The impact of game outcomes on fantasy football participation and National Football League media consumption.

Dwyer, Brendan

Introduction

With more than 32 million participants, the activity of fantasy sports has become a popular endeavor for the contemporary sport fan (Fantasy Sports Trade Association [FSTA], 2011). Fantasy football, in particular, has garnered the most attention as an intoxicating complement to traditional National Football League (NFL) fandom. Defined as an ancillary sport media service wherein individual participants compete weekly in an online environment based on the statistical output of real-world NFL players, the activity has blossomed from a niche vocation into a pop-culture phenomenon. As a result, sport industry practitioners and researchers have focused their inquiry on the evolving habits associated with this activity as a means to better understand this highly-engaged group of sport consumers.

Recent research has suggested fantasy sport participants tend to spend more time engaged in professional sports, whether online or through television viewership, and have even been found to spend more money attending sporting events and purchasing merchandise (Drayer, Shapiro, Dwyer, Morse, & White, 2010; FSTA, 2008; Karg & McDonald, 2011; Nesbitt & King, 2010a; Nesbitt & King, 2010b). Additionally, a 2008 study by Ipsos Public Affairs as cited by Fisher (2008) found that fantasy sports participants not only out-consume the general public, but also other non-participating sport fans, in the leading product categories.

As an activity, fantasy sports participation operates in a virtual world and fosters a larger scope of league interest, as a participant may own/manage players on several teams throughout the league (Dwyer, 2011b). Thus, fantasy sports rely heavily on sport media consumption. Streaming scores and statistics, real-time pre-game news and analysis, and league-wide game access from products such as DirecTV's NFL Sunday Ticket and Major League Baseball's MLB.TV are highly sought after media components for this group of sport fans. As a result, Internet use, television viewership (live games and other programming), and cell phone use were reported as important modes of media consumption (Drayer et al., 2010; Dwyer & Drayer, 2010). From a theoretical perspective, researchers have also studied participant motives, involvement level, locus of control, and team loyalty (cf., Dwyer, 2011a; Dwyer & Kim, 2011; Farquhar & Meeds, 2007; Kwak, Lim, Lee, & Mahan, 2010). However, to firmly understand the behaviors associated with an experiential phenomenon (i.e., fantasy sports participation and NFL media consumption), a foundational approach is recommended (Fazio & Zanna, 1978). Therefore, this study examined the congruency between a participant's attitudes and behavior.

Attitude-Behavior Relationship

The attitude-behavior relationship (A-BR) framework was developed by Fazio, Powell, and Herr (1983) to help understand the influence attitudes have on behavior. Defined within the framework as "an association between a given object and a given evaluation" (Fazio et al., 1983, p. 724), an attitude has been established for decades as the fundamental antecedent to behavior (Ajzen & Fishbein, 1977). The A-BR framework suggests a positive attitude toward a product leads to increased consumption and a negative or non-attitude leads to decreased or non-consumption. Fazio, Powell, and Williams (1989) tested the framework on consumers with food products, and it was determined that individuals with stronger attitudes toward products showed greater attitude-behavior consistency.

[FIGURE 1 OMITTED]

Accepted as a foundational construct in marketing, advertising, and consumer psychology (Foxall, 1990), the A-BR framework has also been extended to sport fandom (Mahony & Howard, 1998; Mahony & Moorman, 1999; 2000). The framework was teamed with Zillman, Bryant, and Sapolsky's (1989) disposition theory of sportfanship which suggests sport fans derive enjoyment from watching their favorite team succeed and watching a disliked team fail. As a result, it was found both strong positive and strong negative attitudes toward a sport team led to increased spectator sport viewership (Mahony & Howard, 1998; Mahony & Moorman, 1999). In other words, television viewership of professional sport teams was not limited to one's favorite team, as one's rival team also spurred consumption. Non-viewership was a result of a neutral or non-attitude toward a team. Due to these unexpected findings, spectator sport has emerged as a captivating context for the application of the A-BR framework.

Fantasy sport, in particular, has been a growing context for application of the framework, as this segment of sport fans have displayed over-emotional, illogical, and oftentimes irrational patterns of behavior (Drayer et al., 2010; Dwyer, 2011b; Karg & McDonald, 2011). Thus, guided by the work of Fazio et al. (1983), Drayer et al. (2010) qualitatively developed and proposed a conceptual framework to explain the relationship between fantasy football and National Football League (NFL) consumption. The authors provided the following propositions to explain the phenomenon: (1) the activation of attitudes facilitated by fantasy football participation created new perceptions of the NFL; (2) these redefined perceptions broadened media consumption, and (3) one's perceptions and media consumption of the NFL were constantly changing based on NFL game outcomes related to one's fantasy and favorite NFL team. In all, the researchers concluded that fantasy participation created new attitudes (attraction to players & interactivity) and altered normative attitudes (team identification & loyalty) which led to amplified media consumption habits (TV, internet, & cell phone).

