Journal of Sport Management, 2011, 25 70-83 © 2011 iHuman Kinetics, Inc. For Love or Money: Developing and Validating a Motivational Scale for Fantasy Football Participation Brendan Dwyer Virginia Commonwealth University Yongjae Kim Kutztown University of Pennsylvania The contemporary sport fan has the ability to consume spectator sport through several means including event attendance, television and radio broadcasts, print publications, and Internet applications. Recently, an ancillary sport service, termed fantasy sports, has become one of the most popular activities among sport fans. As a result, the business of fantasy sports is booming. This study examined motivational dimensions underlying fantasy football participation from a Uses and Gratifications perspective. Utilizing Churchill's (1979) five-step method for developing quality marketing measures, this study identified and validated three motivational dimensions: entertainment/escape, competition, and social interaction. The results suggest a pattern of fantasy football participation that is more purposeful and active than traditional media use. Discussed are the gambling associations, future research opportunities, and suggestions for developing fantasy football participation into a more creative and interactive marketing communication tool. The psychological, sociological, and behavioral study of sport spectators and consumers has been a prosperous area of study for sport management researchers over the past two decades. Specifically, the demand for sport has been the primary focus of the majority of this research, for understanding consumer motivation provides vital information for sport marketers (Sloan, 1989). Within this process, the development of measurement instruments is an important step in explaining why people attend, engage in, or watch sport events (Trail, Anderson, & Fink, 2000; Trail & James, 2001; Wann, 1995). For sport managers, this information is pertinent to provide products and services that continually meet the needs and wants of sport consumers. Currently, sport fans have the ability to consume sports through several means including event attendance, television and radio broadcasts, print publications, and the Intemet. Therefore, to fully understand fan consumption, it is necessary to investigate the numerous ways and frequency in which they consume sport. Interestingly, as the study of sport consumer behavior has evolved, the examination of certain segments of sport consumers has been overlooked, specifically, the population of mediadominant fans. While this segment of the sport population represents professional sports' most substantial fan base, a comprehensive psychographic assessment of this group of consumers is lacking (Pritchard & Funk, 2006). Within this large segment of sport fans reside fantasy sport participants. Also known as Rôtisserie and Fanalytics, fantasy sports are games in which participants act as general managers or owners of their own sport team. Completely customizable, interactive, and involving nearly every major professional sport from professional football to bass fishing, fantasy sports are primarily online services that allow fans to simultaneously follow their favorite sports and actively compete and interact with family, friends, acquaintances, or even strangers based upon real-world professional player statistics. The current study investigated fantasy football participation due to its immense popularity (> 20 million participants) and its status as the gateway activity for other fantasy sport games (Fantasy Sports Trade Association [FSTA], 2008). With respect to the process of the activity, fantasy football leagues are formed throughout the summer and teams are traditionally drafted as the National Football League (NFL) season begins. Individual NFL players are drafted based on their potential to perform on the field as fantasy team performance is determined by their statistical output in the form of touchdowns, yards from scrimmage, receptions, field goals, etc. Competition between fantasy participants typically mirrors the NFL's weekly head-to-head matchup, and as a result, participants engage in weekly sometimes even daily activities involving their team and league including adding and dropping players, trading players, and altering their starting line-ups. To perform these actions, however, access to Dwyer is with the Center for Sport Leadership, Virginia Commonwealth University, Richmond, VA. Kim is with Kutztown University of Pennsylvania, Kutztown, PA. 70 Mofiviafions for Fantasy Foofbail Participation 71 player and team information is required, and as a function of this requirement, the Intemet plays a significant role in the process. Specifically, fantasy sport websites, such as Yahoo.com, ESPN.com, and CBSsports.com, have become the primary vehicle through which fantasy owners draft players, check scores, conduct research, and perform daily and weekly functions associated with the activity. Simultaneously, as the Intemet has grown into a primary means of communication and entertainment for consumers, fantasy football have become one of the most popular activities among NFL fans. The popularity of fantasy sports, in general, is evidenced by the fact that nearly 30 million Americans and