfaction with Social Life Scale (Diener, Emmons, Larsen, & Griffin, 1985) as an index of social well being. One hundred fifty -five university students completed a battery of questionnaires that assessed basic demographic information, level of sport fandom (SFQ), level of team identification (for both local and distant teams) (SSIS), and psychological health (i.e., social self esteem [Collective Self -esteem Scale; Luhtanen & Crocker, 1991] and social life satisfaction [modified Satisfaction with Social Life Scale; D iener et al., 1985]). The combined effect of fandom, identification with the local team, and identification with a distant team, was a predictor of social psychological health. However, independently, identification with the local team accounted for the la rgest proportion of variance in psychological health scores, with neither general sport fandom nor identification with a distant team contributing independently. In sum,
58 level of identification with a local team was a predictor of social psychological wellbeing, whereas identification with a distant team and mere sport fandom were not. Taken together, the extant data indicated that: 1) higher levels of identification with a local sport team are associated with positive psychological health, 2) there is n o relationship between identification with a distant sport team and psychological health due to the lack of readily available societal connections, and 3) high levels of general fandom are not related to psychological health, again due to the lack of value d social connections. However, support for these three tenets has been garnered primarily from correlational data, therefore limiting the confirmation of a causal (e.g., directional) relationship between the variables. Recognizing this limitation, Wann ( 2006a), implemented a cross lagged, longitudinal design to document the association between team identification and psychological well -being. The central hypothesis was that a chronological association would be found between team identification at Time 1 and psychological well -being at Time 2. Moreover, no relationship was anticipated between Time 1 psychological well -being and Time 2 team identification. To reiterate, the positive association between team identification and psychological well -being was e xpected because ones identification with a local sport team allows for the individual to become part of a larger social group, which in turn aids in the reduction of feelings of isolation and loneliness. During the first testing interval (i.e., Time 1) of Wanns (2006a) investigation, 214 university students completed a demographic questionnaire, the SSIS, and four measures of various components of psychological health (i.e., personal self -esteem, collective self -esteem, loneliness, and perceived stress). Twenty -one percent of the original sample participated in a second identical assessment three months following the initial session. As predicted, a positive
59 path was evident between team identification at Time 1 and psychological well -being at Time 2, pro viding evidence to suggest a temporal pattern between identification and well -being. Wann (2006a) however cautioned, that although the results of this investigation support the idea that identification with a local team generates numerous psychological benefits, these fans are also subjected to negative consequences such as increased negative emotion (e.g., Bernhardt et al., 1998; Wann et al., 1994), depression (e.g., Schwarz et al., 1987; Schweitzer et al., 1992), and reduced perceptions of personal abilit y (e.g., Hirt et al., 1992) following a poor performance by their team. Fortunately, in general, highly identified fans are able to cope with threats to their identity (e.g., blasting) and continue to reap the social -psychological benefits of their identif ication. In sum, Wanns 2006(a) study is the most recent attempt by researchers to reveal a directional relationship between identification with a sport team and psychological health. As suggested throughout this review, identification with a sports team holds a significant place in the lives of millions of people worldwide, and the emotional impact that sports provides is the most important measure of its size and value (Rascher, 2005, p. 2). Based on the extant literature, is clear that there are numer ous behavioral and psychological correlates of identification with a sports team. Wann (2006b) developed a theoretical model to formally explain the association between identification with a team and social psychological health (see Figure 1). According t o the Team Identification -Social Psychological Health Model, enhanced psychological well -being is the result of team identification, social connectedness, and coping mechanisms. Wann asserts that as an individual develops an affiliation with a sports team and he/she embraces the identity of being a team supporter, social connections are developed with other fans (i.e., a sense of belonging and solidarity). The social connections that are derived from ones identification with a team can be either enduring or temporary. The enduring connections
60 result when the presence of other fans is clearly evident (i.e., other supporters are easy to find/readily available). In contrast, temporary social connections are not permanent and only arise when an individual ente rs an environment in which his/her connection with a team and similar others is made salient. As an identified fan, threats to ones identity are abundant and it is ones ability to cope with such threats (e.g., teams loss/poor performance) which will mod erate the their level of social well -being. Like the social connections resulting from identification, enhancements in social psychological well -being are both long term (trait) and transitory (state) in nature; fans acquire the psychological benefits fro m engagement in social interactions with fellow team supporters. Figure 2 1. The Team Identification -Social Psychological Health Model (adapted from Wann, 2006b). Despite the recent development of a model of psychological health pertaining to sports f ans (Wann, 2006b), numerous questions remain regarding the psychological implications of identification. For example, several negative consequences have been tied to high levels of identification (e.g., aggression [Bernhardt et al., 1998], abandonment of responsibilities [Wann et al., 2003], extreme emotion [Wann et al., 1994]), yet there remains uncertainty regarding factors Antecedents (causes) of sport team identification Sport team identification Enduring social connections resulting from identification Temporary social connections resulting from identification Threat to identity Coping with threat State and trait social psychological health
61 that may moderate the association among the components of the model. Furthermore, there is a lack of conceptual clarity with respect to temporary and enduring social connections and their interactive effects on psychological health (Wann, 2006b). In light of this, Wann (2006b) has urged researchers to employ approaches to study both the positive and negative consequences of sport fand om and identification (p.288) in order to develop a more balanced and complete perspective of the effects of team identification. As noted by Wann (2006b), it would be imprudent to ignore the numerous negative consequences that have been correlated with h igh levels of team identification. It is especially important to further explore the foundation for these negative effects, since a number of the behavioral and emotional experiences of fans parallel those encountered by individuals with various psychopathologies. Individuals with affective disorders and substance abuse problems, for example, exhibit similar macro behaviors (e.g., preoccupation with or avoidance of people, places, and things) that are specific to their given situation. Highly identifie d fans exhibit similar behavioral tendencies such as compulsively seeking the latest team related news (e.g., statistics, scores, trades). This need is often met through ritualistic (e.g., daily or hourly) consumption of sports media, a behavioral trend that has become more popular with the increased prevalence and accessibility of media outlets that cater to the avid sports fan (e.g., numerous 24 hour sports television networks, countless websites, and radio programming) (Wann & Branscombe, 1992). In ot her words, highly identified fans exhibit a distinct need (or at a minimum, a strong perceived need or desire) for team related information, and will often plan their schedules around retrieving the information, as it is a critical component of their dai ly routine. In addition to actively seeking team insight and watching competitions, highly identified fans experience elevated emotional reactions (both positive and negative) to game outcome and sporting news, as
62 their self -concept is directly linked with their target team and its performance (Wann et al., 2001). As such, sport consumption often takes precedence over other commitments and responsibilities. Moreover, addicts continue to use drugs and/or alcohol despite the fact that it detracts from personal responsibilities, while individuals with chronic anxiety and phobias actively avoid potential exposure to threatening stimuli through behavioral and environmental changes (Williams et al., 1997). The explicit behavioral parallels that exist between highl y identified fans and such clinical populations suggest that the two groups may also engage in similar micro -behaviors (e.g., information processing, functional alterations in the allocation of attention). At the most basic level, each of these groups sh are an intense emotional connection and reactivity to a distinct stimulus (e.g., highly identified sports fans: their team; addicts: narcotics, alcohol, caffeine; phobics: spiders, snakes; clinically anxious: social situations). For these individuals, such stimuli possess motivational salience and according to Lang, Bradley, and Cuthbert (1997), it is motivation that drives attention. These micro-behavioral tendencies, therefore, should be reflected in environmental conditions that index differential alloc ation of attention to stimuli that are differentially salient as a function of affective disposition. Attentional Bias Attention is a fundamental component of cognition, is a significant mediator of human behavior, and is associated, in various forms, with mental health. By developing an understanding of the factors, both internal and external, which influence the direction and maintenance of attention, insight is gained into the cognitive mechanisms that drive behavior and potential psychopathology. As reviewed, much research has explored the behavioral and psychological correlates of fandom; however, the cognitive factors that drive the thoughts, emotions, and behaviors of highly identified sports fans have been relatively ignored. Therefore,
63 by adopting a cognitive orientation and exploring variations in attentional allocation, we can gain insight into the micro -behaviors of fans, which can aide in understanding their explicit actions. As Williams and colleagues (1997) noted, people have a tendency to be d rawn to content that is important to them. For example, if one is buying a house, one seems to notice For Sale signs. If one buys a new car, the number of people driving just that type of car suddenly seems to increase. (p.74) In other words, an increa sed sensitivity to elements in the environment that are personally relevant/motivationally salient is reflected in an attentional bias towards these meaningful stimuli. Williams and colleagues outlined a number of assumptions with respect to the nature of attentional bias. First, a distinct shift in the direction of attentional focus occurs and the individual becomes aware of a particular element of his/her surrounding environment. This awareness can occur via any of the senses and is often initiated by an environmental (internal or external) change. Although, this shift is generally regarded as involuntary, it can be voluntarily driven. Moreover, the salience of environmental cues can vary, thereby increasing/decreasing the degree of bias. Finally, the bia s can facilitate and/or inhibit behavior. Parkinson and Rachman (1981) conducted an early experiment that illustrated the facilatory effects of attentional bias. In their study, two groups of mothers (one group who had children scheduled for a tonsillectom y on the day of the study and another that had children, but were not scheduled for a surgical procedure) listened to music. The music was coupled with: medically -related, auditorily confusable neutral, or neutral/dissimilar terms that were presented at increasing volume. The mothers were asked to repeat the words they heard. During all but the loudest condition, mothers whose children were having surgery repeated more medically related (i.e., concern relevant) terms than the control group. Although no di fferences were noted between the groups for the neutral/dissimilar terms, differences did emerge for auditorily
64 confusable neutral terms when presented at the lowest volumes. Despite some methodological criticisms, Parkinson and Rachmans study provides an example of how the presence of motivationally salient material can alter cognitive processes. Numerous protocols have been developed to explore the nature of attentional bias among individuals with emotional disorders, and have been more recently extende d to better understand the cognitive mechanisms associated with substance abuse. For example, the emotional Stroop paradigm and subsequent design modifications (e.g., counting, masked presentation, pictorial) have been used to assess attentional bias by m easuring interference on color naming/ identification task performance when affective stimuli (words or pictures) are used (e.g., Lim & Kim, 2005; Moradi, Taghavi, Heshat Doost, Yule, & Dalgleish, 1999; Williams, Mathews, & MacLeod, 1996). Other researcher s (e.g., hman, Flykt, & Esteves, 2001; hman, Lundqvist, & Esteves, 2001; Rinck, Reinecke, Ellwart, Heuer, & Becker, 2005) have employed visual search tasks, using the time taken to detect a target image among distracters as an index of bias. Additionally MacLeod, Mathews, and Tata (1986) developed one of the leading protocols used by researchers of attentional bias, the dot -probe task. The dot probe task is grounded in the assumption that individuals respond more quickly to stimuli that appear in areas of the display where attention is directed (Posner, Snyder, & Davidson, 1980). In a typical task design, attentional allocation is assessed using briefly presented (i.e., 500ms) pairs of stimuli (neutral or affective), which are immediately followed, on se lected trials, by the appearance of a small dot. Participants respond to the appearance of the dot by rapidly pressing a button. If a bias exists, individuals will potentially respond faster to probes that replace affective rather than neutral stimuli, given that attention should be attracted by the
65 most salient material. In general, the dot -probe task has been able to successfully discriminate between clinical and control populations. Affective Disorders In the seminal dot probe study, MacLeod et al. (1986) sought to explore an attentional mechanism underlying anxious individuals cognitive bias. Clinically anxious and nonanxious controls completed the task, which consisted of 288 word pairs, including 24 words related to physical threat and 24 related to social threat. Although participants responded to neutral stimuli (i.e., dots) with neutral responses (i.e., button press), analysis of response latency indicated that the anxious groups attention shifted toward the threat terms. When probes were present ed in both the upper and lower portions of the display, anxious participants responded faster when they were preceded by a threat word (both social and physical). An opposite pattern was evident among the controls; faster responses time were evident when t he probes followed neutral words. The original protocol was rerun with a group of clinically depressed participants to identify whether the anxious groups bias was moderated by depression. The response latencies of the depressed participants did not revea l the presence of a bias towards threat content. Together these results support the notion that high anxiety leads to a bias in selective attention that favors the pickup of emotionally threatening information (MacLeod et al., 1986, p. 18), suggesting th at the cognitive system may functioning differently in anxious and nonanxious populations and that such a processing bias favoring emotionally threatening stimuli may facilitate the maintenance of affective disorders. The pattern of faster reactions to th e presence of probes presented in the location previously occupied by threat related cues exhibited by anxious individuals revealed in MacLeod and colleagues (1986) seminal investigation has been replicated in numerous other studies. These subsequent expe riments have revealed similar trends among individuals diagnosed with:
66 Generalized Anxiety Disorder (GAD) (e.g., Bradley, Mogg, White, Groom, & de Bono, 1999; Dalgleish, Taghavi, Neshat Doost, Moradi, Canterbury, & Yule, 2003), social anxiety disorder (e.g ., Musa, Lpine, Clark, Mansell, Ehlers, 2003; Pishyar, Harris, & Menzies, 2004; Stirling, Eley, & Clark, 2006), Post Traumatic Stress Disorder (PTSD) (e.g., Bryant & Harvey, 1997; Elsesser, Sartory, & Tackenberg, 2004), and Obsessive Compulsive Disorder ( OCD) (e.g., Muller & Roberts, 2005; Tata, Leibowitz, Prunty, Cameron, & Pickering, 1996). This collective body of evidence suggests that bias to negative information is a key feature of anxiety disorders (Williams et al., 1997; Mogg & Bradley, 1998). However cognitive biases are not exclusive to emotional disorders. Recent research has linked chemical dependence (illicit and legal) with biased attention. Chemical Use and Dependence The application of the dot -probe paradigm was further extended by Lubman, Peters, Mogg, Bradley, and Deakin (2000), who implemented a pictorial dot probe task to explore attentional biases in individuals with opiate dependence. It was hypothesized that individuals addicted to opiates would have faster response times to probes t hat replaced drug relevant stimuli than to those that replaced neutral images. Sixteen methadone-maintained heroin addicts and 16 non -drug using controls viewed 160 pairs of images that included drug -relevant images paired with neutral images and pairs of neutral images. Following the presentation of each image pair, a probe appeared in the location previously occupied by one of the images and the participants were asked to identify the location of the probe (i.e., right or left area on the display) as quic kly as possible. Participants also completed a variety of questionnaires assessing numerous psychological attributes. Lubman et al.s (2000) analysis of response times revealed that only the addicts were faster to respond when the drug image and probe appe ared in the same location. Moreover, a bias
67 score (positive scores reflect faster reactions) calculated based on response time to only probes presented after the relevant -neutral pairs indicated that opiate addicts exhibited faster response times to probe s that followed the drug relevant images compared with the control group. Although an attentional bias toward drug relevant stimuli was evident, no relationship was found between bias and depressive symptomology or anxiety for either the controls or addict s. Lubman and colleagues (2000) suggested that the presence of such a bias can increase the visibility of drug paraphernalia in their environment, prompt concentration problems, increase conscious thoughts about drug use, and potentially aid in the maint enance of the addiction. Ehrman, Robbins, Bromwell, Lankford, Monterosso, and OBrien (2002) conducted a similar investigation. In addition to including current nicotine addicts and controls, they also tested a group of former smokers. A difference in bi as was evident between current and non smokers. As predicted, the current smokers exhibited a greater bias (i.e., enhanced attention) toward smokingrelated stimuli than both former and non -smokers, suggesting that smoking related cues capture the attentio n of smokers. Townshend and Duka (2001) studied the selective attention of heavy nondependent and occasional social drinkers using alcohol related words and images in a dot probe task. It was anticipated that the alcohol related stimuli would be perceive d as motivationally salient for the heavy drinkers and thereby promote reduced response latencies to probes presented after such cues. Data supported the hypothesis with respect to the pictorial stimuli; heavy drinkers exhibited an attentional bias for alcohol related images. Townsend and Duka proposed that the lexical stimuli were more abstract than the images and therefore were unable to elicit sufficient emotional reaction and subsequent bias in attentional allocation.