In one follow-up study, Karg and McDonald (2011) surveyed a large sample of Australian Football League (AFL) fans, fantasy participating and non-participating, to compare the attitudes and behaviors associated with AFL fandom. The results suggested fantasy participants out consumed non-participants in the areas of game attendance, TV viewership and other forms of media consumption, merchandise purchases, and sport gambling activities, while favorite team attachment remained similar for both groups. Dwyer (2011b) quantitatively tested and confirmed Drayer et al.'s (2010) propositions #1 and #2 by examining fantasy football involvement in association with intentions to watch televised NFL games. The author found that fantasy participation duplicated a sports fan's dispositional A-BR with sport teams in that a participant was attracted to and consumed one's own fantasy players and one's opponent's fantasy players in addition to one's favorite NFL team and their rival. In all, media consumption habits increased significantly as a result of this widened scope of interest.

The A-BR framework is not without critics. For instance, Azjen and Fishbein (2000) revisited the automaticity of attitude activation and found attitudes to be much less stable than previously hypothesized. Herr (1995) argued the consistency between attitude and behavior may have less to do with a hypothesized relationship and more to do with shared method variance. This criticism has also been extended to sport-consumer behavior. In the process of developing a valid and reliable measure for sport consumer attitudes toward corporate social responsibility, researchers Walker and Heere (2001) found that a non-linear relationship between attitudes (awareness & affect) and behavior, as prescribed in the A-BR framework, resulted in a more valid and reliable model. In all, much debate centers on how attitudes influence behavior and it remains essential that researchers continue to empirically test the evolving relationship.

Identification, Attachment, and Attraction

Defined by Callero (1985) as the process of linking an individual's experience and internal information processing to social networks created through interactions with others, identification with sport, more specifically sport teams, is a well-researched area. Underpinned in Tajfel and Turner's (1986) social identity theory, previous studies have found evidence that identification is an influential predictor of sport consumption (Madrigal, 1995; Trail, Anderson, & Fink, 2000, 2005; Trail, Fink, & Anderson, 2003; Wann & Branscombe, 1993). Trail et al. (2000) defined identification as "an orientation of the self in regard to other objects including a person or group that results in feeling or sentiments of close attachment" (p. 165-166).

From this definition, Trail, Robinson, Dick, and Gillentine (2003) examined sport spectator motives in association with distinct points of attachment. Robinson and Trail (2005) extended this line of research by developing and validating a Points of Attachment Index (PAI) with the following underlying fan connection points: team, player, coach, sport, community, and level of play. The authors concluded the different forms of identification influenced consumption behavior distinctively and supported measuring identification by distinct categories. Researchers Kwon, Trail, and Anderson (2006) also found that various points of attachment influenced future consumption intentions to attend games and purchase merchandise differently.

Extended to fantasy sports, the aforementioned Karg and McDonald (2011) study found evidence of unique attitudes and differing points of attachment for fans who participate in fantasy sport compared to non-fantasy participating Australian Rules Football fans. Shapiro, Drayer, and Dwyer (2011) examined differences in points of attachment and media consumption related to Major League Baseball (MLB) and fantasy baseball. The results suggested attachment to one's favorite MLB team, community, and sport appear to be the most ubiquitous connection points for fantasy baseball participants, in that these points were found to be significant across varying forms of fantasy participation and MLB consumption.

In 2001, Funk and James developed the Psychological Continuum Model (PCM) to conceptually guide how spectators psychologically move from awareness of a sport/team to allegiance to that sport/team (Funk & James, 2001). The PCM includes four stages with specific attributes and applications for individuals within each phase. Awareness and attraction are the first two stages. Each includes a lower-level connection to the sport object, but each lacks the psychological commitment present at the higher levels, attachment and allegiance. According to the authors, movement along the PCM toward fan allegiance provides outcomes that are more durable and impactful for individual sport organizations.

Given the short duration of fantasy football player ownership (typically one year), Drayer et al. (2010) proposed that the psychological connection with one's fantasy players rarely reaches the attachment level. However, the connection during this short span is highly-interactive; thus, the authors suggested it goes beyond mere awareness. As a result, this study utilized one's attraction to his/her fantasy football players in conjunction with media consumption. Conceptualized as the stage on the continuum where spectators connect for hedonic motives and social situational factors, attraction is the level where attitudes form (Funk & James, 2001; 2006). To measure attraction, the authors prescribed the application of the attraction facet of the leisure involvement construct (cf., Havitz & Dimanche, 1997; Iwaski & Havitz, 1998; Laurent & Kapferer, 1985).

The purpose of the current study was to assess Drayer et al.'s (2010) proposition #3 by examining one's attitudinal and behavioral changes toward the NFL with twelve weeks of fantasy and favorite team outcomes serving as an extraneous treatment. Specifically, attraction to fantasy players was tested as the newly acquired attitude and team identification in the form of attachment was included as the normative attitude. This study utilized a pre-post research design to answer the following research questions:

RQ1: How do game outcomes impact attraction to fantasy players?

RQ2: How do game outcomes impact NFL team attachment?

RQ3: How do game outcomes impact fantasy team media consumption?

RQ4: How do game outcomes impact NFL team media consumption?

RQ5: How do game outcomes impact general NFL media consumption?