Canadians currently participate in some kind of fantasy sport league (FSTA, 2008). Many fantasy owners participate in several leagues during a single season and throughout the year. Recently, this habitual commitment has transformed fantasy sports into a highlylucrative enterprise. The FSTA currently estimates the total market impact of fantasy sports to be $4.48 billion, annually, with $800 million spent directly on fantasy sport products and services. Fantasy sport participants also represent corporate America's most highly-coveted group of consumers (Caucasian, Male, 18^5, Bachelor's Degree, and $78,000 annual income). As a result, fantasy sport has emerged as an easy, cost-effective means of reaching an engaged and lucrative group of consumers (Leporini, 2006). Despite the ever-increasing popularity of the activity and the highly-sought after marketability of its participants, there is a significant lack of empirical research studying the motivational factors behind participation. Given this limited amount of investigation, the current study was exploratory in nature. The purpose of this study was twofold: (1) examine what motivational dimensions lead to fantasy football participation, and (2) develop and validate a comprehensive fantasy football motivation scale. Specifically, this study examined fantasy football from a Uses and Gratifications (U&G) perspective to understand how fantasy football participants use this media to fulfill their unique needs and wants. Review of Literature Fantasy Sports and the i\/ledia-Dominant Sport Fan The advent of fantasy sports as an ancillary sport media service has seemingly created a new role for the traditional sport consumer (Drayer, Shapiro, Dwyer, Morse, & White, 2010). For most sport fans, it is the closest way to participate in professional sports without actually putting on a uniform and competing. A typical fantasy team is comprised of a self-selected group of heterogeneous players. The corresponding attachment attributed to these players often distributes a participant's interest throughout an entire professional league. Ergo, instead of passively following one's favorite team, a fantasy participant is given the opportunity to actively engage in operations similar to those that occur in a professional sport team's front office. Furthermore, the premise of fantasy sports allows individual participants to compete weekly against other fantasy team owners in a league-style format. This competition usually lasts the duration of a professional sport league regular season, and is dependent upon the statistical output of real-world athletic performance. Given this unique competitive nature, fantasy sport participation has created a new, highly-engaged sport consumer who craves interactivity and real-time information combined with the traditional, old-fashioned spectatorship associated with professional sports (Shipman, 2001). In general, the scholarly literature in the area of fantasy sports is limited (Lomax, 2006). Previous studies have spanned from gambling concerns associated with fantasy sports, masculinity issues, and communication (Bernhard & Eade, 2005; Davis & Duncan, 2006; Shipman, 2001). Recently, however, researchers Drayer et al. (2010) explored the consumption habits of a group of fantasy football participants with regard to NFL products and services. The authors observed unique behaviors including an increase in media consumption and a newly found gameday experience. According to the authors, these behaviors resulted from fantasy football participation which changed the participants' perception of the league. Research by Comeau (2007) and Woodward (2005) determined that the unique nature of fantasy sport participation has the potential to alter traditional sport fandom through an amplified reliance upon mass mediated sport. Consequently, fantasy sport participants are classified as an important faction of the media-dominant sport consumer population. By definition, media-dominant sport fans primarily consume spectator sports through the mediated broadcast (television, radio, or Internet) or publication (print or electronic) of the event. According to Pritchard and Funk (2006), in comparison with heavy and event-dominant sport consumers, media-dominant fans are "more likely to purchase team-related merchandise, view media advertising and promotions, and are equally involved with the sport" (p. 316). Therefore, a further understanding of this group of spectators will offer strategic insight for sport organizations, specifically when it comes to advertising and sponsorship opportunities. Given the amount of media use associated with fantasy football participation, this study used the U&G perspective as a theoretical framework to investigate fantasy sport participants use of this interactive media. A Uses and Gratifications Perspective In the U&G paradigm, it has been assumed that people are actively involved in media selection and usage to fulfill their needs and wants (Blumler & Katz, 1974). The psychological communication perspective was established in an attempt to understand what individuals do with mass media, rather than what the media does to an individual (Katz, Blumler, & Gurevitch, 1973; Rubin, 72 Dwyer and Kim 1994). The primary objectives of the