Fan community identification: an empirical examination of its outcomes in Japanese professional sport.
Yoshida, Masayuki ; Gordon, Brian ; Heere, Bob 等
Introduction
Many sport fans come together at stadiums in order to enrich their
social ties with others by sharing communal fan experiences such as
talking, clapping, singing, or yelling (Melnick, 1993). Researchers have
suggested sport fans do not only develop vertical ties to their team
(team identification), but also horizontal ties to other team fans (Katz
& Heere, 2013). The importance of these horizontal relationships
between consumers was emphasized by Yoshida, Heere, and Gordon (in
press). They found attachment to the fan community, instead of the
identification with the team itself, as the only significant predictor
of actual live attendance over time. This indicates the importance of
strengthening fan communities, as gate receipts represent one of the
most significant sources of revenue for teams. By developing fan
communities around a sport team, the team may also potentially improve
the stadium atmosphere at sporting events (Melnick, 1993). Therefore,
sport teams routinely encourage fan communities--for example, the Green
Bay Packers "Cheeseheads" in the United States, Newcastle
United "Magpies" in England, FC Barcelona "Blue
Elephant" in Spain, and the Hanshin Tigers "Hanshin Fans"
in Japan--where sport fans come together, co-create social experiences,
and build camaraderie and friendship with other fans (Hunt, Bristol,
& Bashaw, 1999; Oliver, 1999).
Although both academicians and practitioners have recognized the
potential benefits associated with interpersonal relationships in sport
fan communities, at least three important concerns in previous research
limit our understanding. First, there is still much to learn about how
one conceptualizes and measures sport fans' feelings of friendship
and camaraderie in sport fan communities. Researchers in the marketing
field have conceptualized consumers' sense of brand community as
brand community identification (Algesheimer, Dholakia, & Herrmann,
2005; Keller, 2003). However, investigations of brand community
identification have not focused on its potential applicability to the
construct of fan community identification in the sport context. Within
sport marketing, important issues in relation to sport fan communities
have been examined primarily in qualitative research (Grant, Heere,
& Dickson, 2011; Katz & Heere, 2013). Due to the nature of
qualitative inquiry, the collective feelings of those involved with
sport fan communities have not been evaluated using the same criteria
across different sport contexts.
Second, team identification and fan identification have often been
used interchangeably in the literature (Gwinner & Swanson, 2003;
Sutton, McDonald, Milne, & Cimperman, 1997). However, Mahony and
colleagues (2002) posited that an individual can be devoted to multiple
points of attachment revolving a sport team brand. One perspective on
different types of fan identification can be derived from the
distinction between self- and communal-brand connections (Rindfleisch,
Burroughs, & Wong, 2008). Self-brand connection is defined as the
extent to which a consumer incorporates a brand into his or her
self-concept (Rindfleisch et al., 2008). Team identification is similar
to the idea of self-brand connection and refers to a sport fan's
perceived connectedness to a sport team and the tendency to experience
the team's successes and failures as one's own (Gwinner &
Swanson, 2003). In contrast, communal-brand connection is a
consumer's sense of belonging to a brand community (Keller, 2003).
In spectator sport, fans derive positive psychological benefits (i.e.,
friendship and camaraderie) from membership in fan communities. By
conceptualizing the brand community as a triangle, Muniz and
O'Guinn (2001) referred to the self-brand connection as the
vertical tie between consumer and organization, and the communal-brand
connection as the horizontal tie between consumers. While the self-brand
connection has received considerable attention in the field of sport
marketing, a clear understanding of the communal-brand connection has
yet to be achieved. Therefore, the conceptual focus of this study is
centered on fan community identification, which distinguished between
perceived oneness with a fan community and perceived oneness with a
sport team. By conceptualizing fan community identification and
examining its consequences, we attempted to extend previous sport
marketing research that is based mainly on team identification.
Third, limited attention has been devoted to a moderator analysis
that identifies which fans are more or less influenced by the management
of sport fan com munities. From a managerial standpoint, it is important
to understand what variables highlight the influence of a sport fan
community on its members' communal behavior (Algesheimer et al.,
2005; Carlson, Suter, & Brown, 2008). While marketing strategies
characterized by brand and relationship marketing efforts are thought to
be effective in the development of brand communities (Keller, 2003;
Rust, Zeithaml, & Lemon, 2000), little is known about the roles of
team brand equity and relationship-building programs (e.g., fan loyalty
programs) in the context of sport fan communities.