Method

Participants

This study targeted FSTA subscribers through an online survey protocol. The FSTA is a conglomerate of 120 companies ranging in size and scope. The group serves somewhere between five and seven million fantasy sports consumers. To assess a baseline of attitudes and behavioral intentions prior to the NFL season and the same variables after NFL and fantasy-related outcomes, data collection occurred in two phases. The initial email solicitation for participation was sent to a panel of 1,000 email addresses provided by the FSTA. The email was sent during the third week of August 2011. Two follow-up emails were sent to remind potential respondents at one week and three weeks. Data collection for phase 1 ended on September 7th, 2011, so that no regular-season game results could interfere with this phase of the study. The response rate for phase 1 was 31.7%, as 317 respondents started the online questionnaire but only 302 surveys were deemed useable.

Phase 2 of the study was conducted after week 12 of the NFL regular season. Week 12 was selected as it provided the longest amount of time from the start of the regular season but before the fantasy football playoffs which typically begin in week 14. To contact phase 1 participants again, respondents were asked to provide their email addresses as consent to participate in phase 2 of the study. Phase 2 emails were sent in two iterations to remind potential respondents of the survey. The first was sent on November 28, 2011. Response rate for phase 2 was 78.1% as 236 participants completed the phase 2 questionnaire. Sample demographic and descriptive statistics are provided in Table 1. It is important note as a limitation to this study that fantasy sports participants are highly engaged sport fans. While this segment is growing, the generalizability of this study's results will need to be tempered as the consumption habits, team attachments, and viewership behavior may be at higher levels than the general sport fan population.

Instrumentation

The survey questionnaire was hosted online by Formsite.com. Phase 1 included 34 questions including demographic and descriptive items. Phase 2 was limited to 26 items. The study's design was one in which each respondent was pre-tested and post-tested on the following dependent variables: Attachment to NFL Team, Attraction to Fantasy Players, and media consumption intentions (team, fantasy, & general NFL). What follows is a brief summary of each scale.

Attachment Items--To measure participant identification, the current study employed the team-specific section of Trail, Robinson, Dick, and Gillentine's (2003) Points of Attachment Index which segments identification based on a fan's connection with his/her favorite NFL team. This point of attachment was chosen based on the conceptual work of Drayer et al. (2010) which posited team identification as an important normative attitude for fantasy football participants. The scale contained three items measured on a seven point Likert-type scale.

Attraction Items--To measure attraction, three items were developed based on the suggestion by Drayer et al. (2010) and the PCM work of Funk and James (2001). Five leisure attraction items were adapted to the activity of fantasy football, pilot-tested, panel-tested, and refined before utilized. Ultimately, the scores of a three-item scale showed evidence of good reliability and validity (see Table 2). Each item was measured on a seven point Likert-type scale.

Media Consumption Items--Nine total items were used to measure changes in media consumption. To assess fantasy-related and favorite NFL team-related consumption, participants were asked the likelihood of following one's fantasy and favorite team via the internet, television, and cell phone. Lastly, intentions to watch Monday Night Football, pre-game shows, and post-game shows were asked to assess general NFL-related media consumption. These items were chosen to reflect general NFL media consumption based on the results of Drayer et al.'s (2010) conceptual framework and previous media consumption studies related to the NFL (Fortunato, 2004; Oates, 2009). Each item was measured on an eleven point Juster scale.

[FIGURE 2 OMITTED]

To formulate the subject variable, participants were categorized by the level of success/failure experienced during the first twelve weeks of NFL regular season. During Phase 2 of the survey, the respondents were asked to identify their favorite NFL team and the win-loss record from their most preferred fantasy football league. The sample was then split into four groups based on the winning percentage of the NFL team and the fantasy football record identified (see Figure 2). For consistency reasons, a fantasy team with a .500 record was considered a winning record, and based on the selection of week 12 as the report date, no NFL team had a .500 record. Group 1 (BF) experienced fantasy and favorite team failure while Group 4 (NF) experienced success with both teams. Group 2 (FF) experienced favorite team success, but fantasy failure, and Group 3 (TF) experienced the opposite. As an example, if a participant was a Green Bay Packers fan (11-0) and their fantasy football team was 3-9, the participant would be placed in Group 2 (FF).

Analysis

A pre-post research design was used in the study. Descriptive and demographic statistics were used to ensure the characteristics of the sample mirrored the greater population. To answer RQ1 and RQ2, two separate repeated measures analyses of variances (ANOVAs) were conducted to determine the significance of the extraneous treatment within each group and the differences between each group as a result of the interaction effect at week 12 of the NFL season. To answer RQ3, RQ4, and RQ5, nine separate repeated measures ANOVAs were conducted to determine the significance of the extraneous treatment within each group and the differences between each group as a result of the interaction effect at week 12 of the NFL season. Differences between each group during the pre-season were not expected, but were assessed anyway.

Multiple comparison procedures (post hoc tests) were then assessed to determine which group means differ after the overall significance tests demonstrated at least one significant difference (Klockars & Sax, 1986). Lastly, tests of simple effects were conducted to determine the statistical significance of the change in participant attitudes and behavioral intentions from pre-season to week 12.

Results

The current sample of fantasy participants was slightly younger and less affluent, but for the most part matched the demographics and descriptives of the larger fantasy football population (FSTA, 2008). Cronbach's, Average Variance Extracted, and correlation scores were interpreted on the attitudinal constructs to determine preliminary reliability and convergent validity. In all, the scale scores (Table 2) showed adequate to good evidence of internal consistency, convergent validity, and discriminant validity (Fornell & Larcker, 1981; Nunnally, 1978). Further validation is required and encouraged of the Attraction to Fantasy Players scale, as it has only been tested twice on the same sample.

