Introduction

Formula 1 racing has millions of fans worldwide, generating millions in revenue. For the Formula 1 U.S. Grand Prix hosted by Circuit of The Americas (COTA) in Austin, Texas, more than 235,000 spectators attended the three-day event in 2014, but attendance has dropped successively over the years since COTA first hosted the event in 2012 (DiZinno, 2013; Theis, 2014). Though the Formula 1 World Championship is one of the world's most watched international mega sport events--roughly 425 million people watched the Formula 1 World Championship in 2014 (Sylt, 2015)--COTA relies on revenue generated from ticket sales to accrue any kind of profit margin due to the fact revenue generated from television rights and track sponsorships for all Formula 1 events go directly back to ownership and management (Gutman, 2012). A steady decline in race attendance in recent years coupled with limited potential for revenue generation has led several researchers to question whether a U.S. niche market like Austin can sustain hosting the Formula 1 race (e.g., Jenson, Cobbs, & Groza, 2014), particularly given how much bringing Formula 1 to the US cost COTA and the state of Texas (Sylt, 2014).

Making matters worse for host race facilities like COTA is the fact Formula 1 has faced criticism recently regarding the increasing financial commitments demanded of the racing teams and the subsequent withdrawal of racing teams in Formula 1's most popular events. According to Weaver (2013), more than half of the sport's $1.5 billion income goes directly to a private equity firm with a controlling stake in Formula 1, while racing teams spend upwards of $250 million to be involved in Formula 1. Such a lack of revenue sharing and increased investment spending has recently forced two racing teams to file for bankruptcy (Weaver, 2014), and may perhaps explain why a decade-low 18 cars competed in the Formula 1 U.S. Grand Prix last year. Thus, a growing concern for Formula 1 becomes whether the costly nature of the sport for spectators, participants, and organizers might potentially damage future loyalty to the brand (i.e., conative loyalty).

Many researchers have examined conative loyalty (behavioral intentions supporting the brand or team; see Shapiro, Ridinger, & Trail, 2013; Trail, Fink, & Anderson, 2003), or related constructs (e.g., group-supportive behaviors; Fisher & Wakefield, 1998), in the context of sport. For example, Trail, Kwon, and Anderson (2009) found advertising mitigated the effects of losing on conative loyalty for low-involvement consumers of an intercollegiate men's basketball team. Trail, Anderson, and Fink (2005) found attachment was correlated with conative loyalty, while Harrolle, Trail, Rodriguez, and Jordan (2010) showed an indirect relationship, but did not test or report a direct correlation. Gwinner and Swanson (2003) found highly identified fans exhibited higher conative loyalty toward sponsors of the team (e.g., purposely look for sponsors, denote a stronger purchase intent), while Matsuoka, Chelladurai, and Harada (2003) determined high fan identification was related to strong intentions to attend future professional soccer games. Amidst a plethora of research on conative loyalty in the sport marketing literature, however, there exists a relative shortage of research on predictors of conative loyalty among sport consumers.

Based on identity theory and extant literature on identification and spectator consumption behavior, we sought to investigate the differential effects of motives and points of attachment on conative loyalty of attendees of the Formula 1 U.S. Grand Prix. Although researchers in the sport marketing literature have examined motives and points of attachment (e.g., Funk & James, 2006; Robinson, Trail, & Kwon, 2004; Trail, Robinson, Dick, & Gillentine, 2003; Woo, Trail, Kwon, & Anderson, 2009), there is a relative dearth of research devoted to investigating the effects of motives and points of attachment on conative loyalty (e.g., Harrolle et al., 2010). Further, while studies on Formula 1 have grown rapidly in recent years, a comprehensive model of motives and points of attachment predicting conative loyalty has not yet been proposed. The current research provides some meaningful insights and implications in these regards.

Theoretical Framework

Motivation

Motivation has been defined as "the driving force within individuals that impels them to action" (Schiffman & Kanuk, 2004, p. 87), and "the energizing force that activates behavior and provides purpose and direction for that behavior" (Hawkins, Best, & Coney, 2004, p. 354). There are many theories of motivation throughout various disciplines; however, this study concerns the motivation for consumption, predominately spectator sport consumption (e.g., event attendance). While there is still much to learn about motives in sport consumption, there is a well-established stream of sport marketing research dedicated to the topic. Trail and James (2001) claim there are more than 40 different motives that have been proposed by scholars, all of which are considered to predict being a spectator/fan and predict sport consumption behavior or behavioral intentions, including value development (e.g., Milne & McDonald, 1999), eustress (e.g., Gantz & Wenner, 1995), and self-esteem enhancement (e.g., Branscombe & Wann, 1991), to name a few. Wann's (1995, 1997) Sport Fan Motivation Scale (SFMS) measures eight motives of sport consumption, and includes eustress, self-esteem, escape, entertainment, economic (gambling), aesthetic, group affiliation, and family as factors driving sport spectating. Trail and James (2001) developed the Motivation Scale for Sport Consumption (MSSC) that identifies nine motivations: achievement, acquisition of knowledge, aesthetics, drama, escape, family, skill, physical attraction, and social interaction. Funk, Mahony, Nakazawa, and Hirakawa's (2001) Sport Interest Inventory (SII) has also contributed a broad range of motivational factors for sport consumption.

Not only have previous scholars established an inventory of general sport consumption motives, but also they have gone further in suggesting that motivations differ by type of sport (Robinson & Trail, 2005; Wann, Grieve, Zapalac, & Pease, 2008; Wann, Schrader, & Wilson, 1999). For example, work by Wann et al. (1999) showed consumers who preferred individual (versus team) sport were motivated more by aesthetics than by any other motive in the SFMS. Consumers who preferred team sport were motivated by eustress and self-esteem more than any other motives in the SFMS. In addition, consumers who preferred aggressive to non-aggressive sport were most strongly motivated by economic (gambling) factors, and those consumers who preferred non-aggressive sport were most strongly motivated by aesthetics (Wann et al., 1999). To this end, Wann et al. (2008) categorized different sports into three dichotomous pairings--individual/team, aggressive/non-aggressive, and stylistic/non-stylistic--and found motivations for the various sports depended on how they were classified among these dichotomies.

Based on Sloan's (1989) argument that there are distinct differences between observers and fans, and that motivations may differ between the two, Trail, Robinson et al. (2003) proposed several models that categorized motives by those specific to fans, those specific to spectators, and those that applied to both. They suggested that spectator motives were those related to the skill of the athletes, aesthetics of the game, drama, and knowledge acquisition. The only motive that they suggested was specific solely to fans was vicarious achievement. However, they also proposed that there were overarching motives that would be applicable to both spectators and fans--these motives were comprised of escape and social interaction. These categorizations were supported.

Woo et al. (2009) reassessed Trail, Robinson et al.'s (2003) models and extended them, but tested competing models in which they moved social interaction to be a part of fan motives out of the overarching motives, and left only escape as the sole overarching motive. Woo et al. chose this competing model for several reasons. However, there were no significant differences between the fit of the competing models, so either model could have been selected based solely on fit.

Similar to Trail, Robinson et al. (2003) and Woo et al. (2009), we chose the MSSC to base our selection of motives for this research. The motives included social interaction, drama, physical skill, aesthetics, success (vicarious achievement), and acquisition of knowledge. In addition, we attempted to generate sport-specific motives due to the research from Wann and colleagues, who suggest sports differ on motivation (Wann et al., 1999, 2008). To that end, we had a fair amount of input from a pre-study Formula 1 focus group and we generated ideas from the popular press, as well. Furthermore, consistent with Trail, Robinson et al. (2003) and Woo et al. (2009), we hypothesized categories of motives (e.g., spectator motives).

Through this process, we added a motive of nostalgia to the spectator motives second order construct based on input from our pre-study focus group. We also moved social interaction to the spectator motives construct based on the input that true fans do not go to Formula 1 events for the social aspects--these individuals are referred to as "superficial" fans. We removed aesthetics from this second order construct, as well, as we explain later. Therefore, the spectator motives construct included acquisition of knowledge, appreciation of the drivers' skill, drama, nostalgia, and social interaction, leaving vicarious achievement (team success) as the only fan motive and escape as the only overarching motive.