Given the limitations of previous research, the purposes of this
study were to (1) develop a model of fan community identification that
included outcome and moderator variables, and (2) examine the
relationships between the proposed constructs at professional sporting
events. In order to achieve our objectives, the setting we chose was
Japanese professional sport. In Japan, there are two major professional
sport leagues: the Japan Professional Football League (J. League) and
Nippon Professional Baseball (NPB). We attempted to examine fan
communities among those attending sporting events of both leagues. We
selected the Japanese professional sport context because (1) J. League
and NPB teams were excellent illustrations of fan communities with a
strong presence of rituals, traditions, and history; (2) fans of the two
leagues had a rich cultural world with their own fight songs,
ceremonies, and group movements; and (3) all study constructs were
readily identified and assessed in both settings.
Conceptual Background
Defining a Sport Fan
A sport fan is defined as "an enthusiastic devotee of some
particular sport consumptive object" (Hunt et al., 1999, p. 440).
Because the object of devotion underlying fan behavior can vary from fan
to fan (Mahony et al., 2002), there is a growing interest in an
extensive understanding of enthusiastic sport fans. One important
perspective on a typology of sport fans is to classify individuals into
the following five segments: temporary, local, devoted, fanatical, and
dysfunctional (Hunt et al., 1999). While temporary and local fans use
being a sport fan as a peripheral object for self-definition, devoted,
fanatical, and dysfunctional fans see sport-related objects as more
central to their self-concept (Hunt et al., 1999). Furthermore,
fanatical fans are different from the other types. Fanatical fans engage
in a number of behaviors such as body paint, costumes, signs, fight
songs, and group movement (Decrop & Derbaix, 2010; Hunt et al.,
1999). Because fanatical fans tend to exhibit these behaviors that are
supportive of particular sport-related objects (e.g., sport, team, or
player), their behaviors are likely to be accepted by others (e.g.,
family, friends, and other fans; Hunt et al., 1999). Therefore, the
source of fanatical fans' engagement in the aforementioned
behaviors is not only self-identification with a particular
sport-related object, but also communal identification with others who
also support the same sport consumptive object. In this study, our
conceptual focus was on the degree to which sport fans were fanatical on
the basis of their psychological connection to the focal sport team
(i.e., team brand equity), sense of camaraderie formed with other fans
(i.e., fan community identification), and fan-like behavior (i.e., fan
behavior that supports sport-related objects).
Defining a Sport Fan Community
A sport fan community is a specific form of brand community in the
sport context. A brand community is defined as a specialized,
non-geographically bound community based on the relationships among
consumers of a brand (Muniz & O'Guinn, 2001). In the
contemporary marketplace, one can witness brand communities in multiple
product categories due to a rich variety of self-expressive products,
including cars (Algesheimer et al., 2005), motorcycles (McAlexander,
Schouten, & Koenig, 2002), computers (Muniz & O'Guinn,
2001), and spectator sport teams (Grant et al., 2011; Katz & Heere,
2013; Yoshida et al., in press). Focusing on the communal aspect of
sport fans, Oliver (1999) considered the social bonding of a sport fan
community as a blend of personal identity with the cultural milieu
surrounding the focal sport team. Given this implication, a sport fan
community can be defined as a specialized, non-geographically bound
community based on sport fans' personal identity with the cultural
milieu surrounding a specific sport team.