Attraction to Fantasy Players and NFL Team Attachment (RQ1 & RQ2)

The interaction effect between fantasy and NFL game outcomes and attraction to fantasy football players resulted in a statistically significant difference (F[df]=19.204[3,126]; p<.001). As expected, the post hoc analysis (Tukey) of the interaction effect during the pre-season did not result in any statistically significant attraction score differences. At week 12, however, statistically significant attraction score differences existed between groups that experienced fantasy team failure (BF: [sd]=3.816[.856] & FF: [sd]=4.161[.963]) and the groups that did not (TF: [sd]=4.681[1.026] & NF: [sd]=5.406[.755]). The attraction to fantasy players scale was measured on a seven point Likert-type scale. In addition, those who experienced fantasy team success, but favorite team failure indicated significantly higher attraction scores than those who experienced success with both teams. The test of simple effects results (Figure 3) suggest a statistically significant change in attraction scores from the pre-season to Week 12 for BF (negative), FF (negative), and TF (positive), while NF's attraction scores remained constant. The impact of NFL game outcomes on favorite NFL team attachment resulted in no statistically significant differences within-subjects or as a result of the interaction effect (F[df]=.965[3,126]; p=.124). The group mean attachment scores ranged from 5.661 to 6.156 on a seven-point Likert-type scale.

[FIGURE 3 OMITTED]

Fantasy Football-Related Media Consumption (RQ3)

Three separate repeated measures ANOVAs were conducted to analyze the impact of fantasy and NFL game outcomes on fantasy football-specific media consumption, one for each mode of media consumption (inter net usage, televised game broadcasts & cell phone usage). The results suggested a statistically significant interaction effect for televised game broadcasts and cell phone usage (Television: F[df]=5.517[3,126], p<.001 & Cell Phone: F[df]=11.867[3,126], p<.001), but no statistical significance for internet usage (F[df]=.097[3,126]; p=.961). The pre-season post hoc analyses for the consumption of both fantasy team-related televised game broadcasts and cell phone usage did not result in any significant differences between the groups. The televised game broadcast post hoc analysis (Tukey) at Week 12, however, resulted in statistically significant differences between groups that experienced fantasy team failure (BF: [sd]=8.039[1.110] & FF: [sd]=7.951[1.209]) and the groups that did not (TF: [sd]=9.460[.813] & NF: [sd]=9.115[.807]). All of the consumption scales were measured on an 11 point Juster scale. The cell phone usage post hoc analysis at Week 12 resulted in similar likelihood to consume score differences (BF: [sd]= 5.745[1.291], FF: [sd]=5.564[1.364], TF: [sd]=6.923[1.288], & NF: [sd]=6.811[1.399]). The test of simple effects for the televised game broadcast results (Figure 4) suggest a statistically significant change in likelihood to consume scores from the pre-season to Week 12 for BF (negative), FF (negative), and TF (positive), while NF's likelihood to consume scores remained constant. The test of simple effects for the cell phone usage results (Figure 5) suggest a statistically significant change in likelihood to consume scores from the pre-season to Week 12 for BF (negative) and FF (negative), while NF's and TF's likelihood to consume scores remained constant.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

Favorite NFL Team-Related Media Consumption (RQ4)

Three separate repeated measures ANOVAs were also conducted to analyze the impact of fantasy and NFL game outcomes on favorite NFL team media consumption (RQ4), one for each mode of media consumption. The equality of covariance assumption (Box's Test) was violated for the internet usage and televised game broadcasts analyses. Thus, Pilai's Trace statistic was interpreted for these contrasts. Wilk's Lambda was interpreted for the cell phone usage test. The results suggested statistically significant interaction effects for each mode of media consumption (Internet: P[df]=16.307[3,126], p<.001; Television: P[df]=12.657[3,126], p<.001, & Cell Phone: F[df]=5.768[3,126]; p<001). The pre-season post hoc analyses for each mode of favorite NFL team media consumption did not result in any significant differences between the groups. The internet usage post hoc analysis (Tanhame) at Week 12 resulted in statistically significant differences between groups that experienced favorite NFL team failure (BF: [sd]=5.898[1.410] & TF: [sd]=6.850[1.273]) and the groups that did not (fF: [sd]=7.567[1.030] & NF: [sd]=8.400[.789]). The groups that experienced just fantasy or favorite NFL team failure (FF & TF), however, did not differ significantly. The test of simple effects for the internet usage results (Figure 6) suggested a statistically significant change in likelihood to consume scores from the preseason to Week 12 for BF (negative) and TF (negative), while NF's and FF's likelihood to consume scores remained constant.