68 Yeomans, Javaherian, Tovey, and S tafford (2005) also used a dot -probe task to test the hypothesis that high and moderate caffeine consumers exhibit greater bias toward caffeine related lexical stimuli than non -consumers, with high consumers displaying the largest bias. Results partially c onfirmed the hypothesis; high caffeine consumers displayed an attentional bias toward caffeine related stimuli, yet no differences were noted for moderate or non -consumers as a function of stimuli -probe location. As such, Yeomans and colleagues (2005) dem onstrated the presence of attentional bias for caffeine -related words among individuals who consume large amounts of caffeine, but not among individuals who do not to have strong personal association with caffeine and the effects of caffeine. They reasoned that because the presence of caffeine related stimuli has little relevance to moderate and non -consumers, it consequently does not capture their attention. In sum, the dot -probe paradigm has been used successfully to gain insight into the cognitive underpinnings of anxiety disorders and various chemical addictions. The findings are generally consistent in showing that cognitive biases toward motivationally salient (e.g., threatening, state/dependence relevant) stimuli influence how individuals allocate their attention, and subsequently how they function. For highly identified fans, the concept of fan is central to their identity, and therefore team relevant material is personally meaningful and salient to these individuals. As such, fans may possess an increased sensitivity and exhibit increased emotional reactivity to elements in the environment that are related to their team and the teams performance. This elevated emotional investment may manifest itself through fluctuations in their attention and s ubsequent behavior. Attentional Breadth and Flexibility The ability to appropriately allocate and manage attentional focus is vital for effective engagement within ones surroundings and as a result, researchers have been driven to
69 understand the factors t hat alter attentional flexibility and scope. Of particular interest to the current investigation is the dynamic interplay among emotion, motivation, and attention. While considerable research has linked positive affect with broader, more flexible cognition as evidenced by increased creativity and elevated problem solving ability (e.g., Isen (1987, 2000), a more recent line of research has explored the attentional alterations that are tied to affective and motivational changes. Grounded in Isens work (1987, 2000), Friedman and Frster (2005) proposed that approach motivational states also promote greater attentional flexibility, an ability to adaptably shift focus among cognitive operations using executive control (p.70). They suggest that positive affect allows for more information to be available for processing, while increasing the amount of relevant content that can be attended to. Taken together these changes are proposed to facilitate attentional shifting. Preliminary work by Kuhl and Kazn (1999) pr ovided empirical and theoretical grounds for Friedman and Frsters (2005) hypotheses. They found that relative to negative affective states, experimentally induced positive affective states were able to significantly increase attentional flexibility as ev ident by reduced response times when the word and ink color were incongruent. The Stroop task (Stroop, 1935) is commonly employed as an index of attentional flexibility because participants are presented with color words that are printed in avariety of ink colors and are then asked to identify the color ink that the word is printed in rather than the word itself. Because the typical response is to read the printed word, participants must shift their task focus to naming the ink color when the word and ink are incongruent. With that said, the task of naming the ink color is easier under states of elevated attentional flexibility. In sum, this classic task provides a quantitative index of the ease with which one is able to shift focus.
70 Friedman and Frster (2 005) were able to extend the previous work of Kuhl and Kazn (1999) to the experience of motivational states outside of conscious awareness using maze tasks (i.e., seeking reward/avoiding threat) (Experiment 1) and motor actions (i.e., flexion/extension) (Experiment 2) to activate the approach and avoidance motivational systems. In their two experiments they showed that approach motivation (compared to avoidance motivation) increased attentional flexibility (e.g., reduced incongruent ink color naming respon se time) They concluded, approach -related motivational states enhance the flexibility of human behavior (p.79, Friedman & Frster, 2005). In addition to emotion and motivational orientation influencing the ability to shift attentional focus, these affe ctive variables also have been shown to significantly alter the breadth of attention (e.g., Derryberry & Tucker, 1994; Easterbrook, 1959). Frster, Friedman, zelsel, & Denzler (2006) synthesized the current state of the literature, stating that avoidanc e related states such as anxiety (i.e., tense arousal) narrow the focus of attentionand approach-related states such as joy (i.e., elated arousal) broaden the focus of attention, leading to increased responsiveness to peripheral cues on the perceptual level and increased activation of relatively inaccessible mental representations on the conceptual level (p. 134). In line with Frster and colleagues (2006) assertions, Gasper and Clore (2002) showed that state affect influences the processing of global a nd local information, with happiness (i.e., approach motivational state) stimulating a global rather than local perceptual focus. Moreover, participants in the approach condition of Frster and colleagues (2006) study demonstrated faster response times to global targets compared to local targets and participants in the avoidance condition responded faster to local, compared to global targets. The global local target detection task is commonly used to capture the breadth of attention as it simultaneously pre sents
71 participants with both global and local targets (e.g., Frster et al., 2006; Gable & HarmonJones, in press; Navon, 1977). Specifically, response times to target letters presented within composite stimuli (i.e., large letters formed by smaller letter s) are used as a gauge of the scope of attention, with faster times to the large targets indicative of a global focus and faster response times to the smaller targets as a local (i.e., narrow) focus of attention. In general, global features are processed m ore quickly than local features, as the general affective state of most people is typically positive, yet changes in emotion, and consequently motivational state, influence the scope of attention (Gable & HarmonJones, in press; Gasper & Clore, 2002). The interrelationship among emotion, motivation, and attention signifies the functional importance of understanding the factors that influence each individual component in order to better understand the nature of behavior. These variables are of particular in terest to understanding the growing population of highly identified sport fans as they experience a myriad of emotions resulting from their affiliation with a team. Limitations in the Literature In recent years, there has been a surge in the number of studies exploring the various correlates of sport fandom, which have yielded important information concerning the psychological costs and benefits of team identification. However, the existing body of literature is limited due to a number of theoretical, metho dological, and interpretative shortcomings. First, there has been relative inconsistency in the operational definition of fan and the classification of high and low identified fans. Specifically, researchers have unsystematically used the SSIS (e.g., W ann & Branscombe, 1992; Wann, Dimmock, & Grove, 2003), the SFQ (e.g., Wann & Pierce, 2005), consumption behaviors (e.g., Kerr et al., 2005), and self classification (e.g., Wann, Friedman et al., 2003; Wann et al., 2004), along with nonvalidated assessment s of identification (e.g., Hirt et al., 1992) to classify fans. As a result of this
72 variability, the generalizations that can be drawn from existing research are limited, and thereby leave questions concerning the nature, correlates, and consequences of va rying levels of identification unanswered. Next, the majority of what is known about the psychology of sports fans is grounded in correlational evidence. As such, the direction of the relationship between identification and fans thoughts, emotions, and behaviors is uncertain. Moreover, the data extracted from studies exploring the psychology of sports fans has been largely one -dimensional. That is, researchers have either taken a subjective approach using self report assessments of thoughts, feelings, an d behaviors (e.g., Branscombe & Wann, 1991; Schwarz et al., 1987; Schweitzer et al., 1992; Wann et al., 2004) or have employed more covert methods monitoring physiological fluctuations in response to sports cues (e.g., Hillman et al., 2000, 2004). Although each approach has yielded valuable information pertaining to the correlates and consequences of fandom, a gap in extant knowledge exists. Namely, there is presently no evidence linking the physiological changes fans experience in response to sport stimuli to their thoughts and behaviors, nor is there data to indicate if/how increased social -psychological health impacts cognitive processing (e.g., how attention is allocated, where attention is directed) Likewise, there is no evidence to connect extreme spo rt fandom with what is known about other dependencies/pathologies. Third, a majority of the existing research (e.g., Branscombe & Wann, 1991; Wann, Dimmock, & Grove, 2003; Wann & Steele, 1998) has relied solely on subjective report assessment techniques ( e.g., questionnaires and interviews) that are unable to tap the mechanisms that drive fan behavior and emotion. Additionally, the current perspectives on identification and psychological well -being have been primarily founded on data collected from homoge neous populations (e.g., college students who were fans of their schools basketball
73 team). However, the conclusions drawn from this research have been broadly applied to fans in general. Researchers have recognized this limitation and more recent work has become more diverse, exploring if the existing trends in the data are maintained across various fan demographics and sport characteristics. The conceptual and methodological concerns discussed above highlight the modifications necessary to advance the c urrent state of the literature. More systematic exploration of the cognitive, behavioral, and affective factors associated with team identification in controlled environments can provide insight into how fans process information, and specifically, how spor t relevant stimuli influence cognitive processing, attention, daily functioning, and mental health. Conclusion Sport is an essential element of many peoples social and personal lives. Unfortunately to date, research has failed to adequately consider the cognitive and emotional processes that drive sport fans, especially those fans that define themselves by their affiliation with a specific team. While there are countless anecdotes and supporting empirical evidence that highlight the extreme behaviors of highly identified fans, our knowledge has not progressed beyond description of the superficial actions and reactions of fans. That is, in order to truly understand the fan researchers must look to identify cognitive hallmarks that distinguish highly identified fans from their less identified counterparts. While highly identified fans may possess psychologically healthy profiles, I propose that fans emotional reactivity (e.g., Hillman et al., 2000, 2004) and ensuing behaviors (e.g., Kerr et al., 2005) in sport specific situations manifest in maladaptive ways that are similar to those demonstrated by individuals with affective disorders (e.g., excessive consumption, aggression/violence, impaired daily functioning). By inducing pleasant and unpleasant emoti ons through the recall of game specific memories and assessing consequential attentional and
74 behavioral alterations, we can advance our understanding of the cognitive basis and consequences of identification. These findings will aid in discovering the cognitive, affective, and behavioral characteristics of highly identified sports fans. Extending this line of research will help to identify the cognitive processes that play a role in promoting both adaptive and maladaptive behaviors among highly identified s ports fans.