In addition to three markers of brand community (consciousness of
kind, moral responsibility, and rituals and traditions), Muniz and
O'Guinn (2001) further contend that "brand communities can be
relatively stable groupings, with relatively strong (but rarely extreme)
degrees of commitment" (p. 415). On the contrary, we see fan
communities as more diverse groups of people who are high and low in
team commitment. Muniz and O'Guinn discuss brand communities based
on vertical relationships (consumer organization) and horizontal
relationships (consumer --consumer). We contend that for some consumers
the vertical relationships are more important (e.g., team
identification) while for others the horizontal relationships are more
important (e.g., the other fans). Schouten and McAlexander (1995) noted
the diversity in commitment to a brand community and discussed the
hierarchical structure within brand communities. The inner circle of the
brand community are devoted fans of the organization, yet at the
periphery of the community there are many members who are at the event
for a multitude of reasons. For example, while the hardcore fans are
strongly committed to their favorite team, other fans' primary
attachment points might be a particular player, sport, and local city
(Hunt et al., 1999; Mahony et al., 2002). Moreover, some individuals,
such as college football fans who enjoy tailgating parties, do not
necessarily identify with their team or related sport attachment points,
but they have a commitment to the university and other fans (Katz &
Heere, 2013). Thus, our argument is that in order to understand fan
communities, we not only need to understand their vertical ties to the
organization (e.g., team identification), but also their horizontal ties
to the other fans (e.g., fan community identification).
Defining Fan Community Identification
There is a commonly acknowledged conceptualization of
consumer-brand community connection. Muniz and O'Guinn (2001)
considered the consumer-brand community connection as an intrinsic
connection that brand community members feel toward one another and the
collective sense of difference from others that are not in the
community. Similarly, Keller (2003) contended identification with a
brand community may help consumers feel a kinship with others associated
with the brand. Other researchers have reached a similar conclusion that
a consumer's emotional and social bonds with a brand community can
be conceptualized as brand community identification (Algesheimer et al.,
2005; Fuller, Matzler, & Hoppe, 2008). Given this perspective, fan
community identification is defined as the intrinsic connection that fan
community members feel toward one another and the collective sense of
difference from others not in the fan community (Keller, 2003; Muniz
& O'Guinn, 2001).
Behavioral Consequences in Sport Fan Communities
Loyal sport fans engage not only in self-interested tasks (e.g.,
attending, watching, reading, and purchasing), but also in tasks that
benefit their favorite sport teams (e.g., supportive displays of sport
fandom, positive word-of-mouth, and collaborative event attendance) and
other fans (e.g., sharing knowledge about a team with other fans,
cooperative communications in the stands, and consumer-to-consumer
helping behaviors in fan communities; Decrop & Derbaix, 2010; Hunt
et al., 1999; Yoshida, Gordon, Nakazawa, & Biscaia, 2014). Such
team- and others-oriented behaviors are referred to as extra-role
behaviors (Ahearne, Bhattacharya, & Gruen, 2005; Yoshida et al.,
2014). A review of the brand community literature reveals four important
extra-role behaviors pertaining to behavioral consequences in sport fan
communities: fan community engagement, customized product use, member
responsibility, and positive word-of-mouth (Schau, Muniz, & Arnold,
2009; Woolf, Heere, & Walker, 2013).
[FIGURE 1 OMITTED]
Fan community engagement refers to consumers' escalating
behavioral involvement in a fan community that includes socially
committed behaviors such as self-expression, story-telling, and fan
community participation (Schau et al., 2009). In sport fan communities,
the key levels of community engagement include (1) staking a social
space, (2) participating in seminal events, (3) badging the milestones
for symbolic representation, and (4) documenting personal stories in a
narrative format. Customized product use is defined as consumers'
improved use of team-related products in sport fan communities. Such
behaviors include customizing (e.g., designing products to fit
one's self-concept) and commoditizing (e.g., the extensive use of
products to influence other fans to follow their favorite team). Member
responsibility refers to a felt sense of duty and obligation to a fan
community as a whole and to its individual members in order to create,
enhance, and sustain the ties among the fan community members. The key
components of member responsibility are welcoming, empathizing, and
governing. Positive word-of-mouth is defined as consumers'
external, outward focus on creating favorable impressions of a sport
team, enthusiastic fans, and the fan community in the social universe
beyond the fan community.
Hypothesis Development
Figure 1 is an illustration of the proposed fan community
identification model. Building on social identity theory (SIT; Tajfel
& Turner, 1985) and the literature on brand community identification
(Algesheimer et al., 2005; Bagozzi & Dholakia, 2006; Fuller et al.,
2008), fan community identification is hypothesized to influence team
brand equity and fan community-related consequences in the spectator
sport context. Furthermore, drawing on the consumer prosocial behavior
literature (Ahearne et al., 2005; Bettencourt, 1997), team brand equity
is supposed to foster fan community-related consequences because a
consumer's commitment to and identification with a brand is the
foundation for prosocial behavior. The framework also includes fan
loyalty program participation by examining its moderating impact on the
relationship between team brand equity and fan community-related
consequences. In the following section, we develop hypotheses within
this framework.