audience-centered approach are to explain the social and psychological needs that motivate individuals to use media, and to uncover the origins of the needs leading to different pattems of media consumption for gratification (Katz et al., 1973; Rubin, 1994). With the emphasis on the role of audience initiative and activity, the U&G paradigm provides a theoretical approach to examine audiences' psychological processes while using media in particular contexts (Eighmey & McCord, 1998; Papacharissi & Rubin, 2000; Ruggiero, 2000). Given the recent developments of the electronic information age, the notion of audience activity has become an important construct for conducting audience analysis (McQuail, 1994; Ruggiero, 2000). If audiences are aware of not only their desires but also how to gratify their psychological needs by selecting appropriate media (Katz et al., 1973; Rubin, 1994), then perhaps they will take the initiative in selecting and using media content from a number of alternatives (e.g., television) to best satisfy their needs and desires (Ferguson & Perse, 2000; Katz et al., 1973; Rubin, 1994). This activity is directed by prior motivation, interests and preferences, and involvement of users with media (Blumler, 1979; Levy & Windahl, 1984). In addition, this audience activity is especially crucial when examining what leads to an individual's choice for the type of media and the continuing use of a medium and its content to satisfy specific needs (Swanson, 1987). In addition, audience activity research provides insight into a consumer's choice with regard to a given source within a specific medium (Lin, 1993). For instance, the psychological need for social interaction may drive an individual toward a social networking website (e.g., Facebook orYouTube) as opposed to a content provider such as CNN.com or ESPN.com (Haridakis & Hanson, 2009). From the U&G perspective, sport fans are often motivated to participate in activities and select content that will satisfy a specific need. With the advancement of interactive and realistic media technology, sport fans, for example, make use of sport video games to satisfy their needs and wants which cannot be fulfilled in a real life sport context (Kim & Ross, 2006; Kim, Walsh, & Ross, 2008). A study by Kim and Ross demonstrated that sport fans play an active role in media selection and usage. Furthermore, with the explosive growth of the Internet, online technology makes it easier to access information related to sports, and communicate with their peer groups and like-minded sport fans. Sport fans are going online to access information and statistics about their team and players, read expert opinions on fantasy matchups, watch game highlights, and follow their teams and players (Consumer Electronics Association, 2008; Seo & Green, 2008). In an online community, sport fans communicate with their peer groups and like-minded sport fans, representing themselves as a member of their favorite sport teams on team message boards. Likewise, fantasy football participants log-on to fantasy-specific websites as owners of their fantasy teams. From there. they communicate and connect with other league members/competitors via chat rooms and league message boards. As such, sport fans' media consumption patterns in an electronic environment are more purposeful and active than traditional media uses. For sport fans, fantasy football participation is perhaps an additional way to enjoy their favorite sports, teams, or players. In a virtual environment, the possibility to create, manage, and run a fantasy team through a virtual draft enables participants to feel their initial emotional experience, much like real sport owners can in professional sports rather than in front of their televisions or on the Internet. In particular, these highly interactive features of fantasy sport websites enable fantasy owners to mentally and emotionally engage with the chosen players for their own fantasy team (Grigorovici & Constantin, 2004; Levy & Windahl, 1984; Rubin & Perse, 1987; Sundar & Kim, 2005; Swanson, 1987). Furthermore, participating in fantasy sport leagues perhaps enables users to build new social relations with like-minded users and their peer group. Like real-life sport participation, the social context of fantasy sport leagues provides individuals with opportunities to participate in dynamic social interaction among sport fans. In all, a focus on participation in fantasy football leagues from a U&G point of view may aid researchers and practitioners in understanding the motivations and characteristics of these users. From the U&G perspective, a number of studies have investigated audiences' motivation and decision to use a certain type of media as a new technology moves into the stage of mass communication (Elliott & Rossenberg, 1987). However, there have only been a few empirical endeavors to examine psychological and behavioral aspects of fantasy sport participants. Sport Motivations Previous sport spectator research has suggested that different motives may drive different consumer segments and different sport activities (Funk, Mahony, & Ridinger, 2002). In 1995, Wann developed the first comprehensive and psychometrically-sound sport fan motivation instrument termed the Sport Fan Motivation Scale (SFMS). The SFMS focused on eight categories of sport fandom: eustress (i.e., arousal and stress relief), self-esteem or personal enhancement, escape or diversion, entertainment, economic factors (i.e., gambling), aesthetic value (i.e., artistic characteristics), group affiliation, and family affiliation. More recently. Trail, Anderson, and Fink (2000) specified additional factors that contribute to sport consumption: motives, identification, event expectations, disconfirmation or confirmation of those expectations, self-esteem behaviors, and the individual's affective reactions. Trail and James (2001 ) then extended this work and developed the Motivational Scale for Sport Consumption. This scale identified the following nine motives for following sports: vicarious achievement, acquisition of knowledge, aesthetics, social interaction, excitement, escape, family, physical attractiveness of the players, and Motiviations for Fantasy Football Participation 73 the skill exhibited by the players. Despite being developed for sport spectator consumption, many of the characteristics that make up these well-researched motivations can be found in fantasy football participation. Specifically, the motives dealing with vicarious achievement, acquisition of knowledge, social interaction, excitement, and escape may help explain why individuals choose to participate. With respect to sport participation, several studies have investigated the psychological needs satisfied through sport activities. According to a study by Pelletier et al. (1995) in which the Sport Motivation Scale (SMS) was developed, sport behaviors can be directed and prompted by extrinsic motivation as well as intrinsic motivation. Extrinsic motivation refers to engaging in an activity to gain material rewards or avoid external constraints, while intrinsic motivation refers to engaging in an activity purely for the pleasure and satisfaction derived from doing the activity (Deci, 1975). Extrinsically motivated behavior is typified by participation in an activity prompted by such external forces as rewards or imposed pressure on the individual. While the majority of previous motivation research has focused on intrinsic qualities, the SMS offers a unique framework for understanding both intrinsic and extrinsic motives. With regard to the current study, extrinsic motivation may be evident when people engage in fantasy football for external reward in the form of compensation. Once again, despite this confluence of well-researched sport consumer motives, there is a significant knowledge gap with respect to the unique motivations of fantasy sport participants. Motivational Dimensions of Fantasy Sport Participation. A study by the Sports and Entertainment Academy at Indiana University (2000) was the first to explore motivations for fantasy sport league participation and demographic characteristics of fantasy sport owners. In this study, 90% of respondents indicated friendship as the primary reason for playing fantasy sports, and almost 60% of respondents cited fun as the second most important reason, followed by the thrill of competition and passing time. With regard to gender differences in fantasy sport participation, the findings supported the assertion of previous research that fantasy sport owners are predominantly male (FSTA, 2008). However, the results of this study should not be generalized to a larger population due to the small sample size and the lack of a rigorous statistical analysis. More recently, in an attempt to determine a typology of fantasy sport participants, Farquhar and Meeds (2007) identified a set of common underlying motivational dimensions for fantasy sport league participation. Employing a Q-Methodology, the study found the following five primary motives for fantasy sport participation: surveillance, arousal, entertainment, escape, and social interaction. Further, the study indicated that the perceived gratifications of arousal and surveillance led to differences among types of fantasy sport users. The primary difference between the two groups was the way in which they valued the element of chance in fantasy sports. For instance, individuals driven by surveillance, or the need for statistical information, believed fantasy sports were games of skill, while those driven by arousal, or the need to compete, saw them as games of chance. Interestingly, according to Farquhar and Meeds, highlyinvolved participants believed they "get more out of fantasy sports when they put in more time and money" (p. 1217). Thefindingsprovided valuable insight into the use of fantasy sports and the behavior of sport consumers in the virtual environment. It also suggested that participation in fantasy sports is driven by intrinsic motives for learning and understanding something new, experiencing fun and excitement, and obtaining achievement through competition against other players in the fantasy league. However, this study did not verify the underlying factor structure of the motivational dimensions, and further, was limited to intrinsic motivations identified to explain the patterns of media use and sport consumption in the absence of external rewards or constraints. A potentially interesting extrinsic motivation associated with fantasy sport participation is gambling. Since