75 CHAPTER 3 METHODS Participants To determine an adequate sample size for the repeated measures analysis, an a priori power analysis was conducted using the G*Power general power analysis program (Erdfelder, Faul, & Buchner, 1996). Alpha level was set at 0.05 with a corresponding power of .80 to detect a medium effect size. A medium effect size was selected according to the convention established by Cohen (1988) for estimating effect size in the absence of existing data. Based on these requirem ents, a total sample size of 110 was determined to be adequate. One hundred and forty three participants (111 females and 32 males) were recruited from the University of Florida and the surrounding local community. Undergraduate students participated in this study for extra credit. All participants were at least 18 years of age ( M = 20.71, SD = 2.17). Materials Self -Report Measures Demographic questionnaire A questionnaire was developed to acquire basic demographic information including age, sex, ethnici ty, and affiliation with the University of Florida (see Appendix A). Sport Spectator Identification Scale The Sport Spectator Identification Scale (SSIS; Wann & Branscombe, 1993) is a valid and internally consistent (Cronbachs =.91) sevenitem, eightpo int Likert scale inventory developed to assess identification (i.e., ones psychological connection) with a specific sport team. In the current investigation, the SSIS was phrased to reflect ones identification with the University of Florida football team An individuals score on the SSIS is obtained by summing
76 the responses to all seven questions and dividing the total by seven; the lowest identification score is one and the highest identification score is eight. Sport Fandom Questionnaire The Sport Fa ndom Questionnaire (SFQ; Wann, 2002) is a psychometrically sound (i.e., valid, internally consistent: Cronbachs =.96, and reliable: r = .94) five item self -report questionnaire designed to assess an individuals identification with his/her role as a spo rt fan and self -perception as a sport fan (i.e., not the fan of a specific team or player). The SFQ Likert scale is anchored by the statements Strongly Disagree (1) and Strongly Agree (8). Similar to the SSIS, an individuals SFQ score is reported as the average of five responses, with a score of eight representing the highest level of general fandom and a score of one representing the lowest level of fandom. State Trait Anxiety Inventory The trait subscale (TAI) of the State Trait Anxiety Inventory (STAI; Speilberger, 1983) is a twenty item Likert -scale inventory that assesses how people generally feel. For each question participants respond on a scale ranging from 1 (Almost Never) to 4 (Almost Always). The TAI has a long history of use in the social scie nces, including sport psychology, and has been reported to have a strong internal consistency ( r =.90), and test retest reliability (r = .75 across 104 days) (Spielberger, 1983). In the current study the TAI was used to assess whether or not trait anxiety is associated with level of fan identification. Positive and Negative Affect Schedule The Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) is a brief 20 item inventory that assesses two dimensions of mood, positive affect (PA) (i.e., the degree to which an individual feels excited, alert, etc) and negative affect (NA) (i.e., the degree to which an individual feels upset, nervous, etc). The two 10 item orthogonal scales (i.e., PA
77 and NA) that constitute the PANAS are both inter nally consistent (Cronbachs =.88 (PA) and Cronbachs =.87 (NA)) and valid. Although the PANAS can be used to assess mood across various time periods (e.g., momentary, daily, weekly, yearly), the general directions (i.e., indicate to what extent you generally feel this way, that is, how you feel on average) were used for this application. Behavioral Inhibition System / Behavioral Activation System Scales The Behavioral Inhibition System / Behavioral Activation System Scales (BIS/BAS; Carver & White 1994) is a 24 question, four -scale inventory, developed to assess differences in personality related to variations in the sensitivity of two motivational systems, behavioral inhibition (BIS) (i.e., aversive motivation avoid something unpleasant) and be havioral activation (BAS) (i.e., appetitive motivation approach something that is desired). Specifically, the single BIS scale is concerned with responsiveness to cues of punishment and anxiety, whereas the three BAS scales focus on: drive (i.e., persist ence toward goal attainment), fun seeking (i.e., desire for new rewards), and reward responsiveness (i.e., positive reactions to actual rewards or the anticipation of a reward). The behavioral inhibition and behavioral activation systems were explored to f urther understand the relationship between personality (i.e., anxiety and impulsivity) and fandom, as well as the relationship between state motivational orientation and trait system activation. Beck Depression Inventory The Beck Depression Inventory (BDI ; Beck & Steer, 1987) is a 21 item inventory with high internal consistency (Cronbachs = .81 for nonpsychiatric patients; Beck, Steer, & Carbin, 1988) and acceptable concurrent validity (r = .60 to .74 with nonpsychiatric patients; Beck et al., 1988) th at measures the typical symptoms and manifestations of depression (e.g., sadness, guilt,
78 sense of failure). For each item of the inventory, participants read a group of four statements and select the one statement that best describes the way that he/she h as been feeling in the past week. Statement values range from zero to three and the responses for all 21 items are tallied to calculate an overall BDI score. It is suggested that scores ranging from five to thirteen are normal (i.e., no or minimal depres sion), whereas scores greater than or equal to 14 may be indicative of some form of depression (i.e., 14 19 = mild, 2028 = moderate, and 2963 = severe) (Groth -Marnat, 2003). Movement Imagery Questionnaire -Revised The Movement Imagery Questionnaire Revis ed (MIQ -R; Hall & Martin, 1997) is an eight item inventory designed to assess an individuals ability to see (visual imagery) and feel (kinesthetic imagery) movements. For each question, the participant is instructed to produce a described movement and the n to either feel or see themselves executing that same movement again, without actually moving. The participant then rates the ease/difficulty with which the image was created using a 1 (vary hard) to 7 (very easy) Likert scale. Scores from each of the dim ensions are summed to provide an index of visual and kinesthetic imagery ability. Visual analog scales Two visual analog scales were developed to assess immediate affective experiences following the emotional imagery sessions. Visual analog scales are b rief, simple scales that place minimal cognitive demands on the respondent and minimally disrupt the completion of other tasks. The scales are 100 unit horizontal lines without unit markers and include anchor words at each end to identify the maximal and m inimal levels of the state being assessed (Brown, 1994). Descriptive anchors (sad -happy, angry calm) (Crisp et al., 2007) were selected to capture how the participants are feeling at a given instance (Figure 3 1). Scores on the scale range from zero to 10 0, with the scores increasing in value from left to right along the scale.
79 Figure 3 1. Visual analog scales. Experimental Stimuli and Task Emotional imagery scripts Personalized, script driven imagery has been used successfully to evoke emotion and affec tive responses (e.g., Miller, Patrick, & Levinson, 2002; McNally, Lasko, Clancy, Macklin, Pitman, & Orr, 2004; Velasco & Bond, 1998). In the typical design, participants provide a detailed description of a situation they have previously experienced, emphas izing the sensations and feelings they had during the actual experience. In the current study, participants were asked to identify their three best and two worst experiences as a University of Florida fan from a list of 33 team wins and 18 team losses repr esenting championships, rivalries, and other significant competitions. The participant also had the option of selecting games not provided on the lists by simply providing the year of the game, the opponent, and outcome. For each game, participants were pr ompted to identify their thoughts and feelings they experienced during the game, as well as any other information they deem relevant to their experience. Based on the information provided and game descriptions acquired from University of Florida Athletic A ssociation press releases, a total of five scripts were developed for each participant, with scripts across participants being matched in length (30 -seconds) and quality (5 stimulus and 5 response propositions). All scripts were read and recorded by the pr imary researcher for playback in the laboratory to induce two pleasant conditions, two unpleasant conditions, and one pleasant expectancy condition (see Appendix B for sample scripts). Visual analog scale ratings were used as a manipulation check. S a d H a p p y A n g r y C al m
80 Stroop task The classic Stroop task (Friedman & Frster, 2005; Stroop, 1935) is a measure of attentional flexibility (i.e., the ease with which one is able to shift focus ). In this task, participants are presented with color words that are printed in variety of ink colors and are asked to identify the color ink that the word is printed in rather than the word itself. Because the typical response is to read the printed word, participants must shift their task focus to naming to the ink color when the word and ink are incongruent. With that said, the task of naming the ink color is easier when attention is more flexible (i.e., better ability to shift between attending to the meaning of the word and the color of the text). This increased flexibility promotes faster response times to stimuli that are incongruent. In the current study, a modified Stroop task was created in which four color words (red, blue, green, yellow) were each presented 12 times, with the ink color and word meaning matching on half the trials. On the remaining trials the word meaning and ink color were incongruent (e.g., the word blue was printed in red ink). All color words were displayed in each incongruent color twice. Additionally, eight catch trials (i.e., XXXX) were included, with each in k color being presented twice. Participants identified the ink color of the stimulus using the V (red), B (yellow), N (green), and M (blue) keys on the keyboard that were labeled with each color word printed in black ink. Each trial was preceded by a centrally positioned fixation cross that was presented for 500ms and each color word remained on the screen until a response was made. A program run on the E -Prime 2.0 Professional platform (Psychology Software Tools, Inc., Pittsburgh, PA) managed stimulus presentation and data collection. Global local target detection task Breadth of attention was assessed using composite stimuli that consist of large letters formed by smaller letters. In the task, participants were asked to respond as quickly as possi ble
81 to the presence of target letter, which was presented within a composite stimulus (i.e., a large letter formed by smaller letters) (Frster et al., 2006; Gable & Harmon Jones, in press; Navon, 1977). Under normal processing conditions, global feature s are processed more quickly than local features, yet motivational state (approach avoidance) has been shown to alter attentional focus and the processing hierarchy. In general, approach (i.e., pleasant) states promote a global focus (i.e., attending to mo re general information), whereas avoidance (i.e., unpleasant) states promote a more narrow scope of attention (i.e., attending to more specific details) (Frster et al., 2006; Gasper & Clore, 2002). However, high approach motivational states (i.e., states tied to active goal attainment) have been associated with a reduced scope of attention, as Gable and Harmon Jones (in press) suggest that information not relevant to goal attainment must be disregarded. In the current study, a program running on the E -Pri me 2.0 Professional platform (Psychology Software Tools, Inc., Pittsburgh, PA) randomly presented participants with 32 global -local target detection trials. A centrally positioned fixation cross was presented for 500ms and signaled the impending stimulus presentation. One of eight composite letters then replaced the fixation cross and participants were instructed to identify as quickly and accurately as possible if the stimulus contained the letter H or the letter L, whether globally or locally portraye d, by pressing the corresponding letter on the keyboard. Global targets are those in which the H or L was composed of an arrangement of smaller letters (T or F) and local targets are those where the large T or F was made up of smaller letters ( H or L) (see Figure 3 1). Sixteen local and 16 global trials were randomly presented. Faster responses to the global targets indicate a broader, more global focus, whereas faster responses to the small (i.e.,
82 local) targets reflect a more narrow attent ional focus (Frster et al., 2006; Gable & HarmonJones, in press). Figure 3 1. Example of composite stimuli depicting global (left) and local (right) targets. Procedure Participants were recruited from the University of Flori da and the surrounding local community and directed to a secure website that contained the informed consent document and questionnaires. Upon providing consent (see Appendix C), participants electronically completed a demographic questionnaire, SSIS, SFQ P ANAS, BIS/BAS, BDI, and MIQ R (see Figure 3 2). They then were asked to briefly describe their three best (i.e., most pleasant) and two worst (i.e., most negative/ unpleasant) experiences as a University of Florida football fan and rate how they feel when they think about the selected game using the two VAS. Following completion of the initial assessment, a second, laboratory -based session was scheduled. Upon entering the laboratory for the second session, participants were escorted to a designated testing room and were seated comfortably in front of a 19 inch Dell Ultra Sharp flat panel monitor (Dell Inc., Austin, TX) and Dell Optiplex GX620 computer (Pentium D processor; Dell Inc., Austin, TX) equipped with a standard keyboard and mouse. Once acclimated, participants were introduced to the two VAS and were instructed to indicate how they felt at that moment (see Figure 3 2). These ratings served as baseline affective data. Next, the Stroop and global -local target detection tasks were described and partici pants were provided with 32 global -
83 local target detection practice trials and 56 Stroop practice trials (i.e., one trial block of each task) administered via E -Prime 2.0 Professional software (Psychology Software Tools, Inc., Pittsburgh, PA) to familiarize themselves with the task demands. After successful completion of the practice phase, the task sequence was randomized and counterbalanced across participants (see Figure 3 2). Figure 3 2. Experimental design. 1. Pleasant script_1 30-second listen 30-second image 2. VAS 3. Stroop 1. Unpleasant script_1 30-second listen 30-second image 2. VAS 3. Stroop 1. Expectancy Induction 2. Pleasant script_3 30-second listen 30-second image 3. VAS 4. Stroop* 5. Global-Local Task* *randomized & counterbalanced* Experimental Condition *randomized & counterbalanced* Debriefing 1. Pleasant script_2 30-second listen 30-second image 2. VAS 3. Global-Local Task 1. Unpleasant script_2 30-second listen 30-second image 2. VAS 3. Global-Local Task 1. 60-second rest period 2. VAS 2. Global-Local Task Initial Online Assessment Demographic SSIS SFQ BIS-BAS PANAS (trait) STAI (trait) BDI MIQ-R Imagery Development Prompts VAS General Reactions Availability & Contact InfoLab SessionTask Instruction & Practice VAS Stroop Practice* Global-Local Practice* *randomized & counterbalanced*
84 In the experimental porti on of the laboratory session, participants were instructed to carefully listen to a 30 -second audio recording that described one of their personal (pleasant/unpleasant) experiences as a University of Florida football fan (listening stage), and create a viv id mental representation of the experience while listening to the narrative. Following the 30-second listening stage, participants were prompted to again mentally recreate the experience as vividly and accurately as possible for another 30 -second period (i maging stage) (McNally et al., 2004; Velasco & Bond, 1998). A tone sounded at the conclusion of the imaging stage and participants immediately rated their current affective state (i.e., anger, sadness, happiness) using the VAS. After their ratings were gi ven, participants were presented with either the global -local target detection task or the Stroop task. Task presentation was randomized and counterbalanced across participants. This sequence was repeated three times, with the participant completing both the global local target detection and Stroop tasks twice, with each trial block being paired with a new pleasant and unpleasant narrative. Participants were also presented with two neutral conditions that did not contain emotional imagery prompts, but inst ead allowed participants to rest quietly for 60 -seconds prior to completing the VAS and subsequent task. These two conditions were randomly integrated into the experimental sequence. Finally, participants were randomly presented with an electronic messag e notifying them that after completing the next two tasks, they would be asked to share what it means to be a Gator football fan and that responses from all study participants would to be compiled and provided to the Universitys Athletic Association to be used in a new marketing campaign launching in the fall. Moreover, they were informed that the author of the entry which best captured the Gator spirit would receive two tickets to a friends and family team dinner. Participants were then presented with a third pleasant narrative (i.e., the expectancy condition) that was immediately
85 followed by a 30-second imaging stage, completion of the VAS, the Stroop task, the global -local target detection task (task order was randomized and counterbalanced across part icipants), and the opportunity to draft a statement addressing what it means to be a Gator football fan. Once the participant completed all six experimental conditions, they were debriefed and any questions were answered (see Figure 3 2).