The Impact of Fan Community Identification
SIT (Tajfel & Turner, 1985) forms the theoretical base of the
framework and suggests that fan community identification fosters team
brand equity and fan community-related consequences. We begin by
considering the impact of fan community identification on team brand
equity that is the value added to a sport team by the brand name
(Farquhar, 1989). According to Tajfel and Turner (1985), an
individual's identification with a social group creates a social
identity that shapes the person's self-image deriving from the
social category to which he or she perceives himself or herself as a
member of the social group. This psychological state is "the
perception of oneness with or belongingness to some human
aggregate" (Ashforth & Mael, 1989, p. 21) and forms a
collective representation of who one is (Ellemers, Kortekaas, &
Ouwerkerk, 1999). Besides other fields of consumer behavior (e.g.,
consumer-company identification, Ahearne et al., 2005; team
identification, Wann & Branscombe, 1993), social identification has
also been studied in the context of brand communities (Bagozzi &
Dholakia, 2006; Fuller et al., 2008). In general, these studies provide
support for the impact of brand community identification on
brand-related outcomes such as brand commitment (Carlson et al., 2008),
brand trust (Fuller et al., 2008), and brand identification (Bagozzi
& Dholakia, 2006). Bagozzi and Dholakia (2006) suggest as a
consumer's identification with a brand community increases, greater
involvement with the brand occurs and promotes the assimilation of the
brand's image into one's self-concept. Oliver (1999) provides
additional support by suggesting consumers can be placed in
self-sustaining social environments (e.g., fan community) that reinforce
their brand commitment. These studies indicate that sport fans can be
devoted to their favorite team brands by increasing involvement with the
teams in the fan communities. Therefore, fan community identification
will serve as the means of increasing team brand equity among the fans
of the focal sport team. Derived from these arguments, we test the
following hypothesis:
[H.sub.1] Fan community identification has a positive effect on
team brand equity.
Theoretically, the results from previous research provided
additional support for the development of fan community-related
behaviors. From one perspective, a consumer's identification with
other fans strengthens his or her engagement in the fan community
(Algesheimer et al., 2005). Other researchers provided a theoretical
basis for the impact of fan community identification on a number of
community-related behaviors such as integrating and retaining other
fans, participating in team-related discussions, assisting other fans,
and providing feedback to the team for improving event experiences
(Fuller et al., 2008; Katz & Heere, 2013; Schau et al., 2009). To
make the mechanism driving fan community-related consequences more
concreate, we draw on SIT. Because identification with a social group
helps individuals develop a sense of belonging, increase their
self-esteem, raise their aspirations, and invest themselves in altruism
and unselfish behaviors (Mael & Ashforth, 2001), fan community
identification affects not only an assessment of team brand equity, but
also fans' community engagement and prosocial behaviors. Based on
this discussion, we expect fan community identification plays a key role
in achieving consumers' extra-role behaviors in fan communities
(e.g., fan community engagement, customized product use, member
responsibility, and positive word-of-mouth). Therefore, we propose:
[H.sub.2:] Fan community identification has a positive effect on
(a) fan community engagement, (b) customized product use, (c) member
responsibility, and (d) positive word-of-mouth.