its inception in the mid-1950s, fantasy sport leagues have been associated with sport wagering. Given that sport performance is ultimately unpredictable and league winners are traditionally compensated via league entry fees, it is understandable how these gambling associations were derived. In 2005, Bernhard and Eade explored the similarities between traditional sport gambling and fantasy sport participation. Specifically, the authors concentrated on the observation of fantasy baseball participants. From an economic point of view, the authors concluded that the two activities have several similarities including an initial investment to participate and the potential to win money. In addition, comparisons were drawn between the amount of research performed by participants and the value of luck versus skill in both ventures. In all, Bernard and Eade provided a preliminary investigation into the positive and negative aspects of fantasy baseball in comparison with other forms of gambling. While they suggested that the two cultures are similar, they concluded fantasy sport participants do not exhibit the serious forms of pathological gamblers, and therefore, the activity is more like "gambling's cousin, the stock market" (p. 35). With the continued growth of the Internet and the associated growth of fantasy sport leagues, a more complete understanding of this distinct group of consumers is needed among researchers and marketers. Although previous research has some limitations, the findings of previous studies provide insight into the understanding of possible motivations for fantasy football participation and point to a "gratification" approach that can serve as a starting point in understanding audience experiences with fantasy football leagues. In an attempt to add to the knowledge base of sport consumer behavior, this study sought to develop and validate a Motivational Scale for Fantasy Football Participation (MSFFP). The following section presents the methods employed to satisfy this purpose. 74 Dwyer and Kim Methods Using Churchill's (1979) suggested procedure for developing effective marketing constructs, the current study implemented the following five-step procedure for developing and validating an instrument for measuring motives of fantasy sport participation; 1) specification of the construct domain, 2) generation of sample items, 3) data collection, 4) purification of the measure, and 5) assessment of instrument reliability and validity. The following sections provide a detailed discussion of the instrument development procedure. Construct Specification and Item Generation Following a review of fantasy sport motivations, spectator sport motivations, and media behavior literature, the unique attributes and behaviors associated with fantasy sport participation, such as diversion, vicarious achievement, escape, competition, entertainment, and statistical gratification, were included to properly specify the following domain; motives for fantasy football participation. Specifically, the construct of gambling was heavily incorporated due to the aforementioned association with the activity. In all, a semistructured interview guide was composed to lead focus group discussions. The generation of sample items was the second step in the scale development process. Focus group interviews were conducted to identify the most common reasons for participating in fantasy football. For this purpose, focus group sessions were conducted to elicit an open discussion of the construct defined above (Churchill, 1995). Based on focus group results, the authors developed a series of statements provided by the participants as reasoning for fantasy football participation. Two independent content evaluators were then responsible for the cross-reference of the interview transcripts and item generation. That is, the individual statements developed by the authors were validated and parsed down into survey items with a seven-point Likert-type scale ranging from 1 = Strongly Disagree to 7 = Strongly Agree and randomly placed within a questionnaire. A pilot study was then performed on a convenient sample to provide additional content validation of the proposed instrument before being applied to the next procedure. Data Collection, iVIeasurement Purification, and Instrument Validation To determine the number of dimensions and obtain a more complete understanding of the underlying structure of the data, a principal component analysis (PCA) with promax rotation was conducted. The results of the PCA were then analyzed by experts to refine and edit the items to provide parsimony. The experts also theoretically critiqued the PCA-generated dimensions to ensure the factors accurately measured the specified domain. Next, to verify the underlying factor structure, the refined survey was administered to a second sample of fantasy football participants on two Internet sites; ESPN.com andYahoo.com. These sites were chosen because of their large audiences and high levels of interaction between members. Once potential respondents had accessed the hyperlink address to the online survey questionnaire, they simply entered their date of birth and clicked the submit button at the end of the page. Individuals under the age of 18 were thanked for their interest but