86 CHAPTER 4 RESULTS I assessed changes in attentional breadth and flexibility among high and low identified fans to identify affective and behavioral consequences associated with team identification and competition outcome. One hundred and forty three participants (111 females and 32 males) completed the online assessments, and then participated in the follow up laboratory session. Only data from participants with low (SSIS 3 5; N = 32 [3 males, 29 females]) and high (SSIS > 6; N = 73 32 [23 males, 50 females]) le vels of identification with the University of Florida football team were retained for subsequent analysis. These two remaining groups were significantly different in their level of reported team identification ( F (1, 104) = 419.20, p < .001: see Table 4 1 ). Table 4 1. Descriptive statistics and results of univariate analysis of variance of self -report measures. Asterisks (*) denote significant differences ( p < .05) between groups. Measure Low ID High ID Statistics Age 20.97 (2.02) 20.81 (2.30) F (1, 104) = .117, p = .73 SSIS 4.25 (.61) 6.95 (.63) F (1, 104) = 419.20, p < .001* Years a Fan 4.00 (3.09) 7.45 (6.31) F (1, 104) = 8.60, p = .004* SFQ 3.72 (1.40) 6.38 (1.34) F (1, 104) = 86.07, p < .001* STAI 39.50 (8.07) 37.85 (6.68) F (1, 103) = 1.19, p = .28 BDI 6.28 (5.59) 5.86 (4.31) F (1, 103) = .17, p = .68 Drive 2.88 (.53) 2.91 (.60) F (1, 103) = .07, p = .80 Fun 3.16 (.41) 3.22 (.54) F (1, 103) = .31, p = .58 Reward 3.59 (.34) 3.61 (.30) F (1, 103) = .12, p = .73 BIS 2.96 (.44) 2.93 (.44) F (1, 103) = .10, p = .75 PA 34.8 (5.40) 36.59 (5.25) F (1, 103) = 2.50, p = .12 NA 19.75 (6.63) 18.22 (5.02) F (1, 103) = 1.69, p = .20 Kinesthetic 5.18 (1.55) 5.24 (1.34) F (1, 103) = .05, p = .83 Visual 5.66 (1.13) 5.95 (.95) F (1, 103) = 1.82, p = .18
87 Psychological Variables High and low identified fans were assessed for general psychological well -being as indexed by trait anxiety (STAI), depressive symptomology (BDI), approach avoidance tendencies (BIS BAS), and affect (PANAS). Group differenc es were compared using a multivariate analysis of variance (MANOVA) to minimize Type I Error (Schutz, Smoll, & Gessaroli, 1983). The Boxs Test was not significant, indicating that the assumption of homogeneity of variance was met ( F (36, 13063.1) = 1.37, p =.07). An overall significant difference was not found between high and low identified fans on measures of personal psychological well -being (Wilks Lambda = .96, F (8, 96) = .51, p = .85, 2 = .04: see Table 4 1). However, to account for variance attribut ed to level of general sport fandom, the number of years as a fan of the target team or to any of psychological variables of interest (e.g., trait anxiety, depressive symptomology, general affect), covariate analyses were also conducted for the dependent m easures of attentional flexibility (i.e., response time to incongruent Stroop conditions) and breadth (i.e., response time to global and local targets). The application of covariate analyses are susceptible to reduction in cell size due to listwise deletion resulting from response omission. However in the current study, the covariate analyses were not compromised because complete response sets were recorded across participants. Changes in Emotion The ability of participants to create vivid and realistic i mages depicting five personal experiences as a University of Florida football fan was crucial to the experimental design as mental imagery was the means for evoking team relevant affective responses. Researchers (e.g., Callow, Hardy, & Hall, 2001; Callow & Waters, 2005) have recommended that a score of four or greater on each subscale of the MIQ R is necessary in order to demonstrate adequate imagery ability requisite for effective imagery based interventions. Participants in the current study
88 reported acce ptable visual ( M = 5.86, SD = 1.01) and kinesthetic ( M = 5.22, SD = 1.40) imagery ability, suggesting a relative ease in being able to see and feel imagined movements. A multivariate analysis of variance (MANOVA) indicated that high and low identified fans reported similar kinesthetic and visual imagery abilities (Wilks Lambda = .98, F (2, 102) = .95, p = .39: see Table 4 1) as measured by the MIQ R (Hall & Martin, 1997). Additionally, a 2 (identification: low, high) x 4 (condition: neutral, pleasant, expec tancy, unpleasant) mixed model multivariate analysis of variance (MM MANOVA) was conducted as a manipulation check to ensure that the personal imagery scripts were effective in inducing distinct pleasant and unpleasant emotional states among high and low i dentified fans. Significant differences in affect were noted across conditions (Wilks Lambda = .33, F (6, 96) = 32.46, p < .001, 2 = .67). Significant differences were noted for both anger ( F (3, 303) = 78.78, p < .001, 2 = .44) and sadness ( F (3, 303) = 124.53, p < .001, 2 = .55) across conditions. A series of follow up paired samples t tests with adjusted alpha levels ( = .008) to prevent inflation of the Type I error rate were used to identify affective differences (Liu, 2002). Pairwise analyses revea led increased anger and sadness (i.e., lower VAS scores) following the unpleasant condition compared with the neutral, pleasant, and expectancy conditions ( p < .001) (Figure 4 1). No significant differences in anger or sadness were noted between the neutra l and pleasant or expectancy conditions ( p > .30). The multivariate analysis did not reveal a significant identification x condition interaction (Wilks Lambda = .93, F (6, 96) = 1.27, p = .28, 2 = .07), nor was there a significant main effect for identi fication (Wilks Lambda = .98, F (2, 100) = .40, p = .84, 2 = .02). However a significant identification x condition interaction was noted for sadness ratings ( F (3, 303) = 3.28, p = .02, 2 = .03), indicating that highly identified fans reported lower le vels of sadness (i.e.,
89 40 50 60 70 80 90 100 Neutral Pleasant Expectancy UnpleasantVAS Ratings Low Id High Id higher VAS scores) following neutral, pleasant, and expectancy conditions, yet reported greater levels of sadness following unpleasant conditions, compared to low identified fans (Figure 4 2). Figure 4 1. VAS ratings by con dition across participants. Figure 4 2. Identification by condition VAS sadness ratings. Changes in Attention Stroop Task An initial 2 (identification: low, high) x 4 (condition: neutral, pleasant, expectancy, unpleasant) mixed model analysis of variance (MM ANOVA) was conducted to assess changes 0 10 20 30 40 50 60 70 80 90 100VAS Ratings Anger Sadness Neutral Pleasant Expectancy Unpleasant
90 in attentional flexibility among fans with varying levels of identification across various affective conditions. Only data for accurate responses were included in the analysis and a minimum of 80% accurac y was required. The mean response accuracy across affective conditions was 96.43% and data from two participants were removed from the analysis. The response times to incongruent stimuli met the assumption of sphericity ( 2(5) = 11.17, p = .05), therefore no adjustments were necessary. The identification x condition interaction was not significant ( F (3, 300) = .24, p = .87, 2 = .002) and the response times did not vary across the four conditions (F (3, 300) = .31, p = .82, 2 = .003), indicating that attent ional flexibility was not influenced by the participants affective state. Moreover, Stroop task performance did not vary as function of level of identification ( F (1, 100) = .83, p = .37 2 = .008: Figure 43). Figure 4 3. Stroop task response times on incongruent trials. When controlling for participants general level of sport fandom (i.e., SFQ) through a repeated measures analysis of covariance (RM ANCOVA), however, a significant condition main effect ( F (3, 297) = 3.22, p = .02, 2 = .03) and condition x identification interaction ( F (3, 297) = 2.73, p = .04, 2 = .03) emerged (Figure 4 4). The significant interaction was followed up by a series of 2 (identification) x 2 (condition) RM ANCOVAs, each with an alpha level adjusted 0 200 400 600 800 1000 1200Response Time (ms) Low Id High Id Neutral Pleasant Expectancy Unpleasant
91 700 750 800 850 900 950 1000 Neutral Pleasant Expectancy UnpleasantResponse Time (ms) Low Id HIgh Id to .008 in order to prevent the inflation of Type I error. A significant interaction was noted for the pleasant -expectancy comparison, indicating that low identified fans responded faster to incongruent Stroop stimuli following the pleasant condition (i.e., demonstrated greater attentional flexibility), whereas highly identified fans responded faster following the expectancy condition ( F (1, 101) = 7.57, p = .007, 2 = .07). Additionally, the neutral unpleasant analysis produced a significant condition main effect ( F (1, 102) = 7.65, p = .007, 2 = .07), demonstrating that the unpleasant condition evoked greater attentional flexibility (i.e., reduced response times) across participants, compared to the neutral condition. Figure 4 4. Stroop task performance marginal means by condition accounting for SFQ. No significant differences in Stroop task performance were noted within or between groups when accounting for the covariates of general anger, state anger, general sadness, state sadness, trait anxiety, positive affec t, negative affect, or years as a fan of the target team. Global Local Target Detection Task An initial 2 (identification: low, high) x 4 (condition: neutral, pleasant, expectancy, unpleasant) mixed model multivariate analysis of variance (MM MANOVA) was c onducted to
92 assess if high and low identified fans breadth of attention changed under varying emotional states. Response times to the presence of global and local targets were used to index breadth of attention. Only data for accurate responses were inclu ded in the analysis and a minimum of 80% accuracy was required. The mean response accuracy across affective conditions was 98.14% for global targets and 98.29% for local targets and data from one participant was removed from the analysis. Task performance on the global local target detection task did not vary as a function of condition (Wilks Lambda = .97, F (6, 97) = .44, p = .85, 2 = .03: see Figure 4 5) or identification (Wilks Lambda = .96, F (2, 101) = 1.93, p = .15, 2 = .04: see Figure 4 6). The i dentification x condition interaction (Wilks Lambda = .92, F (6, 97) = 1.42, p = .22, 2 = .08) was also non -significant (Figure 4 7). Figure 4 5. Global and local target detection response times by condition. Significant differences, however, w ere found both between and within groups across several of the repeated measures multivariate analyses of covariance (RM MANCOVA). First, when accounting for general level of fandom (i.e., SFQ), a significant condition main effect 0 100 200 300 400 500 600 700 800 900Response Time (ms) Global Targets Local Targets Neutral Pleasant Expectancy Unpleasant
93 580 600 620 640 660 680 700 720 740 Neutral Pleasant Expectancy UnpleasantResponse Time (ms) Global_Low Global_High Local_Low Local_Highemerged for both global ( F (3, 303) = 2.87, p = .037, 2 = .028) and local ( F (3, 303) = 2.71, p = .045, 2 = .026) response times across participants (Figure 4 8). Follow up analyses, with adjusted alpha levels ( = .008), only revealed significant differences on local target det ection times between the neutral and pleasant conditions ( F (1, 102) = 8.96, p = .003, 2 = .081), with faster response times being evident following the neutral condition. Figure 4 6. Global and local target detection response times by identificati on. Figure 4 7. Condition by identification global local target detection task performance. 0 100 200 300 400 500 600 700 800 900 1000Response Times (ms) Global Targets Local Targets Low Id High Id
94 0 100 200 300 400 500 600 700 800 900Response Time (ms) Global Targets Local Targets Neutral Pleasan t Expectancy Unpleasant Figure 4 8. Global -local task performance marginal means by condition accounting for SFQ. Asterisks (*) denote significant differences ( p < .008) between conditions. Next, taking into consideration the number of years that participants were a fan of the target team, a significant identification main effect emerged for local task performance across affective conditions ( F (1, 101) = 4.41, p = .038, 2 = .042: Figure 49). Highly identified participants responded significantly faster to the local targets than their less identified counterparts, indicating a more narrow scope of attentional focus. Figure 4 9. Local task performance marginal mea ns by identification accounting for number of years a fan. Asterisks (*) denote significant differences ( p < .05) between groups. 300 400 500 600 700 800 900 1000Response Times (ms) Low High
95 Additionally, when accounting for participants state level of anger, as well as state level of sadness, significant differenc es emerged between high and low identified fans across conditions on global (anger: F (1, 101) = 4.32, p = .040, 2 = .041: Figure 4-10; sadness: F (1, 101) = 4.63, p = .034, 2 = .044: Figure 411) and local target detection times (anger: F (1, 101) = 4.12, p = .045, 2 = .039: Figure 4 10; sadness: F (1, 101) = 4.46, p = .037, 2 = .042: Figure 4 11) with less identified fans demonstrating an overall delayed response time to both targets compared to highly identified fans. Finally, when negative affect was us ed a covariate, a significant main effect for identification was evident for global task performance ( F (1, 101) = 4.13, p = .045, 2 = .039), with high identified fans responding faster than their less identified counterparts (Figure 4 12). The analyses t hat included the covariates of general anger, general sadness, trait anxiety, and positive affect were not significant. Figure 4 10. Global local task performance marginal means by identification accounting for state level of anger. Asterisks (* a nd **) denote significant differences ( p < .05) between groups. ** ** Global Local Low Id High Id 400 500 600 700 800 900 1000Response Time (ms)
96 400 500 600 700 800 900Response Time (ms) Low Id High Id Figure 4 11. Global local task performance marginal means by identification accounting for state level of sadness. Asterisks (* and **) denote significant differences ( p < .05) be tween groups. Figure 4 12. Global task performance marginal means by identification accounting for general negative affect. Asterisks (*) denote significant differences ( p < .05) between groups. Supplemental Exploratory Analyses Several supplemen tary analyses were conducted to explore other factors that may have influenced attentional breadth and flexibility. 400 500 600 700 800 900 1000Response Time (ms) ** ** Global Local Low Id High Id
97 Temporal Effects To ensure that the affective changes induced by the emotional imagery scripts did not dissipate over the course of the ta sk and thereby promote additional attentional changes, each task was divided into eight trial blocks, with each Stroop task consisting of three trials and each global -local target detection task consisting of two trails. Trial block data was not available for the expectancy condition due to limitations in the coding script used in E Prime, therefore the following analyses of temporal effects only examine differences among the neutral, pleasant, and unpleasant conditions. Stroop task performance across trial blocks A 2 (identification: low, high) x 8 (trial block: 1 8) x 3 (condition: neutral, pleasant, unpleasant) mixed model analysis of variance (MM ANOVA) was conducted to assess fluctuations in Stroop task performance. The condition data met the assumpti on of sphericity, therefore no adjustments were necessary ( 2(2) = 4.05, p = .13). The assumption of sphericity, however, was not met for the trial block or condition x trail block interaction data (trial block: 2(27) = 41.38, p = .04; condition x trial: 2(104) = 189.47, p < .001), therefore the Greenhouse Geisser adjustment was used. The only significant difference that emerged was the identification x trial block x condition interaction ( F (10.65, 1043.68) = 2.32, p = .009, 2 = .023). This three way in teraction was followed up with individual 2 (identification) x 8 (trial block) mixed model analysis of variance (MM ANOVA) for each affective condition to isolate if the identification x trial block interaction varied across affective condition. The assump tion of sphericity was violated for trial block data for both the neutral ( 2(27) = 89.46, p < .001) and pleasant ( 2(27) = 60.40, p < .001) conditions, therefore the Greenhouse Geisser adjustment was used. Significant identification x trial block interact ions were evident for both the neutral ( F (5.24, 529.53) = 2.21,
98 0 150 300 450 600 750 900 1050 1200 1 2 3 4 5 6 7 8 Trial Block High Id Low Id 0 150 300 450 600 750 900 1050 1200 1 2 3 4 5 6 7 8 Trial Block High Id Low Idp = .049, 2 = .021) and unpleasant ( F (7, 693) = 2.54, p = .014, 2 = .025) conditions (Figures 4 13 and 4 14, respectively). Based on the plotted data, high and low identified fans Stroop task performance with both the neutral and unpleasant conditions were analyzed with a series of one way analysis of variance with adjusted alpha levels ( = .01). No significant differences in response time were noted between high and low identified fans during the neutral condition on trial blocks 1, 5, 6, 7, and 8 ( p > .01) or between the groups during the unpleasant condition on trial blocks 2, 3, 5, 7, and 8 ( p > .01). Figure 4 13. High and low identified fans Stroop task response times during the neutral condition across trial blocks. Figure 4 14. High and low identified fans Stroop task response times during the unpleasant condition across trial blocks.