The Impact of Team Brand Equity
In addition to the effect of fan community identification on fan
community-related consequences, we also hypothesize the impact of team
brand equity on the outcome variables. As suggested by Keller (2003) in
his work with the brand equity pyramid model, fans' behavioral and
social engagement is beyond team brand equity. Team brand equity
influences the attitudinal and behavioral responses of fans to the fan
community (Keller, 2003). More specifically, the literature on brand
community leads us to conclude that high levels of team brand equity are
likely to engender high levels of fan community-related behaviors such
as participation in the fan community, group behavior, and co-creation
(Algesheimer et al., 2005; Fuller et al., 2008; Oliver, 1999). This
reasoning is in line with research revealing prosocial outcomes of
consumers' brand commitment (Bettencourt, 1997; Jones, Taylor,
& Bansal, 2008). As an example, brand commitment has been shown to
be positively related to prosocial behaviors including altruism (Jones
et al., 2008), cooperation (Bettencourt, 1997), participation in service
delivery (Bettencourt, 1997), and advocacy (Jones et al., 2008). Because
we include team brand equity as an antecedent of four fan
community-related behavioral consequences (fan community engagement,
customized product use, member responsibility, and positive
word-of-mouth) in the proposed framework, the following hypothesis is
derived:
[H.sub.3:] Team brand equity has a positive effect on (a) fan
community engagement, (b) customized product use, (c) member
responsibility, and (d) positive word-of-mouth.
Moderating Effects of Fan Loyalty Program Participation
There is little known about the role the organization can play in
enhancing the relationship between consumers. Managers often appear
reluctant to do anything to encourage these relationships, claiming they
are concerned about getting in the way of an organic process (Katz &
Heere, 2013). One strategy that might be appropriate in this context is
a fan loyalty program, which might facilitate recurring behavior and
could encourage consumer interaction (Rosenbaum, 2008). One conclusion
drawn from the existing literature is that consumers' fan loyalty
program participation moderates the link between consumers'
attitudinal and behavioral responses (Evanschitzky & Wunderlich,
2006; Seiders, Voss, Grewal, & Godfrey, 2005). Evanschitzky and
Wunderlich (2006) found relationship program participants were more
likely to allocate resources (money, time, and effort) to service
providers that offer superior customer value. In the case of loyalty
programs that provide consumers with social interaction opportunities
rather than simple financial incentives, such communal programs elicit
various types of social support from other consumers: intimate
interaction, social participation, physical assistance, feedback,
guidance, and material aid (Rosenbaum, 2008). By providing superior
value, fan loyalty program participation will promote social behavior
among participants. Because brand equity is a significant defining
element of customer value (Rust et al., 2000), we propose the following
hypothesis:
[H.sub.4]: The effects of team brand equity on (a) fan community
engagement, (b) customized product use, (c) member responsibility, and
(d) positive word-of-mouth are stronger for fan loyalty program
participants than for non-participants.
Control Variables
In addition to the main effects of the proposed fan community
identification framework, other variables might influence the outcome
variables. According to the theory of planned behavior, a
consumer's past behavior can explain his or her actual behavior
(Ajzen, 1991). Due to individuals' psychological commitment to
habitual behavior and their desire to minimize the cost of thinking,
consumers prefer to attend sporting events of the same team they
followed before. Therefore, we control for consumers' past
attendance frequency in the current season and length of time as a fan.
Both variables have been proposed to be influential for consumer
decision-making in the sport context (Nakazawa, Mahony, Funk, &
Hirakawa, 1999).
Method
Research Setting and Sample
In fan communities, the role of sporting events is associated with
the idea of "brandfests" (McAlexander et al., 2002).
Brandfests are corporate-sponsored events where consumers come together
to experience and celebrate brand ownership (McAlexander et al., 2002).
Because fans attend sporting events to support and celebrate the
successes of their team, sporting events can be viewed as a type of
brandfest in the sport setting. Conceptual and empirical support for the
phenomenon that sporting events have a role as brandfests is provided by
Katz and Heere (2013) and Woolf et al. (2013). In this study, we
examined the psychology and behavior of spectators who attended sporting
events to support their favorite teams.
This study was conducted in two major professional sport settings
in Japan: professional soccer and professional baseball. First, data
were collected from spectators attending a J. League Division I game
([N.sub.total attendance] = 9,550) in a large city in west Japan. We
conducted data collection in all seating sections (except for the
section of the opposing team's fans) and used a mixture of
convenience and proportionate sampling, which was stratified by both age
and gender. Questionnaires were distributed in the stands prior to the
start of the game. Brief instructions were given to the respondents
about the purpose of the study, voluntary participation, and
confidentiality of the data. Before distributing the questionnaires, 20
trained surveyors observed an assigned block of the stands in order to
estimate the percentage of those attending based on gender (male/female)
and age (ages between 18-29, 30-49; and 50 and older). From the 440
questionnaires distributed, 427 were returned for a response rate of
97.0%. Thirteen questionnaires were not returned. Among the
questionnaires returned, 26 were rejected because many items were left
blank, yielding a usable response rate of 91.1% (n = 401). Of the soccer
sample, 62.4% of the respondents were male. Approximately one-third of
the subjects were in the 30-39 age range (30.5%), 27.2% were between
40-49 years old, 18.1% were between 20-29 years old, and 17.6% were 50
years old and older.