However, in addition to our review of sport motivation literature and our focus group with longtime Formula 1 fans, we assessed there might be supplementary motives for following Formula 1 or being a Formula 1 fan that were specific to the sport and not included in any of the prior measures of motives. One of the recent changes that Formula 1 has implemented has been a new look and sound for the cars, including a new engine that is considerably quieter than the old engine (Johnson, 2014). The longtime fans in the focus group were not in favor of these changes at all.

However, some of the popular press articles noted that some people liked the new look and sound because it was being marketed as more eco-friendly (e.g., Golson, 2014). Thus, we created items to measure whether the new look, sound, and engine of cars were motives for people. Also, one of the enhancements to the new cars was an upgrade in technology that reportedly helped the drivers perform more efficiently. However, there was disagreement in the focus group as to whether this was a good thing or not. As such, we created items to determine if this was a motive. Finally, the U.S. Grand Prix has a relatively new track that is supposed to be a very aesthetically pleasing track (Alberstadt, 2014). Therefore, we modified the aesthetics motive scale to be track specific and included it in the category that we call product motives. In the end, we had five new motive subscales that were specific to the product (Formula 1 racing), which created a fourth dimension. Based on the above information we created our first hypothesis:

Hypothesis 1: The four separate motive dimensions (i.e., the spectator motive dimension, the product motive dimension, team success--a first order latent variable that by itself represents fan motives--and an overarching motive of escape) will be correlated with each other (see Figure 1).

Points of Attachment

Within identity theory, Stryker and Burke (2000) suggest motives impact identification, or role identities. Specific to sport, several researchers have examined the relationships between motives and team identification, or more recently points of attachment. Much of the extant work on sport consumption has been based on the framework of identity theory, which integrates both motives and identification, as previously noted. Stryker and Burke (2000) define identity as "parts of a self, composed of the meanings that persons attach to the multiple roles they typically play in highly differentiated contemporary societies" (p. 284). A role is an identity standard, which is a cognitive representation (belief) about what the particular role entails (Stryker & Burke, 2000). A person's identity standard interacts with perceived situational meanings and causes a cognitive comparison between the ideal role and the situationally impacted role. The cognitive comparison, if positive, results in positive emotions and eventually behaviors (Stryker & Burke, 2000).

Fink, Trail, and Anderson (2002) defined identification within the context of sport consumption as the "orientation of the self in regard to other objects (the team) that results in feelings or sentiments of close attachment" (p. 198). This definition distinguishes between the cognitive "orientation of self' and the affective results of feelings or sentiments regarding close attachment. Based on identity theory and research of Trail et al. (2000) and Trail, Robinson et al. (2003), Robinson and Trail (2005) determined there are several points of attachment associated with the connection to a team or sport organization (e.g., player, coach, community). From the viewpoint of identity theory, different roles may be formed based on different points of attachment and the commitment to those roles. For example, a person can be a fan of the Los Angeles Dodgers (an identity role), where the point of attachment is the team (Dodgers). However, the same person might also be a baseball fan (another role). The point of attachment in this case is the sport. In addition, perhaps that individual is also a fan of Clayton Kershaw, the Dodgers' pitcher (player attachment). Each of these identities could be salient at the same time if Kershaw was pitching for the Dodgers on a particular night. However, the player-attachment role might not be salient on nights when he is not pitching. The identity of being a Dodger fan might also not be relevant if the person is watching the World Series game between the Kansas City Royals and New York Mets, but the baseball-fan identity would be. Roles also have a hierarchy of importance (Stryker & Burke, 2000). Being a Dodger fan might be more important to a person than being a Kershaw fan. If Kershaw was traded to the Tampa Bay Rays, perhaps the individual might no longer be a Kershaw fan. On the other hand, if the individual's commitment to the player is strong enough, then a new role might be formed during the trade, and the person now becomes a Rays fan.

Robinson and Trail (2005) suggested that there were at least seven points of attachment (i.e., player, team, coach, community, type of sport, university, and level of sport). Other studies found additional points of attachment as well (e.g., golf tour and charities; Robinson et al., 2004; athletic department and general sport fan; Shapiro et al., 2013). Trail, Robinson et al. (2003) proposed and tested a model in which attachment to the team, coach, community, university, and player all comprised a higher order construct called organizational identification (attachment). Trail, Robinson et al. (2003) distinguished organizational identification from sport identification--the latter included attachment to the level of sport and attachment to the sport itself. They found that organizational identification was distinct from sport identification. Woo et al. (2009) also found similar results, though community attachment was not included in the organizational identification dimension.

Similar to motives, we had to modify some of the attachment subscales and eliminate others due to the unique aspects of Formula 1 and racing in general. We modified attachment to player to attachment to driver, which made sense since the driver is really the only "player," but oftentimes is also the face of the organization. We did not need to modify the attachment to team subscale at all, nor the attachment to community subscale; however, we re-termed it attachment to place and made its focus the entire state of Texas where the U.S. Grand Prix is hosted. Further, based on information we gathered from popular press surrounding the U.S. Grand Prix, we added another attachment subscale for attachment to the track facility (COTA). We kept the attachment to sport (auto racing) and attachment to type of sport (Formula 1 racing). Based on research from Trail, Robinson et al. (2003) and Woo et al. (2009), we kept the idea of attachment categories and included organizational attachment (team and driver), sport attachment (sport and type of sport), and place attachment (Texas and COTA). However, rather than create second order latent variables like Woo et al. (2009), we kept the first order latent variables for two reasons. First, based on what we gathered from the focus group, the popular press, and track managers, they seemed to feel the order of some of the relationships between some of the points of attachment were important. For example, those who were attached to Texas would be more likely to be attached to the track itself (COTA) since a lot of state taxes were put into funding it. Further, the longtime fans in our focus group were adamant that the drivers were the face of most teams and drove the connection to the team, unlike in other sports where the team was paramount. Hence, we proposed our second and third hypotheses:

Hypothesis 2: Attachment to Texas will lead to attachment to COTA.

Hypothesis 3: Attachment to the driver will lead to attachment to the team.

The relationship between attachment to the sport of racing and attachment to Formula 1 was not clear among those involved in our focus group. Most of the focus group felt attachment to Formula 1 would lead to some degree of attachment to racing in general, but others disagreed. The popular press and track managers gave the impression that being a racing fan leads to being a Formula 1 fan. Trail, Robinson et al. (2003) and Woo et al. (2009) demonstrate the type of sport and the specific sport are correlated but no directional relationships have been suggested or tested. With little to guide us, we followed the rationale of the aforesaid parties and posited the following:

Hypothesis 4: Attachment to the sport of racing will be related to attachment to Formula 1 racing.

The second reason that we did not create second order constructs with the two first order latent variables is solely statistical. Structural equation modeling (SEM) programs often show that when there are only two first order latent variables loading on a second-order latent variable, one first-order variable will load perfectly, creating a boundary parameter violation (Harrolle et al., 2010). To circumvent this issue and to test the hypothesized order of the points of attachment, we tested the model as depicted in Figure 1. Further, based on theory and previous research that has shown motives impact fan identification and points of attachment, we believed that motives would vary across points of attachment, which we will discuss later.

Motivation and Points of Attachment

In regards to identity theory, Stryker and Burke (2000) suggested motives influence the interaction between the identity standard and the situational meanings, resulting in the cognitive comparison. Within the sport context, the relationship between motives and team identification has long been evaluated (e.g., Trail & James, 2001; Wann, 1995; Wann & Branscombe, 1993). For instance, Trail, Robinson et al. (2003) found that motives explained 70% of the variance in team identification, whereas Fink et al. (2002) showed the motive of vicarious achievement explained vastly more variance in fan identification than did other motives (e.g., drama/excitement, escape, and acquisition of knowledge). Relatedly, Wann and Branscombe (1993) found that individuals who were highly identified made more ego-enhancing attributions for their teams' success than did those with low to moderate fan identification. Furthermore, Wann, Ensor, and Bilyeu (2001) found intrinsic motivations for originally following a team were strongest for those individuals with high identification.