prohibited from taking the survey. The information from the completed surveys was entered into a secure database for further analysis. A confirmatory factor analysis (CFA) was then performed to verify the underlying factor structure of the fantasy football motivation scale developed in this study. Cronbach's Alpha coefficient and interitem correlations were analyzed to assess the reliability of the scale scores. The Average Variance Extracted (AVE) score for each dimension was interpreted to ensure convergent validity. Discriminant validity was evaluated through the application of Fornell and Larcker's (1981) AVE test. T'o assess criterion-related (predictive) validity, the scores for each of the MSFFP dimensions were regressed simultaneously on to three measures of fantasy football participation selected due to both theoretical and practical relevance (DeVellis, 1991). The three participation measures were; ( 1) total number of fantasy football teams managed per year, (2) approximate number of hours spent following their fantasy football team per week, and (3) a self-reported level of fantasy football competitiveness. Results Three focus groups were conducted with a total of 23 individuals. Participants were Caucasian males between the ages of 24-35. The individuals were selected by their willingness to participate and their level of involvement ranged from avid to recreational. Analysis of the focus group sessions resulted in 81 initial statements provided by the participants. After interpreting and reviewing the focus group transcripts and each statement, two content evaluators retained 27 statements viewed as potential discriminators. The 27 statements were then developed into items for the initial questionnaire. To adequately test the instrument before the next stage of implementation, a pilot study was administered to 57 graduate students (male = 49, female = 8) enrolled in a midsized public institution in the Western United States. The questionnaire was scrutinized for content validity, and correspondingly, simple reordering occurred. Overall, it was deemed an adequate and usable questionnaire. A PCA with promax rotation was conducted using a convenient sample of 345 undergraduate students at a midsized Western university and a large Midwestern university of which 235 web-based surveys were completed and deemed usable. The 235 respondents were mostly male (88.5%), Caucasian (73.6%), and between the age of 18 and 36 (mean = 23; median = 22). A variety of criteria was then used to decide on an appropriate number iVIotiviations for Fantasy Footbali Participation 75 of factors to retain (Tabachnick & Fidell, 2007). ScreePlot test and the Kaiser criterion, which considers all eigenvalues greater than one as common factors (Kaiser, 1970), suggested four dimensions. In conjunction with these results, a model was chosen that produced the most readily interpretable and theoretically sensible pattern of results. Accordingly, a four factor solution containing 20 variables was accepted as the most appropriate. The resulting model explained 63.2% of the variance, and the primary factor loadings from the pattern matrix for the 20 items ranged from .627 to .869. The factors identified were competition (4 items; eigenvalue = 1.521 ), social interaction (5 items; eigenvalue = 3.476), entertainment/ escape (5 items; eigenvalue - 2.337), and gambling (6 items; eigenvalue = 5.301). The operational definitions of each factor are presented in Table 1. Three independent experts then scrutinized the PCA results to purify the measure. The experts suggested the rewording of one item to eliminate possible confusion, and the removal of two others due to high interitem correlations. In addition, one expert, a measurement specialist, recommended combining two of the items due to similar wording and a high correlation. Based on the feedback from experts, a four factor scale with 17 items was developed for the next stage of the study. A CFA using LISREL 8.8 was performed on an additional 201 fantasy football participants drawn from two Internet sites. The aim of this procedure was to validate the subscales measuring the four factors of motives for fantasy football participation. The demographic profile of the sample is presented in Table 2, and the primary factor loadings and item wording are offered in Table 3. Table 1 Operational Definitions of the 4 Motivational Factors for Fantasy Sport Participation Factor Definition Gambling Social Interaction Competition Entertainment/Escape The gambling motive represents participation driven by, or focused upon, the chance of winning money. The social interaction motive involves establishing, preserving, or renewing relationships with family, friends, and/or coworkers. The competition motive embodies participation not only to win each week, but driven by the need to be the best in the eyes of one's competitors. The entertainment/escape motive represents participation for pure enjoyment of the activity as well as the diversion it provides from one's day-to-day activities. Table 2 Demographic of Fantasy Football Participants—Robust IVIaximum Likelihood CFA Sample N 201 (100)% Age SEX Male Female ETHNICITY Asian/Pacific Islander Black Caucasian/White Hispanic/Latino Multiracial Other Would rather not