[FIGURE 6 OMITTED]

Similarly, the televised game broadcasts post hoc analysis (Tanhame) at Week 12 resulted in statistically significant differences between groups that experienced favorite NFL team failure (BF: [sd]=7.531[1.111] & TF: [sd]=7.684[1.003]) and the groups that did not (FF: [sd]=9.456[.789] & NF: [sd]=9.863[.681]). The test of simple effects for the televised game broadcast results (Figure 7) suggest a statistically significant change in likelihood to consume scores from the pre-season to Week 12 for BF (negative) and TF (negative), while NF's and FF's likelihood to consume scores remained constant. The cell phone usage post hoc analysis (Tukey) at Week 12 resulted in statistically significant differences between groups that experienced favorite NFL team failure (BF: [sd]=4.259[2.005] & TF: [sd]=3.816[2.413]) and the groups that did not (FF: [sd]=5.967[1.295] & NF: [sd]=5.530[1.382]). The test of simple effects for the cell phone usage results (Figure 8) suggest a statistically significant change in likelihood to consume scores from the pre-season to Week 12 for BF (negative) and TF (negative), while NF's and FF's likelihood to consume scores remained constant.

General NFL Media Consumption (RQ5)

Lastly, three separate repeated measures ANOVAs were conducted to analyze the impact of fantasy and NFL game outcomes on general NFL media consumption, one for each form of general NFL media programming (pre-game television shows, post-game television shows, & MNF). The results suggested a statistically significant interaction effect for each form of general NFL media programming (Pre-Game TV: P[df]=7.014[3,126], p<.001; Post-Game TV: P[df]=13.790[3,126], p<.001, & MNF: P[df]=14.780[3,126]; p<001). The preseason post hoc analyses for each form of general NFL media programming did not result in any significant differences between the groups. The pregame television show post hoc analyses (Tukey) at Week 12, however, resulted in statistically significant differences between groups that experienced fantasy team failure (BF: [sd]=5.987[1.301] & FF: [sd]=6.024[1.164]) and the groups that did not (TF: [sd]= 8.460[.946] & NF: [sd]=7.732[1.190]). The postgame television show post hoc analysis (Tukey) at Week 12 resulted in similar differences (BF: [sd]=5.664[1.329], FF: [sd]=5.390[1.410], TF: [sd]=7.700[1.116], & NF: [sd]=7.306[1.383]). The test of simple effects for the pre-game show results (Figure 9) suggest a statistically significant change in likelihood to consume scores from the pre-season to Week 12 for BF (negative), FF (negative), and TF (positive), while NF's likelihood to consume scores remained constant. The test of simple effects for the post-game show results (Figure 10) suggest a statistically significant change in likelihood to consume scores from the pre-season to Week 12 for BF (negative) and FF (negative), while NF's and TF's likelihood to consume scores remained constant.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

[FIGURE 11 OMITTED]

The MNF post hoc analysis (Tukey) at Week 12 also resulted in statistically significant differences between groups that experienced fantasy team failure (BF: [sd]=5.595[1.368] & FF: [sd]=6.644[1.000]) and the groups that did not (TF: [sd]=8.202[.991] & NF: [sd]=7.904[1.171]). In addition, BF's likelihood to watch MNF was statistically significant different from FF's likelihood. The test of simple effects for the MNF results (Figure 11) suggest a statistically significant negative change in likelihood to consume scores from the pre-season to Week 12 for each group.

Discussion

In 2010, Drayer et al. developed a conceptual model to explain the relationship between fantasy football and National Football League (NFL) consumption (see Figure 1). Within this framework, the authors' proposed that a participant's perceptions and media consumption of the NFL were constantly changing based on fantasy and favorite team outcomes. This study assessed this proposition, by employing a pre-post research design that examined the impact of favorite NFL team and fantasy team game outcomes on attraction to fantasy football players, attachment to NFL team, and media consumption of one's fantasy team, favorite team, and the NFL, in general.

Taken together, the results of this study provide much needed, in-depth psychographic and product usage information related to the continually evolving relationship between fantasy football participants and the NFL. More specifically, the findings support the contention that fantasy football participation is a powerful brand-building activity for the NFL where (1) enhanced media consumption of the NFL product was spurred by successful fantasy participation, and (2) favorite NFL team identification among this group of highly engaged football fans was not negatively affected when media consumption of the team dissipated slightly. For the NFL and media providers (i.e., ESPN, Yahoo!, CBSsports.com, etc.), these findings suggest it may be prudent to take extra measures to ensure fantasy participants remain competitive as long as possible during the NFL season. For instance, encouraging participants to compete in a balanced league could prolong competitive interest. In other words, with skill related to fantasy football participation ranging from one individual to the next, keeping the skill variance within leagues to a minimum may keep participants interested in their fantasy football teams longer as chances for the post season persist. In addition, tweaking league settings to include a more inclusive fantasy postseason may keep participants engaged longer into the season. While individual league commissioners hold a great deal of control with respect to league make-up and rules, nearly each site has a default league setting which is often the most popular format. A slight adjustment to the default setting that increased the number of teams that make the playoffs may create increased NFL media consumption during the second half of the season. The results are discussed in more detail below.

With regard to the attraction results, the findings suggest NFL game outcomes related to one's fantasy team impact attraction to fantasy football players as fantasy team failure had a negative impact on week 12 attraction levels. Team success also positively impacted attraction levels, but only when associated with NFL team failure. This confirms Drayer et al.'s (2010) conceptual framework that stipulates game outcomes impact participant attitudes related to fantasy football. Conversely, game outcomes had no impact on a fantasy participant's NFL team attachment according to this study's results. In other words, it appears that neither success nor failure of one's fantasy or favorite NFL team had any impact on the highly-developed connection between a participant and his/her favorite NFL team. This contradicts the conceptual framework that suggested normative attitudes (NFL team attachment) would change as result of in-season outcomes. It does, however, support previous research about the durability of team identification from an attitudinal perspective (Trail et al., 2003; Wann & Branscombe, 1993). Additionally, it validates the complementary nature of fantasy participation from an attitudinal perspective (Dwyer, Shapiro, & Drayer, 2011; Karg & McDonald, 2011).