Based on identity theory and previous sport marketing literature (Fink et al., 2002; Kwon, Trail, & Anderson, 2005; Robinson & Trail, 2005; Robinson et al., 2004; Trail et al., 2003), Woo et al. (2009) proposed and compared multiple models that examined relationships among fan and spectator motivations and points of attachment. Woo et al. (2009) and Trail et al. (2003) showed spectator motives predicted sport identification--approximately 31% of the variance in the former research and a little over 50% in the latter. Woo et al. (2009) also found fan motives explained 77% of the variance in organizational identification, whereas Trail, Robinson et al. (2003) showed that vicarious achievement (sole fan motive) explained about 73% of the variance in organizational identification. As such, we also posited those relationships in our model. With our newly created product/place motive category, we believed motives would tie in most closely with attachment to place. We thus included that relationship in the model, creating the following three hypotheses:

Hypothesis 5: Fan motives (success of the team) will be related to attachment to the team indirectly through attachment to the driver.

Hypothesis 6: Spectator motives will be directly related to attachment to the sport and indirectly related to attachment to the type of sport (Formula 1) through attachment to the sport.

Hypothesis 7: Product/place motives will be directly related to attachment to the place (Texas) and attachment to the venue (COTA) indirectly through the former.

Conative Loyalty

Team identification has been associated with conative loyalty many times in the sport marketing literature, but the relationships among the remaining points of attachment to conative loyalty have not been investigated much at all. Conative loyalty is defined by Oliver (1999) as a behavioral intention to repurchase or a "deeply held commitment to buy" the brand in the future (p. 35). Though Eagly and Chaiken (1993) mentioned intentions (conative loyalty), they did not make a distinction between behavioral loyalty and conative loyalty. Dick and Basu (1994) gave some insights into the idea of a conative disposition; however, it included sunk costs, switching costs, and expectations. To this end, the expectations aspect of conation postulated by Dick and Basu (1994) was somewhat near to Oliver's (1999) idea of conative loyalty. Crosby and Taylor (1983) indicated that conations include intentions and were impacted by cognitive and affective constructs, which in turn led to purchase behavior. We use Oliver's (1999) definition and make the distinction that conative loyalty is the intention or likelihood of acting in a loyal manner in the future, and is different from cognitive, affective, and behavioral loyalty (as per Crosby and Taylor, 1983).

Several scholars have investigated conative loyalty or intentions within the context of sport, many of who measured the construct differently. Bodet and Bernache-Assollant (2011), Matsuoka et al. (2003), and Shapiro et al. (2013) all used a single-item measure of attendance intention. Wu, Tsai, and Hung (2012) used a three-item scale (intentions to attend games, watch games on TV, and purchase merchandise). Yoshida, Gordon, Nakazawa, and Biscaia (2014) had slightly different items (intentions to attend another sporting event, buy additional products, and spend more than 50% of the sport budget on the team), while Yoshida, Heere, and Gordon (2015) used two of the same items as in Yoshida et al.'s (2014) research, but used a probability-of-making-the-same-choice item for the merchandise item. Kwon et al. (2005) used three items (likely to attend future games, likely to buy team merchandise, likely to buy the team clothing). Trail and colleagues (Trail, Fink, & Anderson, 2003; Trail et al., 2005) used the same three items as Kwon et al. (2005) plus a fourth item (likelihood of supporting the team); whereas, Harrolle et al. (2010) utilized five items--the same four as Trail, Fink, and Anderson (2003) and Trail et al. (2005) plus an item that measured likelihood of watching the game highlights with other people.

Interestingly, in the multi-item scales of conative loyalty, some items may not load highly on the scale, indicating perhaps they represent different aspects of conative loyalty. For example, in Harrolle et al.'s (2010) research, both the attending future games item and watching highlights item had less than 35% common variance, indicating that they may be distinct. In addition, in Yoshida et al.'s (2015) research, one item (the probability of attending another sporting event of my team) only had 29% shared variance. There were comparable results in Trail et al.'s (2005) research, as well. The results of all these studies indicate perhaps the different aspects of conative loyalty should not be measured as one construct, but rather should be treated as in Gray and Wert-Gray's (2012) study in which the authors ran four separate regressions with attendance intentions, media based intentions, purchase of team merchandise intentions, and word-of-mouth communication intentions as the dependent variables.

For the purposes of this study, we chose three items from prior research (the likelihood of attending, likelihood of buying licensed merchandise, and likelihood of continuing to support the team; Trail et al., 2005) and modified them to be specific to Formula 1 racing and the U.S. Grand Prix. However we also added a word-of-mouth (WOM) item from Gray and WertGray (2012) to measure whether attendees would spread the word about their experiences during the U.S. Grand Prix. WOM is quite critical to increasing awareness of and interest in a sport event (Shreffler & Ross, 2013) and allows marketers to reach a greater number of people, thus increasing return on investment (Sernovitz, 2009). Based on the research, we posited the following hypothesis:

Hypothesis 8: The correlation between the four items measuring conative loyalty will be insufficient to warrant joining the items into a single factor (i.e., each item will share less than 50% of the variance with any other of the four items).

Relationship Between Points of Attachment and Conative Loyalty

Identity theory does not directly address conative loyalty. However, Stryker and Burke (2000) suggested role identities are expressed in behaviors that represent the identities. Further, Oliver (1999) and Crosby and Taylor (1983), among others, hypothesized that conative loyalty leads to behavioral loyalty. Hence, it would seem role identities might also influence conations (intentions) before influencing actual behaviors.

This relationship is supported extensively within sport consumer research. For example, researchers have shown team identification to be related to conative loyalty across a multitude of studies (see Bodet & Bernache-Assolant, 2011; Gray & Wert-Gray, 2012; Matsuoka et al., 2003; Shapiro et al., 2013; Trail et al., 2005; Wu et al., 2012; Yoshida et al., 2015). However, there is a relative dearth of research that examines relationships between points of attachment and conative loyalty with the exception being Kwon et al. (2005). These authors found that when all six points of attachment (team, university, sport, player, level, and coach) were utilized in a multiple regression model, attachment to the team explained the most variance in conative loyalty (represented by a likelihood of attending games and buying merchandise or team clothing), with attachment to the coach and attachment to the university adding little. The remaining points of attachment were not significant predictors.

In our model, we hypothesize that attachment to the team, attachment to COTA, and attachment to Formula 1 will have direct effects on various aspects of conative loyalty, while attachment to driver, attachment to place (Texas), and attachment to racing will have indirect effects on aspects of conative loyalty. However, the impacts of each will vary by the aspect of conative loyalty as was shown by Gray and Wert-Gray (2012). These researchers showed team identification explained 10% of the variance in attendance intentions, 38% in WOM communication intentions, 24% in merchandise consumption intentions, and 11% in media consumption intentions. Based on the above research, we created additional hypotheses:

Hypothesis 9: Attachment to the driver (indirectly) and attachment to the team (directly) will be differentially related to intention to buy team merchandise, intention to support the team, intention to attend the next U.S. Grand Prix, and intentions to tell others about their experiences, with the former two more highly related than the latter two.

Hypothesis 10: Attachment to place (indirectly) and attachment to COTA (directly) will be differentially related to intention to purchase team merchandise, intention to support the team, intention to attend the next U.S. Grand Prix, and intentions to tell others about their experiences, with the latter two more highly related than the former two.

Hypothesis 11: Attachment to racing (indirectly) and attachment to Formula 1 (directly) will be differentially related to intention to buy team merchandise, intention to support the team, intention to attend the next U.S. Grand Prix, and intentions to tell others about their experiences, with the latter two more highly related than the former two.