For the baseball sample, data were collected from spectators
attending a professional baseball game ([N.sub.total attendance] =
19,087) in the eastern Tokyo metropolitan area. Questionnaires were
distributed to individuals outside the stadium prior to the start of the
baseball game. In order to collect data as systematically as possible,
one of the authors estimated when, where, and how many people would be
present at various locations around the stadium based on observations of
previous games. Twelve trained surveyors approached potential
respondents in the assigned locations. The surveyors provided
instructions on the study purpose, voluntary participation, and
confidentiality of the results. Of the 360 questionnaires distributed,
347 were returned for a response rate of 96.4%. Among the 347 forms
returned, 21 were incomplete, yielding a final usable response rate of
90.6% (n = 326). Of the baseball sample, 68.4% of the subjects were
male. One-third of the respondents were in the 30-39 age range (32.8%),
28.2 were between 40-49 years old, and 24.5% were between 20-29 years
old.
In order to verify the representativeness of our samples, we
compared our samples with the general population. According to the J.
League Annual Survey Report (J. League, 2013), which was based on the
data collected from 17,286 game attendees of all teams, the gender
distribution of the national population (male = 62.6%, female = 37.4%)
corresponded to that of our sample (male = 62.4%, female = 37.6%). Even
though such data were not available for the baseball sample, the gender
distribution (male = 68.4%, female = 31.6%) was almost parallel to that
of the soccer sample. Therefore, our samples were thought to be an
adequate representation of the overall population to generate data for
this study.
Measurement
We adapted items from previous research to measure fan community
identification (Keller, 2003) and team brand equity (Brady, Cronin, Fox,
& Roehm, 2008). In order to measure positive word-of-mouth, a
three-item scale was adapted from previous research (Jones &
Reynolds, 2006), and the wording was modified to reflect the sport fan
community context. Also, we developed new scale items to measure fan
community engagement, customized product use, and member responsibility.
The information supporting these factors was primarily conceptual (Schau
et al., 2009) and no established scale was available in the existing
literature. Therefore, an initial pool of 17 items was generated for
these factors based on the construct definitions. In order to assess
content validity, four researchers from four different universities were
asked to rate each statement as being "Not Representative
(0)," "Somewhat Representative (1)," and "Clearly
Representative (2)" of the dimension (Tian, Bearden, & Hunter,
2001). Items evaluated as clearly representative by three reviewers, and
as no worse than somewhat representative by a fourth reviewer were
retained. Also, the judges were asked to provide suggestions for
changing words and phrases in the items. This process eliminated seven
items, leaving 10 items.
Finally, we included two control variables that might influence
behavioral responses in the fan community: past attendance frequency and
length of time as a fan (Figure 1). Past attendance frequency was
measured by the number of games attended in the current season (Yoshida
& James, 2010). A spectator's length of time as a fan was
measured by the number of years being a fan of his or her favorite sport
team (Nakazawa et al., 1999).
Back Translation
As a check of meaning equivalence between the original English
instrument and the translated Japanese instrument, the survey
questionnaire was first translated into Japanese by one of the authors
and then back-translated into English by another native of Japan who is
also fluent in English. To ensure the accuracy of the translation, a
U.S.-born American citizen was asked to assess differences in meaning
between the original and back-translated instruments. The comparison of
the two forms indicated both instruments reflected the construct domain.
Results
Assessment of the Measures
Measurement Analysis. The psychometric properties of the items were
assessed through a confirmatory factor analysis (CFA) using LISREL 8.8.