NFL game outcomes, however, did impact media consumption intentions for both one's fantasy team and favorite NFL team for nearly every medium. The impact was primarily due to team failure. Team success had little impact on media consumption intentions as it appears a participant's consumption level enters the season at a high level and remains high if the team is successful. This behavioral optimism is common in sports as high expectations entering a season are how fans maintain a "positive attitude/brand loyalty towards a product of pedestrian quality" (p. 269, Bristow & Sebastian, 2001).

As for specific consumption intentions, fantasy team failure resulted in statistically significant lower consumption intention scores for fantasy team-related cell phone use and viewership of televised game broadcasts. Statistically significant intention changes did not exist with regard to fantasy-related internet use which is somewhat logical given the activity's web dependence. In other words, it is virtually impossible to participate in fantasy football without the internet, so it is intuitive to think that internet use would remain relatively constant despite failure because a decrease in fantasy-related internet use would essentially result in non-participation. Interestingly, favorite NFL team failure also impacted fantasy team-related consumption as the group that experienced fantasy success and favorite team failure indicated a statistically significant increase in televised game viewership intentions. From an A-BR perspective, this behavioral intention is congruent with the attitudinal finding wherein a participant's attraction to fantasy players increased as a result of fantasy success and favorite team failure. In general, the fantasy team-related attraction and intention results paralleled Fazio et al.'s (1983) A-BR framework in that groups with higher attraction scores (attitudes) also indicated higher media consumption intention scores (behaviors) and vice versa. In addition, these results confirm Drayer et al.'s (2010) conceptual framework where game outcomes influence both fantasy-related attitudes and media consumption.

With regard to favorite NFL team media consumption, only favorite team failure negatively impacted consumption intentions. While this confirms Drayer et al.'s (2010) conceptual framework that suggests a change in consumption as a result of game outcomes, it appears to occur without the respective attitudinal antecedent. The A-BR framework suggests behaviors directed at a product are guided or influenced by a corresponding attitude toward said product. Specifically, consumption is the result of a positive attitude about the product and non-consumption is the result of a negative or non-attitude about a product (Fazio et al., 1989). Extended to the sport industry, it was determined that only a weak or non-attitude influences non-consumption, as even a strong negative attitude toward a sport team resulted in media consumption (Mahony & Howard, 1998; Mahony & Moorman, 1999). Therefore, the statistically significant decrease in media consumption intentions (behavior) as a result of favorite NFL team failure appears to circumvent the A-BR framework as the attachment to favorite NFL team scores (attitude) remained constant for all groups.

This counter intuitive phenomenon supports previous fantasy sports research that suggests a sport-distinct disconnect between one's highly-developed attitudes and behaviors with regard to one's favorite team (Dwyer, 2011a). However, several questions remain as to why a participant's behavior changes without an attitudinal guide. Perhaps it speaks to the uniqueness of sport as a product or the depth and power of the favorite NFL team bond? Perhaps it is the result of having additional competitive viewership options provided through fantasy football participation (Dwyer & Kim, 2011)? Regardless, further research in this area is needed to more deeply explain this indirect relationship.

[FIGURE 12 OMITTED]

The results regarding general NFL media consumption provide new knowledge about fantasy participation as it appears fantasy-related outcomes are more impactful in driving viewership of NFL pre-game shows, post-game shows, and ESPN's MNF. In other words, fantasy failure is more likely to inhibit one's interest in following the NFL from a league-wide perspective than favorite team failure. This somewhat confirms the conceptual framework that suggests fantasy participation leads to a league-wide interest. However, this study's results suggest that as a season progresses, it is not fantasy participation per se that drives media consumption it is fantasy success. If one's fantasy team is not successful, media consumption will decrease significantly. Favorite NFL team outcomes, however, had little impact in league-wide media consumption.

The Fantasy Football/NFL Consumption Framework Revisited

For the most part, the current study's findings validate proposition 3 of Drayer et al.'s conceptual framework for fantasy football participation and NFL consumption. In-season game outcomes impact the attitudes and behavioral intentions of fantasy football participants. However, one major revision is suggested, and as a result, further research is strongly advised. The revision centers on the feedback loop to one's normative attitudes related to their favorite NFL team (Figure 12). This study's results suggest game outcomes do not impact team identification. Thus, the feedback loop proposed by the authors is not supported. However, game outcomes do impact behavioral intentions. This implies something else, perhaps another variable or another stage within the framework, is driving this change in behavior. The current revision suggests this may occur in the Definition of the Event stage, but further inquiry is needed to validate this proposition.