In sum, we seek to extend the aforementioned research with a few newly developed factors specific to Formula 1 and to posit different motives will differentially impact points of attachment, which in turn will be hierarchically ordered. We further posit points of attachment will subsequently be related to aspects of conative loyalty differentially, all of which is depicted in Figure 1. Although this figure depicts only one model, we are testing four models that are the same except that each conative loyalty item is interchanged as the dependent variable. In Model 1, the conative loyalty item is the intention to attend the next year's U.S. Grand Prix. In Model 2, the conative loyalty item is the intention to purchase team merchandise in the future. In Model 3, the conative loyalty item is the intention to continue to support the participant's favorite Formula 1 team in the future. In Model 4, the conative loyalty item is the intention to tell others about the participant's experiences at the race. We chose to test these four models separately rather than to include all four conative loyalty items in one model to determine the differential effects on each, a design that is similar to Gray and Wert-Gray (2012). If we had included all of them in the same model, they would modify other relationships in the model, making the overall effects less clear.

Method

Procedure

The sample (N=247) comprised attendees of the 2014 Formula 1 U.S. Grand Prix held in Austin, Texas. Data were collected over the course of three days, during which Formula 1 racing teams participated in several practice sessions, qualifying trials, and actual competition racing. A nearly equal number of questionnaires were completed during each of the three days of the event. A random sampling design was utilized whereby researchers intercepted attendees at various site locations across the approximately eight miles of concourse and spectator traffic space at COTA. Respondents were intercepted as they entered and exited the track facility, as well as during race breaks and qualifying rounds. The respondents were asked to complete a short questionnaire on which they were to respond to different scale items measuring motives and points of attachment. The purpose of the study was provided in the questionnaire with the instructions for completion of the survey. Researchers obtained institutional approval for research containing human subjects prior to survey collection. Participation was voluntary, and confidentiality was assured by efforts to hide participant information using a numeric coding system as recommended by Hair, Black, Babin, Anderson, and Tatham (2006).

Participants

The sample was made up of 54.0% males, and the average age of participants was 41 years. The racial makeup comprised 64.6% Caucasians, 10.7% Hispanics, 3.3% Asians, 2.4% African Americans, and 1.3% other races. Ten percent of the respondents were international attendees with a primary residence outside of the US. Respondents had an average range for household income that was $100,000-$149,999 (28.2%).

Measures

The scales used in this study were the Motivation Scale for Sport Consumption (MSSC) and the Points of Attachment Index (PAI). Trail and James (2001) initially developed the MSSC to examine nine motives, but more recent researchers have examined relationships among these nine motives and found two categories of fan motives and spectator motives (Woo et al., 2009). Consistent with the results of Woo et al. (2009), we measured five "spectator motives" (social interaction, nostalgia, drama, skill, and knowledge). Nostalgia was included in this dimension and social interaction was moved here due to input from the focus group. We also included the overarching motive of escape and the motive of perceived success of the team (adapted from Funk & James, 2006) as opposed to vicarious achievement, which is generally included in the MSSC. Furthermore, based on a focus group among longtime Formula 1 fans, we created some sport-specific motives we categorized as "product and place motives" (track aesthetics, sound of new cars, look of new cars, engine of new cars, and technological aspects). All of these had three items, except for drama (four), engine of new cars (one), and technological aspects (two), which led to a total of 31 items. All items were measured using 7-point Likert-type scales that ranged from strongly disagree to strongly agree. Previous sport marketing researchers have utilized the MSSC and shown good construct reliability, discriminant validity, criterion validity, and internal consistency (e.g., Fink et al., 2002; Robinson & Trail, 2005; Robinson et al., 2004; Trail & James, 2001).

Based on Trail, Robinson et al. (2003), a total of six subscales were used to measure points of attachment for (a) driver, (b) team, (c) place, (d) venue, (e) auto racing, and (f) Formula 1. Each subscale had three items; thus, the PAI had a total of 18 items with a 7point Likert-type scale response format (strongly disagree to strongly agree). Previous scholars have shown the PAI to have a strong reliability and convergent validity (e.g., AVE .48-68; Robinson & Trail, 2005). Further, three scale items from Trail et al.'s (2005) conative loyalty scale were included and a fourth item measuring WOM from Gray and Wert-Gray (2012) was also included. As was described in the previous section, each of the conative loyalty items was employed separately as an outcome dependent variable in each of the four models. Each of the items was prefaced with the following statement: "We are interested in what motivates you to attend the Formula 1 U.S. Grand Prix. Please rate the extent to which you disagree or agree with the following items."

Results

We used SEM to test the measurement models and the structural models. The CFA on the motives revealed an adequate fit (RMSEA = .069, CI = .061-.077, [chi square]/df= 787.4/440 = 1.79), as did that on points of attachment (RMSEA = .077, CI = .064-.090, [chi square]/df = 254.6/120 = 2.12). Cronbach's alpha values for the motives scales ranged from .719 to .950, and the AVE values ranged from .627 to .880. Alpha values were good for points of attachment, and they ranged from .721 to .888. The AVE values ranged from .409 to .750 (Table 1). All of the subscales were distinct from each other as AVE values exceeded the squared correlations (Table 2). The structural models fit adequately well (Model 1 [attendance intentions]: RMSEA = .083, CI = .078-.088, [chi square]/df = 2519/1302 = 1.93; Model 2 [merchandise intentions]: RMSEA = .083, CI = .078-.087, [chi square]/df = 2517/1302 = 1.93; Model 3 [support intentions]: RMSEA = .083, CI = .078-.088, [chi square]/df = 2526/1302 = 1.94; Model 4 [WOM intentions]: RMSEA = .084, CI = = .079-.089, [chi square]/df = 2560/1302 = 1.97). As per Cohen (1988), all paths in all of models were significant and meaningful, ranging from [beta] = .279 to [beta] = .859 (see Table 3), except for a few paths to some of the conative loyalty variables.

Discussion

Based on identity theory and extant literature on motivations, points of attachment, and spectator consumption behavior, we sought to investigate the differential effects of motives and points of attachment on conative loyalty by testing four versions of a model of conative loyalty of Formula 1 U.S. Grand Prix attendees. In addition, this study has significant implications for sport consumer research as it demonstrates how motives related to spectatorship, success of the team, and product/place significantly impact points of attachment, which mediate the motive-conative loyalty relationship. Lastly, we have developed and shown there are specific motives relevant to Formula 1 racing.

Overview of Significant Findings

Not only did we find that the four proposed models fit the data moderately well, we determined motives and points of attachment combined to explain about 37% of the variance in intentions to attend the U.S. Grand Prix the next year, 41% of the variance in intentions to buy team licensed merchandise in the future, 52% of the variance in supporting a favorite Formula 1 team in the future, and a little more than 23% of WOM intentions. Interestingly, different points of attachment contributed different amounts of the variance. Attachment to COTA explained the majority of variance in attendance intentions (32%), whereas attachment to the team explained a minor amount (less than 1%) and attachment to Formula 1 explained less than 4%. Conversely, attachment to the team explained most of the variance in supporting the team in the future (44%), whereas attachment to COTA and Formula 1 explained very little (less than 4%). Explaining the variance in intentions to buy team merchandise was more evenly distributed with attachment to Formula 1 explaining 15% and attachment to the team explaining 14%, but attachment to COTA explained only 4%. Although not much variance in WOM intentions was explained, most of it was done by attachment to COTA (19%).

Support for Hypotheses

Our model shows support for the theoretical framework that we used as well. Identity theory (Stryker & Burke, 2000) suggests motives impact role identities, which in turn results in behavior or, as in our research, behavioral intentions. We were able to show these relationships. Specifically, theory and applied research on motives suggest there are multiple motives (Sloan, 1989) and that they can be distinguished by category (Trail, Robinson et al. 2003; Woo et al., 2009). We found a second order spectator motive construct was apparent and well represented by social interaction, nostalgia, drama, appreciation of physical skills, and acquisition of knowledge (see Figure 2). We also found that a second order product/place motive construct existed and was moderately represented by our newly created motives specific to Formula 1 racing: aesthetics of the track, sound of the cars, look of the new cars, new engine, and technology (see Figure 2). In support of Hypothesis 1, these two higher order constructs were also correlated with the success of team motive and the overarching motive of escape. Although these motives categories were correlated to some degree, they were still distinct from each other--the largest amount of shared variance was between success of the team and spectator motives at less than 44% (see Table 3).