99 0 150 300 450 600 750 900 1050 1200 1 2 3 4 5 6 7 8 Trial Block Local Global Global local task performance across trial blocks A 2 (identification: low, high) x 8 (trial block: 1 8) x 3 (condition: neutral, pleasant, unpleasant) mixed model multivariate analysis of variance (MM MANOVA) was conducted to assess fluctuations in global local task performance. The multivariate analysis revealed a significant trial block main effect (Wilks Lambda = .48, F (14, 83) = 6.45, p < .001 2 = .52), that was supported by significant univariate trial block main effects for both global ( F (7, 672) = 11.50, p < .001 2 = .11) and local ( F (7, 672) = 11.16, p < .001 2 = .1 0) task performance (Figure 4 15). Additionally, the identification x trial block interaction was significant for global task performance ( F (7, 672) = 2.14, p = .04 2 = .02: Figure 4 16). Figure 4 15. Response times by trial block to global and local stimuli across affective conditions. Based on the trends evident in the plotted data, global task performance on trial blocks 1 and 2 were compared with performance on trial blocks 3 through 8 using pairwise analyses with adjusted alpha levels ( = 003). Trial block 1 response times were significantly slower compared to all other trials blocks ( p < .001), except for trial block 7 ( t (102) = 2.88, p = .005).
100 0 100 200 300 400 500 600 700 800 900 1000 1 2 3 4 5 6 7 8 Trial Block High Id Low IdAlso, response times to global targets during trial block 2 were significantly faster than resp onse times elicited during trial block 7 ( t (101) = 3.56, p = .001). Local task performance between trial block 1 and all other trial blocks was compared using pairwise analyses with adjusted alpha levels ( = .007). Response times to local stimuli durin g trial block 1 were significantly slower than response times to local stimuli in all other trial blocks ( p < .001). Figure 4 16. High and low identified fans response times by trial block to global stimuli across affective conditions. The signifi cant identification x trial block interaction for global target detection was followed up with a series of eight one -way analyses of variance with adjusted alpha levels ( = .006) comparing response times of high and low identified fans. No significant dif ferences in the detection speed of global targets ( p > .009) were noted between the groups for any of the eight trial blocks. Anger According to the original hypotheses, the specific negative emotion (i.e., sadness/anger) evoked by unpleasant team content was proposed to vary as a function of level of team
101 identification, and was hypothesized to be a significant factor in influencing motivational orientation and subsequent attentional changes. Therefore, participants were dichotomized according to the inte nsity of their anger response to unpleasant team scenarios. That is, all participants who reported some level of identification (SSIS 3) with the target team were then classified according to their anger ratings (VAS 25 = low anger, VAS 25 = high anger). As predicted, those participants who reacted with more intense anger, also reported higher levels of team identification ( F (1, 135) = 21.16, p < .001). Anger reactivity and Stroop task performance A 2 (anger: low, high) x 4 (condition: neutral, plea sant, expectancy, unpleasant) mixed model analysis of variance (MM ANOVA) was conducted to assess changes in attentional flexibility among participants with differing anger reactivity across affective conditions. The data met the assumption of sphericity ( 2(5) = 9.19, p = .10), therefore no adjustments were necessary. The anger x condition interaction was not significant ( F (3, 387) = 1.49, p = .22, 2 = .01), and the response times did not vary across the four conditions ( F (3, 387) = .34, p = .80, 2 = .00 3), indicating that attentional flexibility was not influenced by the participants affective state (Figure 4 17). Stroop task performance did not vary as function of anger reactivity ( F (1, 129) = .86, p = .35 2 = .007). Anger reactivity and global -loca l target detection task performance A 2 (anger: low, high) x 4 (condition: neutral, pleasant, expectancy, unpleasant) mixed model multivariate analysis of variance (MM MANOVA) was conducted to assess if participants with high and low anger tendencies exper ience changes in attentional breadth under varying emotional states.
102 0 100 200 300 400 500 600 700 800 900 1000Response Time (ms) Low Id High Id Neutral Pleasant Expectancy Unpleasant 0 200 400 600 800 1000 1200Response Time (ms) Low Anger High Anger Neutral Pleasant Expectancy Unpleasant Figure 4 17. Anger group by condition Stroop task response times. Task performance on the global local target detection task did not vary as a function of condition (Wilks La mbda = .97, F (6, 124) = .62, p = .72, 2 = .03: Figure 418) or anger classification (Wilks Lambda = .98, F (2, 128) = 1.25, p = .29, 2 = .02: Figure 4 19). The identification x condition interaction (Wilks Lambda = .93, F (6, 124) = 1.56, p = .16, 2 = .07) was also non -significant. Figure 4 18. Anger group by condition global target detection task performance
103 0 100 200 300 400 500 600 700 800 900 1000Response Time (ms) Low Id High Id Neutral Pleasant Expectancy Unpleasant Figure 4 19. Anger group by condition local target detection task performance
104 CHAPTER 5 DISCUSSION As the number of indiv iduals identifying themselves as sport fans continues to surge, there becomes a heightened interest in understanding this population, with both the popular media and researchers attempting to understand the role of sport in contemporary life. To date, rese arch has yielded data that has provided insight into the emotional experiences (e.g., Branscombe & Wann, 1991; Wann et al., 2003), cognitions (e.g., Hirt et al., 1992; Schweitzer et al. 1992), and behaviors (e.g., Wann et al., 2003) of highly identified fa ns. Yet, there has been minimal progress made in understanding the relationship among the elements that characterize high fan identification. In an effort to gain inferences toward the cognitive and affective mechanisms that underlie fan behavior, the prim ary purpose of this project was to empirically investigate fluctuations in attentional allocation among high and low identified fans as a function of the emotional state elicited following imagery of pleasant and unpleasant sport experiences. Specific find ings are discussed, strengths and limitations of the investigation are addressed, directions for future research are proposed, and applied implications are suggested. Review of the Findings Psychological Variables Both high and low identified fans demonst rated healthy psychological profiles that were not significantly different from each other. Specifically, both groups reported trait anxiety scores that were within the mean range of scores for college populations (Spielberger, 1983) and demonstrated mini mal depressive symptomology Groth-Marnat, 2003). Not only are these findings consistent with the work of Wann and colleagues (Wann et al., 2003; Wann et al., 2004; Wann, 2006a; Wann, 2006b), they further demonstrate that high and low identified fans do not
105 inherently differ in psychological health using previously ignored measures of personal psychological well -being (i.e., PANAS, STAI, and BDI). Emotion Induction The use of emotional imagery scripts was effective in altering participants emotional states Specifically, unpleasant memories increased all participants rating of anger and sadness beyond their neutral state and the data showed a trend of reduced anger and sadness following both pleasant and expectancy conditions across participants as hypothe sized. The participants relatively positive affective state revealed during the neutral condition may have promoted a ceiling effect which mitigated statistical significance from emerging, as their neutral ratings considerably exceeded the midpoint on the visual analog scale. Previous research (e.g., Diener & Diener, 1996; Gasper & Clore, 2002) has found similar trends, with the typical mood of people being predominately positive. It was unexpected that the level of identification did not interact with the affective condition to promote changes in ratings of anger. I hypothesized that the intense psychological connection highly identified fans have with the target team would influence the magnitude of their affective reactions while imaging, with the highly identified fans reporting more intense positive (i.e., higher VAS ratings) and negative (i.e., lower VAS ratings) emotions following the pleasant and unpleasant scripts, respectively. Highly identified fans did, however, report lower ratings of sadness fo llowing the neutral, pleasant, and expectancy conditions (i.e., they were happier than the less identified fans), and higher ratings of sadness following the unpleasant condition (i.e., they were more sad). These differences were not, however, statisticall y significant. The comparable positive reactions among high and low identified fans were not surprising as previous research (e.g., Cialdini et al., 1976) has shown that both high and low identified fans
106 exhibit a tendency to bask in the reflected glory ( i.e., BIRG), thereby reaping the benefits of team success (i.e., increased positive affect). Yet, the high and low identified fans were hypothesized to experience distinct negative emotions following exposure to unpleasant scenarios (e.g., Branscombe & Wan n, 1992; Crisp et al., 2007; Wann, 1993). When the team performs poorly, highly identified fans perceive that their identity is threatened because the team is viewed as an extension of them and is a meaningful component of their self -concept. Therefore, hi ghly identified fans have a propensity to experience anger and subsequently engage behaviors that degrade the opposition and are aggressive in nature (i.e., blast) (Branscombe & Wann, 1992; Wann, 1993). The current data however does not support the anger -s adness distinction revealed by previous work. Although personalized imagery scripts were used to modulate emotional states, the mental images may not have evoked emotions comparable to those originally experienced when the event took place, and therefore m otivational orientation and attentional allocation were not affected to the same extent as in the sporting context. While mental imagery has been used in a variety of domains to recreate experiences and evoke affective and physiological reactions, the magnitude of such reactions are diminished (e.g., Jacobson, 1930). Previous research (e.g., Gable & HarmonJones, in press) has emphasized that both emotion type and magnitude are influential factors in producing changes in motivational orientation. The curren t study may not have evoked sufficient changes in the type and/or magnitude of emotions (i.e., valence and arousal) experienced by participants beyond their neutral state to warrant discernable differences in attentional breadth and flexibility. To confirm this postulate, it would be necessary to assess affective reactions (subjective or physiological) during or immediately after watching a competition and compare the magnitude with those evoked through emotional imagery scripts.