say CURRENTLY PLAYING Yes No LEVEL OF COMPETITIVENESS Very Competitive Competitive Recreational 187(93.1%) 14 (6.9%) 14 (6.9%) 4 (2.0%) 163(80.7%) 9 (4.6%) 4(2.1%) 3(1.6%) 4(2.1%) 201 (100%) 0 (0%) 122(60.7%) 66 (32.8%) 13(6.5%) Average Median Standard Deviation Range NUMBER OF YEARS PLAYED Average Median Standard Deviation Range NUMBER OF TEAMS MANAGED Average Median Standard Deviation Range WILL YOU PLAY NEXT YEAR? Yes No Maybe 28.70 25 9.50 18-68 3.62 2 3.70 1-25 3.49 2 4.07 1-30 192(95.5%) 0 (0%) 9 (4.5) 76 Dwyer and Kim Table 3 Factors, Adjusted Items, and Primary Factor Loadings for the MSFFP—Robust Maximum Likelihood CFA Factor Factors and Items Loading* Social Interaction (4 items) Sample mean = 4.54; Std. deviation = 1.31 Playing fantasy football provides an excellent opportunity to get together with, or stay in contact with, my family and friends. (SOCl) .741 One of the main reasons I play fantasy football is that doing so allows me to belong to a group of my peers. (S0C2) .778 An important reason for playing fantasy football is the ability it gives me to interact my coworkers, friends, family, and/or significant other. (S0C3) .730 Interacting with other fantasy football participants is important to me. (S0C4) .686 Gambling (5 items) Sample mean = 3.08; Std. deviation = 1.83 The amount of money wagered determines how much I follow fantasy football team. (GAMl) .901 To me, fantasy football is just another way to bet on professional football. (GAM2) .968 I play fantasy football to win money. (GAM3) .693 Given the opportunity, I would prefer to wager money on fantasy football than play at no cost. (GAM4) .591 Playing fantasy football is only enjoyable if you can bet on the outcome. (GAM5) .625 Competition (4 items) Sample mean = 5,51; Std. deviation = 1.09 I like to play fantasy football to prove to my fellow competitors that I am the best. (COMÍ) .851 When playing fantasy football, it is important to me to compare my skills with my competitors. (COM3) .762 It is important to me to win my fantasy football league(s). (COM3) .709 An important reason for playing fantasy football is the opportunity it provides to compare my unique knowledge about NFL players and teams with my competitors. (COM4) .650 Entertainment/Escape (4 items) Sample mean = 6.06; Std. deviation = 0.89 I play fantasy football because it makes watching NFL football more enjoyable. (ENTl) .602 Playing fantasy football has provided an excellent opportunity to enjoy the performance of NFL players who are not on my favorite NFL team(s). (ENT2) .637 I play fantasy football because it is a fun way to spend my time. (ENT3) .773 I play fantasy football because it provides an entertaining escape from my day-to-day activities. (ENT4) .804 * Primary factor loading from Pattern Matrix Of the 1,314 individuals who viewed the initial postings, 384 began the survey. Of these, 42 were excluded from analyses because they indicated that they were less than 18 years old, and 141 were excluded because they failed to complete the questionnaire. The 201 respondents remaining represented a 15.3% completion rate. To ensure soundness of the data, Mardia's Statistic was interpreted. As a result, the data showed adequate to good multivariate normality with a Mardia's PK score of 1.991 (Tabachnick & Fidell, 2007). Most factors were significantly correlated with each other (see Table 4), with the exception of entertainment/escape and gambling. In addition, the correlation matrix of individual items (Table 5) portrays the strong relationships within factors and the occasional significant correlation outside of a factor. As Hu and Bentler (1999) suggested, the following fit indices and their cutoff criteria were used to assess the overall fit of the model: the Satorra-Bentler scaled chisquare statistic (X^), the Non-Normed Fit Index (NNFI), the Comparative Fit Index (CFI), the Standardized Root Mean Square Residual (SRMR), and the Root Mean Table 4 Correlations Among Factors Factor Social Interaction Competition Entertainment/Escape Gambling Social Interaction 1.00 .378** .433** .190** * Coefficients are statistically significant at p < .05 ** Coefficients are statistically significant at p < .01 Square Error of Approximation (RMSEA). First, to verify the global fit of the data to the model, a chi-square statistic was analyzed. The X^ value (151.30) was statistically significant at p < .05 {df= 111), and the X^/degrees of freedom ratio (1.36) was greater than one. While the global fit was significant, several factors including sample size and nonnormality may substantially affect the X^ statistic score (Curran, West, & Finch, 1996). Therefore, a variety of other comparative indices (0 Item O) c0 w c o c CM s c CO o u U) a) U) CO o O O ÜM1 18** in •ît 84*149* 35** in * 00* in 106 8 •if- 16*098 10006? tu o a (N en ^ ^ ^ O O O O O O 00 &n 00 00 CJ u os _; S O O o t-~ oo •* 8 , as (NI oo — o — 00 o r-~ Tí o o S § § g —^o ' q -.061 os P -.013 p oo p -.079 Os p P -.095 p o p .048 zzl' o en O -.031 en S1 es p .066 en p Os 00 o .059 00 in * »n * * q * * om »n * * os O p 562** 1 1.00 o •* 00 _^ — m 00 q O O O en in "^ ^ t^" "^ - _ in vo — en m 00 so u-, en — ~ en O O q O Tt oo — vo — — O O o\ en 00 O\ O O vo os — ^ 00 oo <^ s = s s 8 s q f^ * o ^ 5 S 00 r^ es vo vo « vo — — Z. '^, o o