We were also able to support Hypotheses 2-4. Attachment to place (Texas) was related to attachment to COTA. However, shared variance was not as much as anticipated (less than 22%). This may have been due to the discontent with the amount of money (taxes) that was being spent supporting COTA and Formula 1 (see Sylt, 2014). Furthermore, the shared variance between attachment to driver and attachment to team was fairly high (approximately 53%), supporting Hypothesis 3 (as well as the focus group's adamant assertions). This amount of shared variance between attachment to player and attachment to team is much higher than what scholars have typically found in prior research (Robinson & Trail, 2005; Trail, Robinson et al., 2003; Woo et al., 2009), which indicates that Formula 1 is unique in this regard. We also found support for Hypothesis 4 and Woo et al.'s (2009) research in that attachment to sport would be related to attachment to Formula 1 (45%).

Our findings showed that motives differentially impact how people relate to the team and other aspects (Wann, 1995, 1997; Wann et al., 2001, 2008). In support of Hypothesis 5 and Woo et al. (2009), we showed that the motive of success of the team directly impacted attachment to the driver (60%), and indirectly impacted attachment to the team (32%) through attachment to the driver. Further, we found that spectator motives impacted attachment to racing (74%) and indirectly impacted attachment to Formula 1 (33%), supporting Hypothesis 6 and previous work by Woo et al. (2009). Although Hypothesis 7 was supported, the amount of variance explained in attachment to place by the product/place motives was lower than we anticipated (less than 20%) and the indirect effect on attachment to COTA was lower still (4%). This may have been due to the combination of product and place in the motives, but we are unsure.

Our findings showed support for Hypothesis 8, as well. The correlations among the four conative loyalty items were not high, with the largest amount of shared variance less than 35%. This is consistent with the findings of Gray and Wert-Gray (2012) and perhaps explains some of the low item loadings for some conative loyalty items in previous research (Harrolle et al., 2010, Trail et al., 2005; Yoshida et al., 2015).

We also found support for Hypothesis 9, in that we found that attachment to the team and attachment to the driver (indirectly) differentially impacted the four conative loyalty items. The largest impact was on supporting the team in the future, followed by intention to buy team licensed merchandise, supporting the research of Kwon et al. (2005). However, the two points of attachment were not significantly related to attending the U.S. Grand Prix the following year or intentions to tell others about their U.S. Grand Prix experiences. Although we thought the latter relationships would be lower than the former, we did not expect them to be non-significant. We assumed there would be at least some relationship, as previous researchers have shown (Gray & Wert-Gray, 2012), but this was not the case.

Attachment to place (indirectly) and attachment to COTA (directly) showed the most consistent effects across the four conative loyalty items. We posited these points of attachment would be more highly correlated to intention to attend the next U.S. Grand Prix and in telling others about U.S. Grand Prix experiences than the other two items (Hypothesis 10). This was supported, as attachment to COTA explained 32% of intentions to attend and 18% of WOM intentions, respectively, whereas this point of attachment explained only around 4% of intentions to buy team merchandise and support the team. These latter relationships were similar to Kwon et al.'s (2005) findings.

Finally, Hypotheses 11 was only partially supported. We proposed the attachment to Formula 1 (directly) and the attachment to racing (indirectly) would differentially impact the four conative loyalty items. Attachment to Formula 1 explained about 15% of the variance in purchasing team licensed merchandise in the future, but did not explain a significant amount of variance in any of the other conative loyalty items (as per Cohen, 1988). The indirect effects of attachment to racing were not significant. These latter results were consistent with Kwon et al.'s (2005) results, in which attachment to sport or to level of sport was not significantly related to conative loyalty as a whole.

In sum, most of our hypotheses were supported through the findings. Primarily, motives variably impacted points of attachment, which in turn related to different conative loyalty items. Further, all of our models fit adequately well, and we explain relatively fair amounts of variance across most of our proposed relationships.

Implications of the Study

There are several practical implications for sport marketers and sport managers, as well as Formula 1 racing in particular. First, it is requisite to understand that what motivates people to attend a Formula 1 race in particular is varied and different in some aspects than other sports. In addition, the points of attachment are varied, as well. As such, one implication of the research is not surprising--segmenting this market is necessary, and marketing to these segments will vary fairly substantially relative to both motives and points of attachment. Second, with regard to the track facilities that host Formula 1 races, our results indicated that attachment to COTA is what drives people to want to attend the Formula 1 U.S. Grand Prix in the future and to discuss their past experiences at the event with others. To a lesser and small extent, attachment to Formula 1 and attachment to Texas also contribute to intentions to attend. However, our results showed that attachment to the driver and attachment to the team does not. Therefore, marketing should focus on the host track facility and to some extent Formula 1 itself, but not on the teams or drivers, at least for attendance reasons. Mean scores for the six different points of attachment support this argument: attachment to the team was the lowest mean of all the different points of attachment, whereas attachment to COTA was the highest mean score. This might indicate one of two things. Perhaps this was a representative sample of attendees, and they are not huge fans of specific Formula 1 teams, or of the drivers (mean score only slightly higher than team). Or, maybe this particular sample was not representative of all attendees. We doubt the latter is the case as the respondents in this research were randomly surveyed at various entrance, concourse, and exit locations of COTA across multiple days, so the sample should be representative.

Third, success of the team and driver is what drives fans to support their favorite team in the future and buy team licensed merchandise. As noted previously, a reduction in the number of drivers and teams at certain Formula 1 races has anecdotally had a deleterious impact on fandom (Weaver, 2014). This is not surprising based on results of this research. Further, attachment to COTA and to Formula 1 does not drive future loyalty to the team to any substantial extent. That is, the sport and the venue do not drive future team-related revenues, but the team and drivers do.

Fourth, the product and place motives we created and investigated specific to Formula 1 (e.g., new engines, new look of the car, sound of the engines, etc.), do not seem to be related that much to attachment to the place, to the racetrack (COTA), or to future support of the sport. This offers complementary support for the anecdotal information found in the popular press and from our focus group that fans do not enjoy the latest alterations to the engines or cars. The correlation matrix (see Table 2) does show that aesthetics of the track and new car technology is an exception to the rest of the product motives, as these two motives show a moderate amount of shared variance with some of the points of attachment and some of the intentions.

Fifth, support for the model and the variables therein should indicate to sport marketers and sport managers the need to identify the motives and points of attachment that are relevant to the specific environment. For example, we found it necessary to modify the variables and model from previous studies (e.g., Trail et al., 2005; Trail, Fink, & Anderson, 2003; Trail, Robinson et al., 2003; Woo et al., 2009) due to the unique aspects of our research context (i.e., Formula 1 racing as the sport and COTA as the venue). We employed previous research as the foundation to build our model and used many of the variables included in previous studies. Not surprisingly, variables that were previously examined by sport marketing researchers explained a majority of the variance in this model. As a result, the newly created factors and aspects of the model did explain some variance, but it was small relative to previously investigated factors.

Limitations and Future Directions

Since we formulated this model and some of these variables specifically for Formula 1 racing and this venue, this particular model is not generalizable to other sports or even to other types of racing without modifications. We are also limited by the fact that the respondents were all at the venue and thus were sufficiently motivated and attached to want and be able to attend. In addition, there are other statistical models that are plausible; however, the theory and extant research justified we employ the current model so we limited ourselves to testing only the four versions of it. Further, we believe this sample is representative of the population of attendees at the Formula 1 U.S. Grand Prix due to the random selection of respondents; however, if it is not then this is certainly a limitation.