107 In addition to the inhere nt limitations of mental imagery, the controlled laboratory environment removed the social (e.g., likeminded fans, camaraderie) and environmental (e.g., temperature, noise, crowding) elements of sport spectator experience, which may have reduced or altered the type and magnitude of emotions experienced by participants. Similarly, when recalling fan experiences during the laboratory session, participants knew the eventual outcome, both of the game and the season. When initially experiencing the event there i s uncertainty regarding the consequences and benefits associated with the outcome. Yet when reflecting back on the experience, participants have a different understanding of its significance and relevance. As such, they may have experienced reduced or even distinct emotions in the laboratory relative to those originally experienced. Finally, to ensure that affective salience did not decrease from the end of the imagery period to the conclusion of each task, response times from the eight trial blocks were c ompared across affective conditions and level of identification. Although response consistency was evident for Stroop task performance, global and local target detection times were influenced by trial block. Specifically, a speeding of responses was only n oted after the first trial block (i.e., first two trials), suggesting that all participants required a brief period to re acclimate to the task demands and did not experience a lessening of emotion over the time required for task completion (ranging from one to two minutes). Because both high and low identified fans exhibited complementary response patterns that were consistent across affective conditions, the effects of the first trial block did not compromise interpretation of the data. In sum, the emotio nal imagery scripts were successful in modulating the affective reactions of all fans. However, the reactions of both high and low identified fans were relatively homogeneous, a pattern which may be attributed to the experimental context. Specifically, the
108 data from the current study illustrate that high and low identified fans do not differ on general measures of psychological well -being, but instead only on sport relevant measures. As such, removing participants from the context in which their group diffe rences emerge may have promoted comparable performance differences between the groups. Accordingly, subsequent analyses accounted for the variables that further distinguished these groups (e.g., number of years a fan of the target team, general level of sp ort fandom) in order to illuminate changes in motivational orientations and attentional fluctuations related to high levels of team identification. Changes in Attention Stroop task Counter to expectations, changes in attentional flexibility, as indicated by response times to incongruent Stroop stimuli, were not evident between groups or affective conditions. This finding is particularly surprising, considering the unpleasant scripts evoked significant changes in ratings of anger and sadness. Additionally, although the pleasant scripts did not significantly enhance positive affect among participants, research has consistently shown that both high and low identified fans exhibit a tendency to bask in reflected glory (BIRG; Cialdini et al., 1976), a behavioral pattern consistent with an approach motivational orientation. A number of plausible arguments may account for the lack of discernable performance difference between the two groups. First, the emotional imagery manipulation was unable to evoke emotions of comparable magnitude and valence to those originally experienced by participants as outlined above. Additionally, the magnitude of the effects evoked in the current study may have been reduced because the Stroop task was completed in a neutral environmen t and the teams success/failure was inconsequential to participants self -concept. More specifically, there was no public forum by which the participants sense of self could be enhanced or reduced such that the feelings and behaviors typically evoked by competition
109 environments were absent (Dimmock & Grove, 2005). Next, the comparison groups (i.e., high and low identified fans) may have been too similar, as they all were residents of the local community and/or students at the target university. Although t he two groups were partitioned into different groups based on their level of identification with the Universitys football team, their affiliation with the University as a whole may have augmented their sport team identification. Both high and low identifi ed fans reported comparable anger experiences following all affective conditions suggesting that even the low identified fans reacted to the emotional scripts because of their personal relevance and/or because of repeated exposure to team relevant content (e.g., paraphernalia, news media) (Branscombe & Wann, 1991; Dietz Uhler & Murrell, 1999). Participants may also have over represented their affiliation with the Universitys football team and as a result may have been falsely classified as highly identifi ed fans. As Gibson, Willming, and Holnak (2002) note, Gator football and its fans are a major source of community identity and pride (p.422), both within the University population and among residents of the surrounding areas. Therefore, participants in the current study may possess a sense of relative obligation to be avid supporters of the University and its football program, as they are members of the community. For example, although neither level of identification nor affective condition prompted chan ges in attentional flexibility, group differences emerged when participants general level of sport fandom was considered. When general level of fandom was covaried, the results indicated that low identified fans responded faster to incongruent Stroop sti muli following the pleasant condition (i.e., demonstrated greater attentional flexibility), and highly identified fans responded faster following the expectancy condition. Despite the significance of these findings, they are counter
110 to the initial hypothes es, which proposed that positive affect would promote an approach orientation, and consequently enhance Stroop task performance. Moreover, it was hypothesized that the highly identified fans would experience the most notable reduction in response times fol lowing pleasant scripts because of their intense connection with the target team (Hirt et al., 1992). Alternatively, the expectancy condition was predicted to reduce attentional flexibility among the highly identified fans because of the necessity to ignore irrelevant stimuli under conditions of heightened approach orientation (i.e., the expectancy to act) (Gable & Harm Jones, in press). Recognizing that the manipulation of emotion through imagery may have been inadequate, it is plausible that the addition of the expectancy to act manipulation with the pleasant imagery script was necessary to evoke a basic positive affective experience and approach orientation among highly identified fans. Importantly, when general level of fandom was accounted for, greater attentional flexibility was evident across participants, following the unpleasant condition, compared with the neutral condition. This trend indicates that all participants demonstrated greater attentional flexibility and an approach motivational orientat ion. Therefore, although participants reported both an increase in sadness and anger following exposure to unpleasant fan memories, the anger response was apparently more dominant. In support of this notion, review of the personalized scripts indicates tha t the opposition, officials, and/or supporters of the other team were often deemed responsible for their teams loss (i.e., the unpleasant fan experience) by participants. This tendency to devalue those with whom one is negatively associated with is charac teristic of blasting, a defensive behavior that is highly approach orientated. In sum, differences in attentional flexibility were not evident across affective conditions or between high and low identified fans until participants general level of fandom was statistically
111 controlled. When accounting for general fandom, a greater percentage of Stroop task performance variability was explained by the interaction between level of team identification and affective condition. Global local target detection task Neither affective condition nor level of identification influenced attentional breadth as indicated by response times to global and local targets. The lack of variability in performance across factors suggests that motivational orientation remained stable across the manipulations. Performance differences on the global local target detection task emerged, however, when other related factors were controlled. When participants general level of sport fandom was considered, differences in response time to loc al targets were faster following the neutral condition compared with the pleasant condition across participants, indicating attentional narrowing following the neutral, rest condition. Likewise, the number of years that participants were a fan exerted an i nfluence on the association among the factors of level of identification and task performance; highly identified fans exhibited a more narrow scope of attention across all affective conditions compared to low identified fans. When controlling for participa nts state levels of anger and sadness, highly identified fans responded faster to all targets across affective conditions. The facilitation of performance on tasks with varying attentional demands (i.e., broad versus narrow breadth of attention) suggest that it is not the hedonic valence of the fans emotional experience which is driving performance in the current study, but instead the emotional arousal; the faster response times exhibited by highly identified fans are the result of increases in arousal a fter mentally recreating fan experiences (both positive and negative). To this point, Hillman and colleagues (2000, 2004) have clearly illustrated that highly identified fans perceive team relevant content, depicting both success and failure, to be more ar ousing than irrelevant content and that this
112 heightened arousal appraisal is exclusive to highly identified fans; low and moderately identified fans do not perceive team relevant and irrelevant images differently. Moreover, according to Easterbrook (1959), under normal task conditions (i.e., low stress/anxiety, as in the current study), moderate arousal facilitates task performance. Although arousal was not directly assessed in the current design, the performance differences between high and low identified fans are representative of increased attentional and motivational engagement (Easterbrook, 1959; Hillman et al., 2000, 2004). In sum, when the two groups of fans were parsed using relevant covariates, performance differences emerged between the high and l ow identified fans. Specifically, the trends in the global -local task performance data highlight the integral role of identification related factors in understanding the nature and consequences of team identification. Additionally, these findings suggest that it is the intensity of the emotional experience (e.g., arousal) which promotes behavioral differences between high and low identified fans. Strengths, Limitations, and Directions for Future Research Despite the failure to find support for the original hypotheses, the data from the current study illuminate several important facets of team identification. First, team identification is not a global construct. That is, the affective and behavioral correlates of team identification appear to be highly conte xtual. The mere observation of spectators at a competition reveals distinct behavioral patterns among those who are assumed to be the allegiant and highly identified (St. John, 2004). The traits, characteristics, and mechanisms underlying these distinct a ffective and behavioral reactions however remain somewhat illusive to researchers who are tied exclusively to laboratory based methodologies, as was the case with the current investigation. To clarify, this study suggests that high and low identified fans may only be distinct within the sport
113 environment and that by removing the highly identified fan from the context where both his/her identification is most salient (and emotional reactions are most naturally elicited) we may also have neutralized the facto rs responsible for the differences between high and low identified fans. The contextual specificity explanation for why our hypotheses were not supported is congruent with the finding that our two comparison groups do not differ in measures of personal ps ychological well -being. It appears that at their core, high and low identified fans are more similar than they are distinct, and that differences between these two groups may only emerge in the context of sport spectating. With that said, future research e xploring the consequences and mechanisms of team identification should be conducted within a sport context, and incorporate commonly employed laboratory based measures. Equipping sport spectators with heart rate monitors and personal digital assistants (PD As; handheld computers) would allow for the time locked tracking of fans arousal, as well as the monitoring of cognitive and affective changes through performance differences on basic cognitive tasks (e.g., Stroop task, Global local target detection task) and brief self -report measures (e.g., VAS) which could be completed during stoppages of play. By immersing the design within the sporting context, researchers will be able to directly compare the experiences of high and low identified fans and chronicle c hanges in emotion and attention. Moreover, the incorporation of physiological indices of arousal would provide support for the proposal that fans emotional intensity (e.g., arousal) primes subsequent attentional and behavioral changes. This methodology w ould also facilitate the understanding of the temporal endurance of the physiological and psychological changes experienced by fans, as this design would enable extension beyond the competition venue and incorporation into fans daily routine.
114 Similarly, it is reasonable to conclude that team identification is a complex construct that should not be examined in isolation. As indicated by the covariate analyses conducted in the current study, general interest in sport, the number of years as a fan, and stat e specific reactions are influential factors when exploring the association among level of identification, affective reactions, motivational orientation, and attentional allocation. As such, it is strongly recommended that future researchers account for th ese factors, both from methodological, as well as statistical/analytical perspectives. Structural Equation Modeling (SEM) may provide additional insight into the complex nature of team identification. SEM is a statistical technique that illustrates how s ets of variables define constructs and how these constructs are related to each other, while providing an index of how well the data set satisfies the proposed theoretical model. The two primary variables of interest in SEM are latent/construct variables which are indirectly observed or measured, and hence are inferred from a set of variables that we do measure using test, surveys, and observed/indicator variables which are used to define or infer the latent variable or construct and are able to be m easured directly. Typically, several observed variables are used to define a single latent variable in order to reduce measurement error and add to the strength of the technique. In sum, SEM permits complex phenomena to be statically modeled and testeda llowing for confirming (or disconfirming) theoretical models in a quantitative fashion (Schumacker & Lomax, 2004, p.2 7). The initial stage of SEM is model specification and consists of incorporating pertinent theory and the extant literature to determi ne which variables to include in the theoretical model to be tested. Accordingly, based on the synthesis of previous research and theory from sport and general psychology, as well as the fields of sport marketing and management, it is proposed,
115 Attentional Flexibility STAI error BIS error BAS error Affective Disposition Fanship Years a Fan SSIS SFQ error error error BDI NA PA error error error Mood State Attentional Breadth Global target detection error error Local target detection Stroop task performance error using the c urrent sample, that affective disposition (i.e., positive affect, negative affect, depressive symptomology, trait anxiety, behavioral inhibition, behavioral activation), coupled with fanship (i.e., number of years a fan, level of team identification, gener al sport fandom) influence emotional reactivity and attention. Figure 5 1 depicts the proposed preliminary model. Figure 5 1. The proposed SEM model demonstrates that attentional breadth and flexibility will vary as a function of mood state. Fans mood states are linked to their affective dispositions and fanship characteristics. Fans personal experiences of positive affect, negative affect, depressive symptomology, trait anxiety, behavioral inhibition, and behavioral activation collectively represent ones affective disposition. The number of years a fan, level of team identification, and magnitude of general sport fandom characterize ones fanship. An additional factor that was not included in the current investigation, but that may increas e our understanding of the nature of team identification, is the motives/antecedents for fanship, as the majority of the motives for team identification are based on social and
116 psychological needs (Trail & James, 2001, p.109). The classification of all highly identified fans together may be inappropriate, as this subset of the population may be quite heterogeneous and require further sub-classification. Research (e.g., Fink, Trail, & Anderson, 2002; Trail & James, 2001) has identified numerous reasons wh y people adopt the role of fan and examination of these factors suggest that the motives for sport consumption may play a significant role in emotional experience of fans. For example two fans may report comparable levels of identification with a target t eam (i.e., it is very important to them that they are a fan of the target team, they very much dislike the target teams greatest rival, they always display the target teams logo at home or work, their friends see them very much as a fan of the target tea m), but may be a fan for distinct reasons (e.g., they enjoy spending time with others, it is a simple pastime, increased self esteem/vicarious achievement). As such it is plausible that those individuals who possess motives that are more closely linked to competition outcome may react more intensely than those whose motives are more social in nature. As such, motives for fanship may aid in further distinguishing among high identified fans, permitting factors associated with functional and dysfunctional beha vioral patterns to be isolated. Implications and Conclusions Two hundred million adults in the United States consider themselves sports fans (King, 2005) and this growing population invests significant amounts of time and money into following their sport teams. The aim of this study was to systematically explore the affective, cognitive, and behavioral factors associated with team identification in a controlled environment and to gain insight into how pleasant and unpleasant fan experiences influence atte ntion. Attention is inherent in all aspects of daily functioning, as it drives how we extract and subsequently process, store, retrieve, and apply information. By exploring the associations among team identification,
117 emotion, arousal, and attentional fluct uations, we are able to gain inferences towards the mechanisms that foster the overt differences between high and low identified fans. However, in order to ascertain the implications of the interactive effects of level of identification, emotion, and behavior, future research must explore this question contextually, assessing the type and magnitude of emotions experienced during competition. Researchers can then determine the functional costs and/or benefits of noted changes, and monitor how long the effect s may endure. Also, researchers should manipulate the sport context to assess if fan engagement activities or post -competition affective buffer zones can be used effectively to neutralize any adverse effects. Although it would be premature and inappropria te based on the current studys findings to make recommendations for how this data may be used in an applied context, the trends in the data suggest that moderating spectators arousal will influence their behavior. The concept of attempting to alter fans arousal has already been adopted by many sport organizations in an effort to reduce spectator violence. Specifically, research and policies have focused on limiting or banning the sale of alcoholic beverages, as alcohol has been shown to influence spectat ors arousal levels and consequently their tendency to engage in aggressive and other socially unacceptable behaviors (Russell, 2008; Wann et al., 2001). Additionally, attention is being paid to environmental factors such as heat, noise, and overcrowding which have been shown to elevate fans arousal levels (e.g., Branscombe & Wann, 1992). The implications, however, for arousal regulation may extend beyond public safety issues to include implications for functional behavior. In conclusion, fans are not me rely consumers of sport, but rather they are psychologically invested (Wann et al., 2001). The current study clearly supports this affective engagement of
118 highly identified fans and reveals that fans level of identification interacts with personal and co ntextual variables to influence attentional breadth and flexibility.
119 APPENDIX A DEMOGRAPHIC QUESTIONNAIRE Please complete the following information about yourself. Name : _________________________________ E-mail address (this will be used to contact you to schedule your second session) : ________________@________ Age : _____ Sex : _____ Marital Status : ___ Single ___ Married ___ Domestic Partnership ___ Divorced ___ Widowed ___ Other (please specify) ___________________ Please select all of the followi ng that currently describe your affiliation with the University of Florida (UF). Select as many classifications as appropriate : ___ Alumni ___ Current student ___ Donor/financial supporter ___ Fan of athletic teams ___ Spouse/sibling of current student ___ University of Florida faculty ___ University of Florida staff ___ I have no relationship with the University of Florida ___ Other (please specify) ___________________ How did you find out about this study : ___ Campus flyer ___ Community flyer ___ Instruc tor ___ Word of mouth ___ Other (please specify) ___________________
120 APPENDIX B SAMPLE IMAGERY SCRIP TS Sample Pleasant Scripts It was a hot and sticky evening in September of 2006 and we were crammed in my living room. I was so excited to be with m y friends to watch the Florida Tennessee game, the first SEC game of my college career. The Gators wore their white tops as they took on the Vols. I couldnt help but to feel nervous in the 4th quarter as Tim Tebow needed to gain one yard on a 4th and one. I was overjoyed when not only did he get one yard, but he got two! That play gave us a first down and led us to a touchdown and we won 2120 win! We were so psyched and gave each other hugs and high fives. It was about 6:45pm on a breezy, but warm Satur day in November when the Florida South Carolina game was coming to an end and I could feel the knot in my stomach grow. I was at Ben Hill Griffin Stadium with my family and friends and you could feel the tension building. With only 8 second left in the gam e South Carolina had a field goal chance. My hands were shaking and I could not help but to think what if we lose? But Jarvis Moss blocked the kick and secured the Gator win. I was so overcome with joy that I jumped up and down, cried and screamedit was a mazing! It is was cool, dreary October evening at Alltel Stadium in Jasonville and I was excited to be with my brother and friends at the Florida Georgia game, but it was a frustrating game, as we could not seem to make a field goal the entire game! And in the last seconds of the game Florida had another field goal opportunity. I was nervous as I anxiously stood praying to see the refs signal that it was good. It was! And I was so excitedIt felt so satisfying to beat our rival UGA. Sample Unpleasant Scr ipts The Gators were battling Auburn at Florida Field in September of 2007. In the second half, Auburn came out fighting and the game rested on the shoulders of the Auburn kicker. Thankfully, Urban Meyer called a timeout just as he made his kick and I was relieved. But as he took his second kick, I clenched my fists, held my breadth, and prayed that theyd block the kick. My heart sank as I saw the ball sail through the uprights, giving Auburn the 20 17 win. Silence fell over the Swamp, it was the Gators first loss of the season and I was so disappointed. It was a bright, sunny afternoon in November of 2003 and I was in the Super Gator Block in Ben Hill Griffin Stadium for the annual Florida -Florida State Game. The entire game was full of awful, one -sided calls by the ACC refs that made my blood boil. I was amazed that the refs could blow a game. To add to my frustration, Chris Rix threw a bomb of a pass down the right side of the field and the Gators lost 34 38. All I could do was scream and boo I was s o angry!