Limitations to the study certainly exist. For instance, only attraction to fantasy players and team identification were examined from an attitudinal perspective. Drayer et al. (2010) identified interactivity and team loyalty as other attitudes attributed to fantasy football participation and NFL fandom, respectively. In addition, the attraction items need to be validated. The application within this study just piloted, panel-tested, and sample-tested twice on the same group. The current results related to the framework should also be tempered slightly as the entire model was not examined. In addition, the attitudes and behavioral intentions examined represent only a small fraction of a very large population of fantasy football participants. The replication of the same design with larger samples would be fruitful. Furthermore, it is prudent to remind the readers that fantasy sports participants are hardcore sports fans and these results do not parallel the general sport fan population. The extension to other fantasy sports is also advised. The current results are limited to fantasy football.

References

Azjen, I., & Fishbein, M. (2000). Attitudes and the attitude-behavior relation: Reasoned and automatic processes. European Review of Social Psychology, 11, 1-33.

Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84, 888918.

Bristow, D. N., & Sebastian, R. J. (2001). Holy cow! Wait 'til next year! A closer look at the brand loyalty of Chicago Cubs baseball fans. The Journal of Consumer Marketing, 18, 256-275.

Callero, P. L. (1985). Role-identity salience. Social Psychology Quarterly, 48, 203-215.

Drayer, J., Shapiro, S. L., Dwyer, B., Morse, A. L., & White, J. (2010). The effects of fantasy football participation on NFL consumption: A qualitative analysis. Sport Management Review, 13, 129-141.

Dwyer, B. (2011a). Divided loyalty? An analysis of fantasy football involvement and fan loyalty to individual National Football League (NFL) teams. Journal of Sport Management, 25, 445-457.

Dwyer, B. (2011b). The impact of attitudes and fantasy football involvement on intentions to watch NFL teams on television. International Journal of Sport Communication, 4, 375-396.

Dwyer, B., & Drayer, J. (2010). Fantasy sport consumer segmentation: An investigation into the differing consumption modes of fantasy football participants. Sport Marketing Quarterly, 19, 207-216.

Dwyer, B., & Kim, Y. (2011). For love or money: Developing and validating a motivational scale for fantasy football participation. Journal of Sport Management, 25, 70-83.

Dwyer, B., Shapiro, S.L., & Drayer, J. (2011). Segmenting motivation: An analysis of fantasy baseball motives and mediated sport consumption. Sport Marketing Quarterly, 20, 129-137.

Fantasy Sports Trade Association. (2008a). Fantasy sport consumer spending: Fantasy sports products and services. Proceedings from the FSTA Winter Business Conference. Denver, CO.

Fantasy Sports Trade Association. (2011, June 10). Fantasy sports-participation sets all-time record, grows past 32 million players. Retrieved from http://www.fsta.org/blog/fsta-press-release

Farquhar, L .K., & Meeds, R. (2007). Types of fantasy sports users and their motivations. Journal of Computer-Mediated Communication, 12, 1208-1228.

Fazio, R. H., Powell, M. C., & Herr, P. M. (1983). Toward a process model of the attitude-behavior relation: Accessing one's attitude upon mere observation of the attitude object. Journal of Personality and Social Psychology, 44, 723-735.

Fazio, R. H., Powell, M. C., & Williams, C. J. (1989). The role of attitude accessibility in the attitude-to-behavior process. Journal of Consumer Behavior, 16, 280-288.

Fazio, R. H., & Zanna, M. P. (1978). Attitudinal qualities relating to the strength of the attitude-behavior relationship. Journal of Experimental Social Psychology, 14, 398-408.

Fisher, E., (2008, November 17). Study: Fantasy players spend big. Street & Smith's SportsBusiness Journal, 11, 1-2.

Fornell, C., &, Larcker, D. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39-50.

Fortunato, J. A. (2004). The rival concept: An analysis of the 2002 Monday Night Football season. Journal of Sport Management, 18, 383-397.

Foxall, G. (1990). Consumer psychology in behavioural perspective. London, UK: Routledge.

Funk, D. C., & James, J. (2001). The Psychological Continuum Model (PCM): A conceptual framework for understanding an individual's psychological connection to sport. Sport Management Review, 4, 119-150.

Funk, D. C., & James, J. (2006). Customer loyalty: The meaning of attachment in the development of sport team allegiance. Journal of Sport Management, 20, 189-217.

Havitz, M. E., & Dimanche, F. (1997). Leisure involvement revisited: Conceptual conundrums and measurement advances. Journal of Leisure Research, 29, 245-278.

Herr, P. (1995). Whither fact, artifact, and attitude: Reflections on the theory of reasoned action. Journal of Consumer Psychology, 4, 371-380.

Iwasaki, Y., & Havitz, M. E. (1998). A path analytic model of the relationships between involvement, psychological commitment, and loyalty. Journal of Leisure Research, 30, 256-280.

Karg, A. J., & McDonald, H. (2011). Fantasy sport participation as a complement to traditional sport consumption. Sport Management Review, 14, 327-346.

Kwak, D. H., Lim, C. H., Lee, W. Y., & Mahan, J. (2010). How confident are you to win your fantasy league: Exploring the antecedents and consequences of winning expectancy. Journal of Sport Management, 24, 416-433.

Kwon, H. H., Trail, G. T., & Anderson, D. (2006). Points of attachment (identification) and licensed merchandise consumption among American college students. International Journal of Sport Management 7, 347-360.

Laurent, G., & Kapferer, J. (1985). Measuring consumer involvement profiles. Journal of Marketing Research, 22, 1-53.