Future researchers should examine whether there are unique motives and/or points of attachment specific to the investigated environment. In addition, it is necessary to investigate whether there are multiple aspects of conative loyalty or that this research was unique in that regard. Finally, researchers need to continue to assess the varied relationships among motives and points of attachment. It is now relatively apparent that motives do in fact impact points of attachment, which seemingly mediate (or perhaps moderate, as per Kim, Trail, and Magnusen [2013]) the relationship to conative loyalty. Nonetheless, both motives and points of attachment should be incorporated in a more comprehensive model of conative loyalty; a combination of what Harrolle et al. (2010) did and what we have done here.

Khalid Ballouli, PhD, is an assistant professor in the Department of Sport and Entertainment Management at the University of South Carolina. His research interests include sport consumer behavior, music in contemporary sport, and branding.

Galen T. Trail, PhD, is a professor of sport management in the Department of Sport Administration and Leadership at Seattle University. His research interests include sport consumer behavior and marketing sustainability through sport.

Todd C. Koesters, JD, MSA, is an assistant professor in the Department of Sport and Entertainment Management at the University of South Carolina. His research interests include sport consumer behavior, sales and marketing, and sponsorship and ROI/ROO.

Matthew J. Bernthal, PhD, is an associate professor in the Department of Sport and Entertainment Management at the University of South Carolina. His research interests include consumer behavior in sport and entertainment, and marketing ethics.

References

Alberstadt, B. (2014, June 17). Top 10 most beautiful Formula 1 destinations. Retrieved from http://www.therichest.com/sports/racing-sports/top-10most- beautiful-formula-1-destinations/

Bodet, G., & Bernache-Assolant, I. (2011). Consumer loyalty in sport spectatorship services: The relationships with consumer satisfaction and team identification. Psychology & Marketing, 28, 781-802.

Branscombe, N. R., & Wann, D. L. (1991). The positive and self-concept consequences of sports team identification. Journal of Sport & Social Issues, 15, 115-127.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers.

Crosby, L. A., & Taylor, J. R. (1983). Psychological commitment and its effects on post-decision evaluation and preference stability among voters. Journal of Consumer Research, 9, 413-431.

Dick, A. S., & Basu, K. (1994). Customer loyalty: Toward an integrated conceptual framework. Journal of the Academy of Marketing Science, 22, 99-113.

DiZinno, T. (2013, November 17). More than 113,000 Sunday, 250,000 for weekend, in year two at COTA. Retrieved from http://motorsportstalk.nbcsports.com/2013/11/17/more-than-113000-sunday-250000- for-weekend-in-year-2-at-cota/

Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX: Harcourt Brace Jovanovich College Publishers.

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, 195-207.

Fisher, R. J., & Wakefield, K. (1998). Factors leading to group identification: A field study of winners and losers. Psychology and Marketing, 15, 23-40.

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

Funk, D. C., Mahony, D. F., Nakazawa, M., & Hirakawa, S. (2001). Development of the sport interest inventory (SII): Implication for measuring unique consumer motives at team sporting events. International Journal of Sports Marketing and Sponsorship, 3, 291-316.

Gantz, W., & Wenner, L. A. (1995). Fanship and the television sports viewing experience. Sociology of Sport Journal, 12, 56-74.

Golson, J. (2014, March 11). The new F1 cars sound like living, breathing machines. Retrieved from http://www.wired.com/2014/03/2014-f1-sound/

Gray, G. T., & Wert-Gray, S. (2012). Customer retention in sports organization marketing: Examining the impact of team identification and satisfaction with team performance. International Journal of Consumer Studies, 36, 275-281.

Gutman, B. (2012, November 16). How brands are activating theirs sponsorships at Formula 1. Retrieved from http://www.forbes.com/sites/marketshare/2012/11/16/how-brands-are-activating- their-sponsorships-at-Formula1/

Gwinner, K., & Swanson, S. R. (2003). A model of fan identification: Antecedents and sponsorship outcomes. Journal of Services Marketing, 17, 275-294.

Hair, J., Black, W., Babin, B., Anderson, R., & Tatham, R. (2006). Multivariate data analysis (6th ed.). Upper Saddle River, NJ: Pearson Educational International.

Harrolle, M. G., Trail, G. T., Rodriguez, A., & Jordan, J. S. (2010). Conative loyalty of Latino and Non-Latino professional baseball fans. Journal of Sport Management, 24, 456-471.

Hawkins, D. I., Best, R. J., & Coney, K. A. (2004). Consumer behavior: Building marketing strategy. Boston, MA: McGraw-Hill Irwin.

Jensen, J. A., Cobbs, J., & Groza, M. D. (2014). The niche portfolio strategy to global expansion: The influence of market resources on demand for Formula One racing. Journal of Global Marketing, 27, 247-261.

Johnson, D. (2014, March 27). Sebastian Vettel says sound of Formula One's new engines is s***. Retrieved from http://www.telegraph.co.uk/sport/ motorsport/formulaone/sebastian-vettel/10726491/Sebastian-Vettelsays-sound-of- Formula-Ones-new-engines-is-s.html

Kim, Y. K., Trail, G. T., & Magnusen, M. J. (2013). Transition from motivation to behaviour: examining the moderating role of identification (ID) on the relationship between motives and attendance. International Journal of Sports Marketing & Sponsorship, 14, 190-211.

Kwon, H. H., Trail, G. T., & Anderson, D. F. (2005). Are multiple points of attachment necessary to predict cognitive, affective, and conative, or behavioral loyalty? Sport Management Review, 8, 255-270.

Matsuoka, H., Chelladurai, P., & Harada, M. (2003). Direct and interaction effects of team identification and satisfaction on intention to attend games. Sport Marketing Quarterly, 12, 244-253.

Milne, G. R., & McDonald, M. A. (1999). Sport marketing: Managing the exchange process. Sudbury, MA: Jones and Bartlett.

Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63, 33-44.

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

Robinson, M. J., Trail, G. T., & Kwon, H. (2004). Motives and points of attachment of professional golf spectators. Sport Management Review, 7, 167-192.

Schiffman, L. G., & Kanuk, L. L. (2004). Consumer behavior (8th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.

Sernovitz, A. (2009). Word of mouth marketing: How smart complaints get people talking. New York, NY: Kaplan.

Shapiro, S., Ridinger, L., & Trail, G. T. (2013). An analysis of multiple spectator consumption behaviors, identification, and future behavioral intentions within the context of a new college football program. Journal of Sport Management, 27, 130-145.

Shreffler, M. B., & Ross, S. D. (2013). The word-of-mouth phenomenon: Its presence and impact in a sport setting. International Journal of Sport Communication, 6, 1-18.

Sloan, L. R. (1989). The motives of sports fans. In J. H. Goldstein (Ed.), Sports, games, and play: Social & psychological viewpoints (2nd ed.; pp. 175-240). Hillsdale, NJ: Lawrence Erlbaum Associates.

Stryker, S., & Burke, P. J. (2000). The past, present, and future of an identity theory. Social Psychology Quarterly, 63, 284-297.

Sylt, C. (2014, August 23). Formula One's billion dollar man. Retrieved from http://www.forbes.com/sites/csylt/2014/08/23/formula-ones-billion-dollar-man/

Sylt, C. (2015, February 1). Formula 1 loses 25 million viewers driven by switch to pay TV. Retrieved from http://www.forbes.com/sites/csylt/ 2015/02/01/f1-loses-25-million-viewers-driven-by-switch-to-pay-tv/

Theis, M. (2014, November 4). Attendance down again at Formula 1. Retrieved from http://www.bizjournals.com/austin/news/2014/11/ 04/attendance-down-again-at-Formula 1.html

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, 98-112.

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

Trail, G. T., & James, J. D. (2001). The motivation scale for sport consumption: Assessment of the scale's psychometric properties. Journal of Sport Behavior, 24, 108-127.