121 APPENDIX C INFORMED CONSENT Protocol Title: Understanding the effects of emotion on cognitive task performance. You are being asked to participate in a research study. Before you give your consent to volunteer, it is important that you read t he following information and ask as many questions as necessary to be sure you understand what you will be asked to do. Investigators: Melanie Mousseau, MSESS University of Florida Department of Applied Physiology and Kinesiology Christopher Janelle, Ph. D. University of Florida Department of Applied Physiology and Kinesiology Purpose of the research study: To compare how performance varies among individuals on two cognitive tasks. What you will be asked to do in the study: In the first session you w ill be asked to: complete a demographic questionnaire. You have the right to withhold any information you do not wish to share. complete a series of brief questionnaires pertaining to your attitudes, feelings, and behaviors. Please note that some of the qu estions deal with rather personal feelings, issues of well -being, and mental health. perform basic movements and answer questions regarding the movement. You should not perform any of the behaviors if you have injury or medical conditions that could make the movements painful or damaging. If you are uncertain, you should refrain or contact your primary care physician before doing the physical action. describe experiences you have had as a University of Florida football spectator/fan. provide contact infor mation so that a second, in lab session can be scheduled. In the second session, you will be asked to: complete a series of brief questionnaires pertaining to your feelings. listen to several brief narratives. complete letter and color identification tasks At the conclusion of your participation you will have the opportunity to ask any additional questions pertaining to the experiment. Time required: Session 1 30 minutes maximum Session 2 45 minutes maximum Risks and Discomforts: There are no anticipated risks or discomforts associated with your participation, however you are given the option to stop or withdraw from the study at any time
122 with no consequence. If however, as a result of any activities in the research you would like to discuss yo ur responses and reactions with a counselor, you can contact Student Counseling Services located in 301 Peabody Hall at 352.392.1575 (http://www.counsel.ufl.edu/). Benefits: Benefits include an awareness of how emotions are evoked by mental imagery. Compe nsation: Participation is voluntary and the investigators will not compensate you. However, if extra credit is offered by an instructor for you participation you may earn up to 2% extra credit. Confidentiality: Your identity will be kept confidential to th e extent provided by law; any information obtained through this testing that can be identified with you will remain confidential. Voluntary Nature of Participation: Your participation in this study is completely voluntary. Your choice of whether or not to participate will not influence your future relations with the University of Florida. If you decide to participate, you are free to withdraw your consent and to stop your participation at any time without penalty or loss of benefits to which you are entitl ed. Questions about the Study: If you have any questions about the study or results, please ask or contact: MELANIE MOUSEAU, MSESS; Doctoral Student Department of Applied Physiology and Kinesiology, 004H Yon Hall, PO Box 118205, 392.7520 x1426, mousse firstname.lastname@example.org OR CHRISTOPHER M. JANELLE, Ph.D.; Director Performance Psychology Laboratory, College of Health and Human Performance, Department of Applied Physiology and Kinesiology, Florida Gym Room 132E, PO Box 118206, 392.0584 x1270, email@example.com u Student Counseling Services: 301 Peabody Hall, 352.392.1575, http://www.counsel.ufl.edu/ Whom to Contact about your Rights as a Research Participant: UFIRB Office, Box 112250, University of Florida, Gainesville, FL 326112250; phone 392.0433 I have r ead the procedure described above. I voluntarily agree to participate in the procedure and I have printed a copy of this description.
123 LIST OF REFERENCES Beck, A, & Steer, R. (1987). Manual for the Beck Depression Inventory San Antonio, TX: Psych ological Corporation. Beck, A.T., Steer, R.A., & Carbin, M.G. (1988). Psychometric properties of the Beck Depression Inventory: Twenty -five years of evaluation. Clinical Psychology Review, 8 (1), 77100. Bernhardt, P.C., Dabbs, J.M., Fielden, J.A., & Lut ter, C.D. (1998). Testosterone changes during vicarious experiences of winning and losing among fans at sporting events. Physiology & Behavior, 65(1), 5962. Boen, F., Vanbeselaere, N., & Feys, J. (2002). Behavioral consequences of fluctuating group su ccess: An internet study of soccer team fans. The Journal of Social Psychology, 142(6), 769781. Booth, A., Shelley, G., Mazur, A., Tharp, G., & Kittok, R. (1989). Testosterone, and winning and losing in human competition. Hormones and Behavior 23(4), 556571. Bradley, B.P., Mogg, K., White, J., Groon, C., & de Bono, J. (1999). Attentional bias for emotional faces in generalized anxiety disorder. British Journal of Clinical Psychology, 38, 267278. Branscombe, N.R., & Wann, D.L. (1991). The positive social and self concept consequences of sports teams identification. Journal of Sport and Social Issues, 15(2), 115127. Branscombe, N.R, & Wann, D.L. (1992). Physiological arousal and reactions to outgroup members during competitions that implicate an important social identify. Aggressive Behavior, 18, 8593. Brown, I. D. (1994). Driver fatigue. Human Factors, 36, 298314. Bryant, R.A., & Harvey, A.G. (1997). Attentional bias in posttraumatic stress disorder. Journal of Traumatic Stress, 10(4), 63 5 644. Callow, N., Hardy, L., & Hall, C.R. (2001). The effects of a motivational general -mastery imagery intervention on the sport confidence of high-level badminton players. Research Quarterly for Exercise and Sport, 72, 389400. Carver, C.S., & White, T.L. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67(2), 319333. Cialdini, R.B., Borden, R.J., Thorne, A., Walker, M.R., Freeman, S., & Sloan, L.R. (1976). Basking in Reflected Glory: Three (football) field studies. Journal of Personality and Social Psychology, 34(3), 366375.
124 Cialdini, R.B. & Richardson, K.D. (1980). Two indirect tactics of image management: Basking and blasting. Journal of Personality and Social Psychology, 39(3), 406415. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). New York: Academic Press. Cohen, S., Kamarck, T., & Mermelstein R. (1983). A global measure o f perceived stress. Journal of Health and Social Behavior, 24(4), 385396. Coombes, S.A., Cauraugh, J.H., & Janelle, C.M. (2007). Dissociating motivational direction and affective valence. Psychological Science, 18 (11), 938440. Corbin, C.B. (1973). Amo ng spectators, trait anxiety, and coronary risk. Physician and Sports Medicine, 1 (2), 55 58. Crisp, R.J., Heuston, S., Farr, M.J., & Turner, R.N. (2007). Seeing red or feeling blue: Differentiated intergroup emotions and ingroup identification in socc er fans. Group Processes & Intergroup Relations, 10(1), 9 26. Dalakas, V., Madrigal, R., & Anderson, K.L. (2004). We are number one! The phenomenon of baskingin reflected -glory and its implications for sports marketing. In L.R.Kahle & C. Riley (Eds.). Sports Marketing and the Psychology of Marketing Communication (pp. 6779). Mahwah, NJ: Lawrence Erlbaum. Dalgleish, T., Taghavi, R., Neshat -Doost, H., Moradi, A., Canterbury, R., & Yule, W. (2003). Patterns of processing bias for emotional information across clinical disorders: A comparison of attention, memory, and prospective cognition in children and adolescents with depression, generalized anxiety, and posttraumatic stress disorder. Journal of Clinical Child and Adolescent Psychology, 32(1), 10 2 1. Davidson, R.J. (1995). Cerebral asymmetry, emotion, and affective style. In R.J. Davidson & K. Hugdahl (Eds.), Brain Asymmetry (pp.361387). Cambridge, MA: MIT Press. Derryberry, D. & Tucker, DM. (1994). Motivating the focus of attention. In P. Niede nthal and S. Kitayama (Eds.), The Hearts Eye: Emotional Influences in Perception and Attention (pp. 170196). San Diego, CA: Academic Press. Derryberry, D. & Reed, M.A. (1998). Anxiety and attentional focusing: Trait, state and hemispheric influences. Personality and Individual Differences, 25, 745761. Dimmock, J.A. & Grove, J.R. (2005). Relationship of fan identification to determinants of aggression. Journal of Applied Sport Psychology, 17, 3747. Diener, E., & Diener, C. (1996). Most people are happy. Psychological Science, 7, 181185.
125 Diener, E., Emmons, R.A., Larsen, R.J., & Griffin, S. (1985). The Satisfaction with Life Scale. Journal of Personality Assessment, 49(1), 71 75. Dietz -Uhler, B. & Murrell, A. (1999). Examining fan reactions to game outcomes: A longitudinal study of social identity. Journal of Sport Behavior, 22(1), 1527. Easterbrook, J.A. (1959). The effect of emotion on cue utilization and the organization of behavior, Psychological Review, 66, 183201. Ehrman, R.N., Robbi ns, S.J., Bromwell, M.A., Lankford, M.E., Monterosso, J.R., & OBrien, C.P. (2002). Comparing attentional bias to smoking cues in current smokers, former smokers, and non -smokers using a dot -probe task. Drug and Alcohol Dependence, 67, 185191. Elias, M. (1981). Serum cortisol, testosterone and testosterone -binding globulin responses to competitive fighting in human males. Aggressive Behavior, 7, 215224. Elsesser, K., Sartory, G., & Tackenberg, A. (2004). Attention, heart rate, and startle response d uring exposure to trauma relevant pictures: A comparison of recent trauma victims and patients with posttraumatic stress disorder. Journal of Abnormal Psychology, 113(2), 289301. Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER: A general power anal ysis program. Behavior Research Methods, Instruments, & Computers, 28, 1 11. Fink, J.S., Trail, G.T., & Anderson, D.F. (2002). An examination of team identification: Which motives are most salient to its existence? International Sports Journal, 6(2), 19 5 207. Frster, J., Friedman, R.S., zelsel, A., & Denzler, M. (2006). Enactment of approach and avoidance behavior influences the scope of perceptual and conceptual attention. Journal of Experimental Social Psychology, 42, 133146. Friedman, R.S. & Fr ster, J. (2005). The influence of approach and avoidance cues on attentional flexibility. Motivation and Emotion, 29(2), 69 81. Gable, P.A. & Harmon Jones, J. (in press). Approach -motivated positive affect reduces breadth of attention. Psychological Sci ence. Gantz, W. & Wenner, L.A. (1995). Fanship and the television sports viewing experience. Sociology of Sport Journal, 12, 5674. Gasper, K & Clore, G.L. (2002). Attending to the big picture: Mood and global versus processing of visual information. Psychological Science, 13, 3440. Groth Marnat, G. (2003). The handbook of psychological assessment (4th ed.). Hoboken, NJ: John Wiley & Sons, Inc.
126 Hall, C. & Martin, K. (1997). Measuring movement imagery abilities: A revision of the Movement Imagery Questionnaire. Journal of Mental Imagery, 21, 143154. Harmon Jones, E., Lueck, L., Fearn, M., & Harmon Jones, C. (2006). The effect of personal relevance and approach related action expectation on relative left frontal cortical activity. Psychological Science, 17 (5), 434 440. Heere, B. & James, J.D. (2007). Stepping outside the lines: Developing a multi dimensional team identity scale based on social identity theory. Sport Management Review, 10, 6591. Heider, F. (1958). The psychology of interperson al relations. New York: Wiley. Hillman, C.H., Cuthbert, B.N., Bradley, M.M., & Lang, P.J. (2004). Motivated engagement to appetitive and aversive fanship cues: Psychophysiological responses of rival sport fans. Journal of Sport & Exercise Psychology, 26 (2), 338351. Hillman, C.H., Cuthbert, B.N., Cauraugh, J., Schupp, H.T., Bradley, M.M., & Lang, P.J. (2000). Psychophysiological responses of sport fans. Motivation and Emotion, 24(1), 13 28. Hirt, E.R., Zillmann, D., Erickson, G.A., & Kennedy, C. (1 992). Costs and Benefits of Allegiance: Changes in fans self ascribed competencies after team victory versus defeat. Journal of Personality and Social Psychology, 63(5), 724 738. Isen, A. M. (1987). Positive affect, cognitive processes, and social behav ior. In L. Berkowitz (Ed.), Advances in experimental and social psychology (Vol 20, pp. 203253). New York: Academic Press. Isen, A.M. (2000). Positive affect and decision making. In M. Lewis & J. HavilandJones (Eds.), Handbook of emotions (2nd edition pp. 417435). New York: Guilford. Jacobson, E. (1930). Electrical measurement of neuromuscular states during mental activities. American Journal of Physiology, 94, 2234. Kerr, J.H., Wilson, G.V., Nakamura, I., & Sudo, Y. (2005). Emotional dynamics of soccer fans at winning and losing games. Personality and Individual Differences, 38, 18551866. King, B. (2005, March 7). The 24/7 fan. Street & Smiths SportsBusiness Journal, 2337. Kuhl, J. & Kazn, M. (1999). Volitional facilitation of difficult in tentions: Joint activation of intention memory and positive affect removes Stroop interference. Journal of Experimental Psychology: General, 128, 382 399. Kwon, H.H., & Armstrong, K.L. (2002). Factors influencing impulse buying of sport team licensed me rchandise. Sport Marketing Quarterly, 11, 151163.