Madrigal. R. (1995). Cognitive and affective determinants of fan satisfaction with sporting event attendance. Journal of Leisure Research, 27, 205-227.

Mahony, D. F., & Howard, D. R. (1998). The impact of attitudes on the behavioral intentions of sport spectators. International Sports Journal, 2, 96-110.

Mahony, D. F., & Moorman, A. M. (1999). The impact of attitudes on intentions to watch professional basketball teams on television. Sport Management Review, 2, 43-66.

Mahony, D. F., & Moorman, A. M. (2000). The relationship between the attitudes of professional sport fans and their intentions to watch televised games. Sport Marketing Quarterly, 9, 131-139.

Nesbit, T. M., & King, K. A. (2010a). The impact of fantasy sports on television viewership. Journal of Media Economics, 23, 24-41.

Nesbit, T. M., & King, K. A. (2010b). The impact of fantasy football participation on NFL attendance. Atlantic Economic Journal, 38, 95-108.

Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York, NY: McGraw-Hill.

Oates, T. P. (2009). New media and the repackaging of NFL fandom. Sociology of Sport Journal, 26, 31-49.

Robinson, M. J., & Trail, G. T. (2005). Relationships among spectator gender, motives, points of attachment, and sport preference. Journal of Sport Management, 19, 58-80.

Shapiro, S. L., Drayer, J., & Dwyer, B. (2011, October). Exploring fantasy baseball consumer behavior: Examining the relationship between identification, fantasy participation, and consumption. Paper presented at the Ninth Sport Marketing Association Conference, Houston, TX.

Tajfel, H., & Turner J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel & W. Austin (Eds.), Psychology of intergroup relations (pp. 6-13). Chicago, IL: Nelson Hall.

Trail, G. T., Anderson, D. F., & Fink, I. S. (2000). A theoretical model of sport spectator consumption behavior. International Journal of Sport Management, 1, 154-180.

Trail, G. T., Anderson, D. F., & Fink, J. S. (2005). Consumer satisfaction and identity theory: A model of sport spectator conative loyalty. Sport Marketing Quarterly, 14, 98-111.

Trail, G. T., Fink, J. S., & Anderson, D. F. (2003). Sport spectator consumption behavior. Sport Marketing Quarterly, 12, 8-17.

Trail, G. T., Robinson, M. J., Dick, R. J., & Gillentine, A. J. (2003). Motives and points of attachment: Fans versus spectators in intercollegiate athletics. Sport Marketing Quarterly, 12, 217-227.

Walker, M., & Heere, B. (2011). Consumer attitudes toward responsible entities in sport (CARES): Scale development and model testing. Sport Management Review, 14, 153-166.

Wann, D. L., & Branscombe, N. R. (1993). Sport fans: Measuring degree of identification with their team. International Journal of Sport Psychology, 24, 1-17.

Zillman, D., Bryant, J., & Sapolsky, B.S. (1989). Enjoyment from sports spectatorship. In J. D. Goldstein (Ed.), Sports, games, and play: Social and psychosocial viewpoints (2nd ed., pp. 241-278), Hillsdale, NJ: Lawrence Erlbaum Associates.

Brendan Dwyer, PhD, is the director of research and distance learning for the Center for Sport Leadership at Virginia Commonwealth University. His research interests center around sport consumer behavior with a distinct focus on the media consumption habits of fantasy sport participants and the consumer psychology surrounding sport pricing transactions.

Table 1
Sample Demographic and Descriptive Statistics (n=236)

Age                    32.525, Mean
                       9.784, St. Dev.
                       18-59, Range

Gender                 87.0%, Male
                       3.8%, Female
                       9.2%, Did not specify

Education              19.2%, High School
                       32.8%, Bachelor's Degree
                       22.9%, Graduate Degree
                       15.2%, Other
                       9.9%, Did not specify

Household              18.3%, Less than $50K
Income                 35.9%, $50K-$99K
                       22.9%, $100K-$150K
                       6.1%, More than $150K
                       16.8%, Did not specify

Ethnicity              81.7%, Caucasian
                       8.4%, Other
                       9.9%, Did not specify

Money Allocated        $107.40, Mean
to League Pool         $111.15, St. Dev.
                       $0-$750, Range

Years Played           6.366, Mean
                       4.616, St. Dev.
                       1-24, Range

Number of              3.840, Mean
Teams Owned            4.423, St. Dev.

Money Spent on         $12.76, Mean
Fantasy Football       $36.81, St. Dev.
Products & Services    $0-$350, Range

Table 2
Reliability and Convergent Validity of Attitudinal Constructs

                       Cronbach's       AVE        Correlation
                        [alpha]
                       Pre    Post   Pre    Post   Pre      Post

NFL Team Attachment    .813   .795   .546   .532   -.102    -.087
(1 dimension; Trail
et al., 2008)

Being a fan of the
[favorite NFL team]
is very important to
me.

I would experience a
loss if I had to
stop being a fan of
the [favorite NFL
team].

I consider myself to
be a "real" fan of
the [favorite NFL
team].

Attraction to          .803   .811   .513   .588   --       --
Fantasy Football
Players * (1
dimension)

Following my fantasy
football players is
a pleasurable
experience.

The performance of
my fantasy football
players is important
to me.

My fantasy football
players interest me.

* New Construct