Trail, G. T., Kwon, H., & Anderson, D. F. (2009). Does advertising mitigate the negative effects of losing on satisfaction and conative aspects of sport attendance? A case study in intercollegiate athletics. International Journal of Sport Communication, 2, 109-124.

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

Wann, D. L. (1995). Preliminary validation of the sport fan motivation scale. Journal of Sport and Social Issues, 19, 377-396.

Wann, D. L. (1997). Sport psychology. Upper Saddle River, NJ: Prentice Hall.

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

Wann, D. L., Ensor, C. L., & Bilyeu, J. K. (2001). Intrinsic and extrinsic motives for originally following a sport team and team identification. Perceptual and Motor Skills, 93, 451-454.

Wann, D. L., Grieve, F. G., Zapalac, R. K., & Pease, D. G. (2008). Motivational profiles of sport fans of different sports. Sport Marketing Quarterly, 17, 6-19.

Wann, D. L., Schrader, M. P., & Wilson, A. M. (1999). Sport fan motivation: Questionnaire validation, comparisons by sport, and relationship to athletic motivation. Journal of Sport Behavior, 22, 114-139.

Weaver, P. (2013, May 24). Seven F1 teams face crisis over looming bill for extra [pounds sterling]1.32bn. Retrieved from http://www.theguardian.com/sport/2013/may/24/f1-teams-crisis-bill-costs

Weaver, P. (2014, November 2). Teams going bankrupt in F1 is nothing new but solution is needed. Retrieved from http://www.theguardian.com/ sport/2014/nov/02/formula-one-bankrupt-marussia-caterham

Woo, B., Trail, G. T., Kwon, H. H., & Anderson, D. F. (2009). Testing models of motives and points of attachment. Sport Marketing Quarterly, 18, 38-53.

Wu, S. H., Tsai, C. Y. D., & Hung, C. C. (2012). Toward team or player? How trust, vicarious achievement motive, and identification affect fan loyalty. Journal of Sport Management, 26, 177-191.

Yoshida, M., Gordon, B., Nakazawa, M., & Biscaia, R. (2014). Conceptualization and measurement of fan engagement: Empirical evidence from a professional sport context. Journal of Sport Management, 28, 399-417.

Yoshida, M., Heere, B., & Gordon, B. (2014). Predicting behavioral loyalty through community: Why other fans are more important than our own intentions, our satisfaction, and the team itself. Journal of Sport Management, 29, 318-333.

Address author correspondence to:

Khalid Ballouli

University of South Carolina

Carolina Coliseum, Room 2026-N

Columbia, SC 29208

Email: ballouli@sc.edu

Table 1
Factor Loadings, Confidence Intervals, Alpha Coefficients, and AVE
Values

                                    [beta]   CI          Alpha   AVE

Acquisition of Knowledge                                 .845    .689
It provides opportunities to        .729     .655-.803
increase my knowledge about the
sport
It provides opportunities to        .832     .780-.885
increase my understanding of the
strategy
It provides opportunities to        .918     .880-.956
learn about the technical aspects
of the sport
Physical Skill of Drivers                                .864    .710
The skills of the drivers           .802     .742-.862
The abilities of the drivers        .831     .777-.885
The talents of the drivers          .893     .850-.936
Escape                                                   .873    .775
It provides an escape from my       .912     .873-.951
day-to-day routine
It provides a distraction from my   .887     .844-.929
everyday activities
It provides a diversion from        .841     .790-.891
"life's little problems" for me
Nostalgia                                                .789    .632
The Grand Prix brings back          .753     .678-.829
nostalgic feelings for me
The Grand Prix reminds me of the    .778     .706-.849
history of F1 racing
The Grand Prix brings back fond     .850     .790-.909
memories for me.
Aesthetic Quality of the Circuit                         .856    .724
The circuit at COTA is beautiful    .791     .727-.855
The circuit at COTA is stunning     .883     .834-.932
The circuit at COTA is              .876     .830-.929
magnificent
Drama                                                    .851    .652
F1 is often very dramatic           .764     .696-.831
F1 is often very exciting           .837     .785-.889
F1 is often thrilling               .829     .776-.883
The rivalries between the drivers   .797     .736-.857
are intense
Technology                                               .719    .627
Recent technologies (KERS/DRS)      .763     .673-.852
have made the racing more
exciting
The improved technology in F1 has   .820     .734-.906
made it more interesting
New Engine (single item)
The new V6 engines have made F1
more exciting
Social Interaction                                       .846    .665
I love having an opportunity to     .726     .649-.803
socialize with other F1 fans.
I enjoy having an opportunity to    .961     .913-1.00
interact with other F1 fans.
It's great being able to talk to    .737     .662-.812
other F1 fans.
Success of the Team                                      .872    .677
My favorite racing team being       .817     .757-.877
successful is important to me
My favorite team winning more       .819     .760-.878
races than it loses is important
to me
My favorite team winning is most    .833     .776-.889
important
Cars                                                     .795    .674
The look of the cars (e.g., new     .711     .630-.791
noses) this year is exciting
The cars are great looking this     .874     .821-.928
year
The cars look really good this      .868     .814-.922
year
Sound                                                    .950    .880
I really like the sound of the      .936     .913-.959
new engines
This year I really enjoy the        .938     .915-.961
sound of the cars
The new engines sound great         .940     .917-.962
Attachment to the Team                                   .774    .576
I consider myself to be a "real"    .761     .685-.838
fan of a particular racing team
I would feel a loss if I had to     .670     .578-763
wasn't a fan of a particular
racing team
Being a fan of a particular         .836     .771-.901
racing team is very important to
me
Attachment to the Driver                                 .887    .750
I am more a fan of the driver       .845     .796-.895
than of his racing team
I am a fan of a specific driver     .887     .846-.929
more than I am a fan of his
racing team
I consider myself a fan of a        .866     .821-.911
certain driver rather than of his
racing team
Attachment to the Sport                                  .888    .735
First and foremost I consider       .699     .621-.777
myself an auto racing fan
Auto racing is my favorite sport    .938     .902-973
Of all sports I prefer auto         .915     .877-.954
racing
Attachment to Formula 1                                  .724    .409
I am a fan of F1 regardless of      .552     .438-.666
what teams are racing
I am not just a fan of one          .598     .491-.705
specific F1 team, but F1 in
general
I consider myself a fan of F1 and   .751     .668-.835
not just one specific team
Attachment to Place                                      .834    .664
I feel connected to the state of    .924     .870-.979
Texas
I am a Texan                        .769     .697-.842
I support Texas as a state          .740     .662-.817
Attachment to COTA                                       .721    .547
COTA is the best circuit in the     .630     .526-.734
U.S.
I am a big fan of COTA              .797     .716-.879
Being a fan of COTA is very         .780     .697-.864
important to me
Conative Loyalty (separate items)
I am likely to come to the
Formula 1 U.S. Grand Prix next
year
I am likely to buy team licensed
merchandise in the future
I am likely to continue to
support my favorite F1 team in
the future
I will tell people about my
experiences here at this race

Table 2
Correlations (below the diagonal), Squared Correlations (above the
diagonal), and AVE Values (on the diagonal)

                1      2      3       4      5      6

Aesthetics    .689   .144   .334    .167   .098   .066
Cars          .379   .674   .046    .227   .026   .108
Drama         .578   .215   .652    .244   .140   .005
Technology    .409   .476   .494    .627   .123   .027
Escape        .313   .161   .374    .351   .775   .030
Attachment    .257   .328   .073    .164   .172   .664
to place
Attachment    .634   .317   .496    .401   .359   .414
to COTA
Attachment    .249   .107   .371    .302   .311   .017
to Driver
Attachment    .332   .090   .462    .310   .286   -.066
to Sport
Attachment    .317   .109   .446    .376   .331   .050
to Team
Attachment    .428   .097   .639    .355   .367   .026
to FI
Acquisition   .434   .260   .548    .483   .466   .115
Knowledge
Nostalgia     .347   .181   .466    .345   .298   .133
Physical      .566   .246   .666    .438   .450   .126
Skill
Social        .449   .186   .556    .485   .406   .198
Interaction
Sound         .058   .500   -.031   .319   .060   .305
Success       .268   .193   .420    .293   .293   .038
of Team
New Engine    .110   .363   .035    .322   .086   .198
Attendance    .363   .120   .439    .343   .242   .248
Intention
Merchandise   .362   .063   .570    .299   .235   .004
Intention
Support       .313   .181   .534    .411   .343   .067
Intention
WOM           .472   .193   .470    .338   .286   .081
Intention