127 Lang, P.J. (1985). The cognitive psychophysiology of emotion: Fear and anxiety. In A. Hussain Tuma & J. Maser (Eds.), Anxiety and the anxiety disorders (pp.131 170). Hillsdale, NJ: Erlbaum. Lang, P.J. Bradley, M.M., & Cuthbert, B.N. (1997). Motivated attention: Affect, activation, and action. In P.J. Lang, R.F. Simons, & M.T. Balaban (Eds.). Attention and Orienting: Sensory and Motivational Processes (pp. 97 135). Mahwah, NJ: Lawrence Erlbaum. L ang, P.J., Bradley, M.M., & Cuthbert, B.N. (1998). Emotion and motivation: Measuring affective perception. Journal of Clinical Neurophysiology, 15(5), 397 408. Laverie, D.A., & Arnett, D.B. (2000). Factors affecting fan attendance: The influence of ident ity salience and satisfaction. Journal of Leisure Research, 32(2), 225246. Lawther, J.D. (1951). Psychology of coaching. New York: Prentice Hall. Lefton, T. (2007, August 1319). Not just hanging around: Consumers stuck on Fatheads giant wall art. S treet & Smiths SportsBusiness Journal, 10(16), 17 19. Lim, S.L., & Kim, J.H. (2005). Cognitive processing of emotional information in depression, panic, and somatoform disorder. Journal of Abnormal Psychology, 114(1), 50 61. Liu, Y. (2002). Analyzing R M ANOVA related data using SPSS10. Measurement in Physical Education and Exercise Science, 6(1), 43 60. Lubman, D.I., Peters, L.A., Mogg, K., Bradley, B.P., & Deakin, J.F.W. (2000). Attentional bias for drug cues in opiate dependence. Psychological Medicine, 30, 169175. Luhtanen R. & Crocker, J. (1991). Self -esteem and intergroup comparison: Toward a theory of collective self -esteem. In J. Suls & T.A. Wills (Eds.), Social comparisons: Contemporary theory and research (pp.211234). Hillsdale, NJ: Erlb aum. Luhtanen R. & Crocker, J. (1992). A collective self -esteem scale: Self -evaluations of ones social identity. Personality and Social Psychology Bulletin, 18, 302318. MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emotional disorde rs. Journal of Abnormal Psychology, 95, 1520. Martens, R., Burton, D., Vealey, R.S., Bump, L.A., & Smith, D.E. (1990). Development and validation of the Competitive State Anxiety Inventory2. In R. Martens, R.S. Vealey, & D. Burton (Eds.), Competitive anxiety in sport (pp.117190). Champaign, IL: Human Kinetics. Mazur, A., Booth, A., & Dabbs, J. (1992). Testosterone and chess competition. Social Psychology Quarterly, 55, 7077.
128 McCaul, K.D., Glaude, B.A., & Joppa, M. (1992). Winning, losing and mood and testosterone. Hormone Behavior, 26, 486504. McNair, D.M., Lorr, M., & Droppleman, L.F. (1971). Manual for the Profile of Mood States. San Diego: Educational and Industrial Testing Service. McNally, R.J., Lasko, N.B., Clancy, S.A., Macklin, M.L Pitman, R.K., & Orr, S.P. (2004). Psychophysiological responding during script -driven imagery in people reporting abduction by space aliens. Psychological Science, 15(7), 493497. Miller, M.W., Patrick, C.J., & Levenston, G.K. (2002). Affective imager y and the startle response: Probing mechanisms of modulation during pleasant scenes, personal experiences, and discrete negative emotions. Psychophysiology, 39, 519529. Mogg, K., & Bradley, B. P. (1998). A cognitive -motivational analysis of anxiety. Beh aviour Research & Therapy, 36, 809848. Moradi, A.R., Taghavi, M.R., Heshat Doost, H.T., Yule, W., & Dalgleish, T. (1999). Performance of children and adolescents with PTSD on the Stroop colour -naming task. Psychological Medicine, 29(2), 425419. Mu ller, J., & Roberts, J.E. (2005). Memory and attention in Obsessive Compulsive Disorder: A review. Anxiety Disorders, 19 1 28. Musa, C., Lpine, J.P., Clark, D.M., Mansell, W., & Ehlers, A. (2003). Selective attention in social phobia and the moderatin g effect of a concurrent depressive disorder. Behaviour Research and Therapy, 41, 10431054. hman, A., Flykt, A., & Esteves, F. (2001). Emotion drives attention: Detecting the snake in the grass. Journal of Experimental Psychology: General, 130(3), 466478. hman, A., Lundqvist, D., & Esteves, F. (2001). The face in the crowd revisited: A threat advantage with schematic stimuli. Journal of Personality and Social Psychology, 80(3), 381396. Navon, D. (1977). Forest before trees: The precedence of globa l features in visual perception. Cognitive Psychology, 9, 353383. Parkinson, L., & Rachman, S. (1981). Intrusive thoughts: The effects of an uncontrived stress. Advances in Behaviour Research and Therapy, 3, 111118. Pishyar, R., Harris, L.M., & Menzi es, R.G. (2004). Attentional bias for words and faces in social anxiety. Anxiety, Stress, and Coping, 17(1), 23 36. Posner, M.I., Snyder, C.R., & Davidson, B.J. (1980). Attention and the detection of signals. Journal of Experimental Psychology: General, 109(2), 160174.
129 Quirk, K. (1997). Not now, honey, Im watching the game. New York: Fireside. Rascher, D.A. (October 2005). What is the size of the sports industry? Sports Economics: The Science Behind the Business of Sports. Rinck, M., Reinecke, A., Ellwart, T., Heuer, K., & Becker, E.S. (2005). Speeded detection and increased distraction in fear of spiders: Evidence from eye movements. Journal of Abnormal Psychology, 114(2), 235 248. Rosenberg, M. (1979). Conceiving the self. New York: Basic Book s. Russell, G.W. (2008). Aggression in the sports world: A social psychological perspective. New York: Oxford University Press. Russell, D., Peplau, L.A., & Cutrona, C.E. (1980). The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39, 472480. Scheier, M.F., Carver, C.S., & Bridges, M.W. (1994). Distinguishing optimism from neuroticism (and trait anxiety, self -mastery, and self -esteem): A reevaluation of the life -orientation test. Journal of Personality and Social Psychology, 67(6), 10631078. Schumacker, R.E., & Lomax, R.G. (2004). A beginners guide to structural equation modeling (2nd edition). New York: Psychology Press. Schutz, R.W., Smoll, F.L., & Gessaroli, M.E. (1 983). Multivariate statistics: A self rest and guide to their utilization. Research Quarterly for Exercise and Sport, 54(3), 255263. Schwarz, N., Strack, F., Kommer, D., & Wagner, D. (1987). Soccer, rooms, and the quality of your life: Mood effects on judgments of satisfaction with life in general and with specific domains. European Journal of Social Psychology, 17, 6979. Schweitzer, K., Zillmann, D., Weaver, J.B., & Luttrell, E.S. (1992). Perception of threatening events in the emotional aftermath of a televised college football game. Journal of Broadcasting & Electronic Media, 36, 75 82. Sloan, L.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 (pp.219262). Hillsdale, NJ: Lawrence Erlbaum Associates. Snyder, C.R., Higgins, R.L., & Stucky, R.J. (1983). Excuses: Masquerades in search of grace. New York: Wiley Interscience. Snyder, C.R., Lassegard, M., & Ford, C.E. (1986). Distancing after group success and failure:
130 Basking in reflected glory and cutting off reflected failure. Journal of Personality and Social Psychology, 51(2), 382388. Spielberger, C.D. (1983). State -Trait Anxiety Inventory for adults sa mpler set: Manual, test, scoring key Redwood City, CA: Mind Garden. St. John, W. (2004). Rammer jammer yellow hammer. New York: Three Rivers Press. Stirling, L.J., Eley, T.C., & Clark, D.M. (2006). Preliminary evidence for an association between soci al anxiety symptoms and avoidance of negative faces in school age children. Journal of Clinical Child and Adolescent Psychology, 35(3), 440 445. Stroop, J. R. (1935). Studies of interference in serial and verbal reactions. Journal of Experimental Psychol ogy, 18, 643662. Svebak, S., Ursin, H., Endresen, I., Hjelmen, A.M., & Apter, M.J. (1991). Psychological factors in the etiology of back pain. Psychology and Health (5), 307314. Tajfel H. (1981). Human groups and social categories. Cambridge, UK: Camb ridge University Press. Tajfel, H., & Turner, J. C. (1979). An integrative theory of social conflict. Reprinted in W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 3347). Monterey, CA: Brooks/Cole. Tata, P.R., Leib owitz, J.A., Prunty, M.J., Cameron, M., & Pickering, A.D. (1996). Attentional bias in Obsessional Compulsive Disorder. Behaviour Research and Therapy, 34(1), 53 60. Townshend, J.M., & Duka, T. (2001). Attentional bias associated with alcohol cues: Diffe rences between heavy and occasional social drinkers. Psychopharmacology, 157, 67 74. Trail, G.T., & James, J.D. (2001). The Motivation Scale for Sport Consumption: Assessment of the Scales Psychometric Properties. Journal of Sport Behavior, 24, 108127 Velasco, C. & Bond, A. (1998). Personal relevance is an important dimension for visceral reactivity in emotional imagery. Cognition and Emotion, 12(2), 231242. Wakefield, K.L., & Wann, D.L. (2006). An examination of dysfunctional sport fans: Method of classification and relationships with problem behaviors. Journal of Leisure Research, 38(2), 168186. Williams, J.M.G., Watts, F.N., MacLeod, C., & Mathews, A. (1997). Cognitive Psychology and Emotional Disorders (2nd Ed). West Sussex, England: John Wiley & Sons.
131 Wann, D.L. (1993). Aggression among highly identified spectators as a function of their need to maintain positive social identity. Journal of Sport and Social Issues, 17, 134143. Wann, D.L. (1994). The noble sports fan: The relationshi ps between team identification, self esteem, and aggression. Perceptual and Motor Skills, 78, 864866. Wann, D.L. (2002). Preliminary validation of a measure for assessing identification as a sport fan: The Sport Fandom Questionnaire. International Jour nal of Sport Management, 3, 103115. Wann, D.L. (2006a). Examining the potential causal relationships between sport team identification and psychological well -being. Journal of Sport Behavior, 29(1), 79 95. Wann, D.L. (2006b). Understanding the positive social psychological benefits of sport team identification: The Team Identification -Social Psychological Health Model. Group Dynamics: Theory, Research, and Practice, 10 (4), 272296. Wann, D.L. & Branscombe, N.R. (1990). Die -hard and fair -weather fans: Effects of identification on BIRGing and CORFing tendencies. Journal of Sport and Social Issues, 14, 103117. Wann, D.L., & Branscombe, N.R. (1992). Emotional responses to the sports page. Journal of Sport and Social Issues, 16(1), 4964. Wann, D.L., & Branscombe, N.R. (1993). Sport fans: Measuring degree of identification with their teams. International Journal of Sport Psychology, 24, 1 17. Wann, D.L., Dimmock, J.A., & Grove, J.R. (2003). Generalizing the team identificationpsychological health model to a different sport and culture: The case of Australian rules football. Group Dynamics: Theory, Research, and Practice, 7(4), 289296. Wann, D.L., Dolan, T.J., McGeorge, K.K., & Allison, J.A. (1994). Relationships between spectator identification and spectators perceptions of influence, spectators emotions, and competition outcome. Journal of Sport & Exercise Psychology, 16, 347364. Wann, D.L., Dunham, M.D., Byrd, M.L., & Keenan, B.L. (2004). The five -factor model of personality and the psyc hological health of highly identified sport fans. International Sports Journal, 8(2), 28 36. Wann, D.L., Friedman, K., McHale, M., & Jaffe, A. (2003). The Norelco sport fanatics survey: Examining behaviors of sport fans. Psychological Reports, 92, 930936. Wann, D.L., Inman, S., Ensor, C.L., Gates, R.D., & Caldwell, D.S. (1999). Assessing the psychological well -being of sport fans using the Profile of Mood States: The importance of team identification. International Sports Journal, 3, 8190.
132 Wann, D. L., Melnick, M.J., Russell, G.W., & Pease, D.G. (2001). Sport fans: The psychology and social impact of spectators. New York: Routledge. Wann, D.L., & Pierce, S. (2005). The relationship between sport team identification and social well -being: Additi onal evidence supporting the Team Identification -Social Psychological Health Model. North American Journal of Psychology, 7(1), 117124. Wann, D.L., Pierce, S., Padgett, B., Evans, A., Krill, K., & Romay, A. (2003). Relations between sport team identifi cation and optimism. Perceptual and Motor Skills, 97, 803804. Wann, D.L., Schrader, M.P., & Adamson, D.R. (1998). The cognitive and somatic anxiety of sport spectators. Journal of Sport Behavior, 21(3), 322 337. Wann, D.L., & Steele, N.L. (1998). Atten tional focus of sport spectators. Perceptual and Motor Skills, 86, 11631167. Watson, D., Clark, L.A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Soc ial Psychology, 54(6), 10631070. Williams, J.M.G., Mathews, A., & MacLeod, C. (1996). The emotional Stroop task and psychopathology. Psychological Bulletin, 120 (1), 3 24. Williams, J.M.G., Watts, F.N., MacLeod, C., & Mathews, A. (1997). Cognitive Psy chology and Emotional Disorders (2nd Ed). West Sussex, England: John Wiley & Sons. Yeomans, M.R., Javaherian, S., Tovey, H.M., & Stafford, L.D. (2005). Attentional bias for caffeine related stimuli in high but not moderate or non -caffeine consumers. Psy chopharmacology, 181, 477485.
133 BIOGRAPHICAL SKETCH Melanie B. Mousseau was born and raised in North Smithfield, RI and began her quest for higher education in 1997 at Mary Washington College located in Fredericksburg, VA, where she majored in psychology. Pursuing her interest in sport psychology, Melanie completed an honors senior thesis entitled, Home Advantage: Does Anxiety Contribute? and was awarded the distinction of outstanding senior psychology major. After receiving her BS in May 2001, Melani e continued her trek south, enrolling in the graduate program in exercise and sport sciences at the University of Florida, specializing in sport and exercise psychology under Dr. Christopher Janelle. During her tenure as a masters student Melanie worked as a tennis instructor within the Sport and Fitness Program at UF and pursued various research interests including cognitive fatigue and mental imagery in sport. Following completion of her masters degree Melanie began pursuing her Ph.D. at UF as an Alum ni fellow. While Melanie reconceptualized her line of research she taught both sport psychology and exercise psychology within the College of Health and Human Performance and actively pursued her interest in consulting, establishing the Performance Psychol ogy Group, LLC in July 2007. Following completion of her doctoral degree in 2009, Melanie plans on further developing the Performance Psychology Groups client list to include sport leagues and organizations interested in learning more about their fans, so that they can maximize the fan experience while implementing effective marketing and outreach strategies that are grounded in cutting edge research.