The fit measures were acceptable for both samples (see Table 1). The
ratios of chi-square to degrees of freedom ([chi square]/df) were within
the acceptable range of 2 to 3 (Hu & Bentler, 1999). The comparative
fit index (CFI) and nonnormed fit index (NNFI) were greater than the
cutoff point of .90 (Hu & Bentler, 1999). The values of the root
mean square error of approximation (RMSEA) were .067 for the soccer
sample and .073 for the baseball sample were smaller than Hu and
Bentler's (1999) criterion of .08.
Scale statistics, including factor loadings ([lambda]), composite
reliability (CR), and average variance extracted (AVE) values, are
presented in Table 1. All items loaded on their respective factors, and
factor loadings ranged from .53 to .88 for the soccer sample and from
.61 to .96 for the baseball sample. In both settings, the CR values for
all factors were greater than the recommended cutoff point of .60
(Bagozzi & Yi, 1988), indicating the proposed constructs were
internally consistent. The AVE values for the proposed constructs ranged
from .50 to .67 in the soccer setting and from .52 to .83 in the
baseball setting, providing evidence of construct reliability (Fornell
& Larcker, 1981). Discriminant validity was assessed by comparing
the AVE estimate for each construct with the squared correlations
between the respective constructs (see Table 2). For the soccer sample,
in 12 cases out of a total of 15 correlations between the six latent
constructs, the AVE values were considerably greater than any squared
correlations between all pairs of the constructs. However, three cases
failed to establish discriminant validity. For the baseball sample, the
AVE values were considerably greater than any squared correlations
between all pairs of the constructs. Next, we used a chi-square
difference test using the soccer sample and compared a model in which
the correlation of each pair of the latent constructs was constrained to
be equal to one with an unconstrained model in which the correlation was
permitted to vary freely (Anderson & Gerbing, 1988). Performing a
total of 15 chi-square difference tests for all pairs, every case
demonstrated a significant difference. Collectively, our results provide
evidence for discriminant validity.
Table 2 also presents descriptive statistics (means, standard
deviations, and correlations) for the measures used in the current
study. In both settings, the mean factor scores pertaining to team brand
equity, fan community identification, and positive word-of-mouth were
slightly higher than those of the factors pertaining fan community
engagement, customized product use, and member responsibility. In terms
of the two control variables, the mean scores of both attendance
frequency (11.77) and length of time as a fan (10.05) for the baseball
sample were higher than those of the variables for the soccer sample
(attendance frequency = 8.70; length of time as a fan = 6.81) because
there were more home games in the regular season for professional
baseball (72 games) than for professional soccer (17 games), and the
professional baseball team had a longer history than the professional
soccer club.
Multi-Group Measurement Invariance. When structural paths are
examined between contexts, the results are meaningful if the measures
are invariant across groups (Steenkamp & Baumgartner, 1998).
Therefore, prior to hypothesis testing, we also tested whether the
measures were invariant across the soccer and baseball samples by
estimating three types of multi-group measurement invariance models
(Steenkamp & Baumgartner, 1998). The first model we estimated was an
unconstrained six-factor measurement model (also called configural
invariance model) across the two samples. This model consisted of the
six latent constructs (fan community identification, team brand equity,
fan community engagement, customized product use, member responsibility,
and positive word-of-mouth) and demonstrated an acceptable fit to the
data ([chi square]2 = 989.38, df = 348; CFI = .98; NNFI = .97; RMSEA =
.071). Second, we estimated a model with the matrix of factor loadings
constrained as invariant across the two samples in order to assess
metric invariance. The chi-square statistic was 1020.35 with 363 degrees
of freedom for the metric invariance model. The difference between the
configural invariance and metric invariance models was significant
([DELTA][chi square] = 30.97, Adf = 15, p < .01), indicating the
factor structure was not invariant across the two contexts.