                7      8      9       10      11      12

Aesthetics    .402   .062   .110    .100    .183    .188
Cars          .100   .011   .008    .012    .009    .068
Drama         .246   .138   .213    .199    .408    .300
Technology    .161   .001   .096    .141    .126    .233
Escape        .129   .097   .082    .110    .135    .217
Attachment    .171   .000   .004    .003    .001    .013
to place
Attachment    .547   .129   .138    .202    .205    .229
to COTA
Attachment    .359   .750   .179    .292    .159    .276
to Driver
Attachment    .371   .423   .735    .248    .348    .222
to Sport
Attachment    .449   .540   .498    .576    .162    .334
to Team
Attachment    .453   .399   .590    .403    .409    .287
to FI
Acquisition   .479   .525   .471    .578    .536    .689
Knowledge
Nostalgia     .439   .366   .440    .449    .479    .488
Physical      .537   .437   .478    .488    .633    .589
Skill
Social        .484   .337   .316    .459    .457    .469
Interaction
Sound         .029   .056   -.122   -.051   -.109   .121
Success       .446   .559   .401    .757    .353    .462
of Team
New Engine    .137   .159   .070    .091    .051    .212
Attendance    .508   .212   .263    .237    .434    .298
Intention
Merchandise   .426   .451   .541    .552    .593    .511
Intention
Support       .490   .495   .472    .671    .509    .484
Intention
WOM           .429   .265   .268    .240    .478    .362
Intention

               13     14      15      16      17     18

Aesthetics    .120   .320    .202    .041    .072   .012
Cars          .033   .061    .035    .250    .037   .132
Drama         .217   .444    .309    .001    .176   .001
Technology    .119   .192    .235    .102    .086   .104
Escape        .089   .203    .165    .004    .086   .007
Attachment    .018   .016    .039    .093    .001   .039
to place
Attachment    .193   .288    .234    .001    .199   .019
to COTA
Attachment    .134   .191    .114    .003    .312   .025
to Driver
Attachment    .194   .228    .100    .015    .161   .005
to Sport
Attachment    .202   .238    .211    .003    .573   .008
to Team
Attachment    .229   .401    .209    .012    .125   .003
to FI
Acquisition   .238   .347    .220    .015    .213   .045
Knowledge
Nostalgia     .632   .190    .131    .010    .139   .039
Physical      .436   .710    .283    .000    .184   .010
Skill
Social        .362   .532    .665    .000    .135   .003
Interaction
Sound         .102   -.009   -.003   .880    .001   .388
Success       .373   .429    .368    -.038   .677   .001
of Team
New Engine    .197   .099    .059    .623    .033   NA
Attendance    .344   .371    .465    -.039   .217   -.001
Intention
Merchandise   .430   .487    .443    -.161   .516   .001
Intention
Support       .330   .530    .473    -.053   .639   .054
Intention
WOM           .268   .432    .398    -.056   .229   -.028
Intention

               19     20     21     22     M     SD

Aesthetics    .132   .131   .098   .223   5.8   1.22
Cars          .014   .004   .033   .037   5.1   1.32
Drama         .193   .325   .285   .221   5.6   1.16
Technology    .118   .089   .169   .114   5.0   1.36
Escape        .059   .055   .118   .082   5.3   1.56
Attachment    .062   .000   .004   .007   4.6   1.94
to place
Attachment    .258   .181   .240   .184   5.2   1.27
to COTA
Attachment    .045   .203   .245   .070   4.7   1.57
to Driver
Attachment    .069   .293   .223   .072   5.1   1.72
to Sport
Attachment    .056   .305   .450   .058   4.5   1.49
to Team
Attachment    .188   .352   .259   .228   5.6   1.11
to FI
Acquisition   .089   .261   .234   .131   5.2   1.32
Knowledge
Nostalgia     .118   .185   .109   .072   4.5   1.48
Physical      .138   .237   .281   .187   5.6   1.22
Skill
Social        .216   .196   .224   .158   5.3   1.26
Interaction
Sound         .002   .026   .003   .003   3.8   2.02
Success       .047   .266   .408   .052   4.7   1.48
of Team
New Engine    .000   .000   .003   .001   3.9   1.81
Attendance    NA     .147   .167   .213   5.7   1.61
Intention
Merchandise   .383   NA     .343   .157   5.2   1.58
Intention
Support       .409   .586   NA     .178   5.4   1.50
Intention
WOM           .461   .396   .422   NA     6.3   1.18
Intention

Table 3
Path Coefficients and Fit Indices Across Different Models

                                     Model    Model    Model    Model
                                      1        2        3        4
Path                                [beta]   [beta]   [beta]   [beta]

Escape [left and right arrow]       .320     .323     .323     .323
Success
Escape [left and right arrow]       .339     .345     .345     .345
Product/Place Motives
Escape [left and right arrow]       .540     .543     .543     .543
Spectator Motives
Success [left and right arrow]      .279     .283     .284     .283
Product/Place Motives
Success [left and right arrow]      .659     .663     .663     .663
Spectator Motives
Product/Place Motives [left and     .561     .564     .564     .564
right arrow] Spectator Motives
Success [right arrow] Attachment    .767     .774     .776     .770
to Driver
Attachment to Driver [right         .717     .724     .731     .718
arrow] Attachment to Team
Product/Place Motives [right        .441     .407     .408     .408
arrow] Attachment to Place
Attachment to Place [right arrow]   .469     .459     .462     .442
Attachment to COTA
Spectator Motives [right arrow]     .856     .858     .858     .859
Attachment to Racing
Attachment to Racing [right         .663     .671     .667     .666
arrow] Attachment to F1
Attachment to Team [right arrow]    -.040
Intention to attend F1 U.S Grand
Prix next year
Attachment to Team [right arrow]             .374
Intention to buy team merchandise
in the future
Attachment to Team [right arrow]                      .667
Intention to continue to support
my favorite F1 team in the future
Attachment to Team [right arrow]                               .077
WOM Intention
Attachment to COTA [right arrow]    .566
Intention to attend F1 U.S Grand
Prix next year
Attachment to COTA [right arrow]             .202
Intention to buy team merchandise
in the future
Attachment to COTA [right arrow]                      .209
Intention to continue to support
my favorite F1 team in the future
Attachment to COTA [right arrow]                               .433
WOM Intention
Attachment to F1 [right arrow]      .194
Intention to attend F1 U.S Grand
Prix next year
Attachment to F1 [right arrow]               .387
Intention to buy team merchandise
in the future
Attachment to F1 [right arrow]                        .064
Intention to continue to support
my favorite F1 team in the future
Attachment to F1 [right arrow]                                 .159
WOM Intention
[R.sup.2]
  Attachment to Driver              58.9%    59.9%    60.1%    59.4%
  Attachment to Team                51.4%    52.4%    53.4%    51.5%
  Attachment to Place               19.4%    16.5%    16.7%    16.6%
  Attachment to COTA                22.0%    21.1%    21.4%    19.5%
  Attachment to Racing              44.0%    45.1%    44.5%    44.4%
  Attachment to F1                  73.3%    73.6%    73.7%    73.8%
  Intention to attend F1 U.S        37.0%
  Grand Prix next year
  Intention to buy team                      40.7%
  merchandise in the future
  Intention to continue to                            52.1%
  support my favorite F1
  team in the future
  WOM Intention                                                23.3%
Model Fit RMSEA                     .083     .083     .083     .084
  [chi square]/df                   1.93     1.93     1.94     1.97