However, because full metric invariance is rare, Steenkamp and
Baumgartner (1998) contend that meaningful comparisons of structural
paths can be made if "at least one item (other than the one fixed
at unity to define the scale of each latent construct) was metrically
invariant" (p. 81). Hence, as suggested by Steenkamp and
Baumgartner (1998), we further evaluated the modification indices (MIs)
of the constrained parameters. Based on an investigation of MIs, the
loadings of five items (TBE4, FCI2, FCE4, MR3, and WOM2) with large MIs
were released. This seemed appropriate because each construct had at
least one factor loading fixed across the two samples in addition to the
item loading set to be equal to 1 to represent the respective latent
construct (Steenkamp & Baumgartner, 1998). For this partial metric
invariance model, the chi-square statistic was 1005.25 with 358 degrees
of freedom. The difference between the fit of the partial metric
invariance model and the fit of the configural invariance model was not
significant (A[chi square] = 15.87, [DELTA]df = 15, p > .05),
providing support for partial metric invariance.
Hypothesis Testing
Structural Modeling. An examination of the hypothesized
relationships was achieved through structural equation modeling (SEM)
using LISREL 8.8 (see Table 3). In the soccer sample, the fit indices
for the hypothesized model were [chi square]/df = 4.11, CFI = .95, NNFI
= .95, RMSEA = .088. Similarly, in the baseball sample, the fit measures
for the structural model were [chi square]/df = 3.96, CFI = .96, NNFI =
.95, RMSEA = .095. Both CFI and NNFI values were greater than the cutoff
point of .90 in the two settings. Although the RMSEA values for the two
samples exceeded the acceptable threshold (.05.08), they indicated a
mediocre fit (.08-.10; Browne & Cudeck, 1993). The ratios of
chi-square to degrees of freedom ([chi square]/df), however, were higher
than the recommended cutoff point (>3.0; Hu & Bentler, 1999) in
both settings. According to Hair and colleagues (2006), the ratio of
chi-square to degrees of freedom is not meaningful when a complex model
is used to analyze data. Because we added two control variables to the
structural model, it was not surprising the ratio of [chi square]:df of
the structural model was larger than that of the measurement model.
Overall, the proposed structural model demonstrated a reasonable fit to
the data, while it was not a close fit.
With respect to hypothesis testing, fan community identification
had a positive effect on team brand equity in both soccer ([gamma] =
.53, p < .01) and baseball ([gamma] = .69, p < .01) settings.
Also, the paths from fan community identification to the four outcome
variables (fan community engagement, customized product use, member
responsibility, and positive word-of-mouth) were positive and
significant for both samples. Hence, we found support for [H.sub.1],
[H.sub.2a], [H.sub.2b], [H.sub.2c], and [H.sub.2d] from both settings.
Furthermore, the findings indicated that team brand equity did not have
a positive effect on the four outcome variables in the soccer setting.
On the other hand, the effects of team brand equity on fan community
engagement ([beta] = .22, p < .01) and positive word-of-mouth ([beta]
= .35, p < .01) were positive and significant in the baseball
setting, in partial support of [H.sub.3a] and [H.sub.3d]. In order to
check the robustness of the hypothesized effects, we also examined
whether the inclusion of consumers' two types of habitual behavior
(attendance frequency and length of time as a fan) influenced these
findings (see Table 3). In the soccer setting, attendance frequency had
significant positive effects on fan community engagement ([beta] = .18,
p < .01) and positive word-of-mouth ([beta] = .14, p < .01). The
respondents' length of time as a fan negatively influenced fan
community engagement ([beta] = -.14, p < .01). In the baseball
setting, only attendance frequency had a significant negative effect on
team brand equity ([beta] = -.10, p < .05). Based on these findings,
it is important to note team brand equity and fan community related
outcomes were more impacted by the proposed constructs than by the
control variables. Our results were robust to the inclusion of these
control variables.
The ability of the exogenous variables to explain variations in the
endogenous variables was assessed by [R.sup.2] values (see Table 3). The
[R.sup.2] values for team brand equity, fan community engagement,
customized product use, member responsibility, and positive
word-of-mouth in the soccer setting were .29, .62, .35, .63, and .77,
respectively, and those in the baseball setting were .46, .51, .19, .46,
and .66, respectively.
Moderating Effects of Fan Loyalty Program Participation. A series
of multi-group SEM analyses were conducted in order to test the
moderating effects of fan loyalty program participation ([H.sub.4a],
[H.sub.4b], [H.sub.4c], and [H.sub.4d]) on the proposed structural model
(Palmatier, Scheer, & Steenkamp, 2007). A dichotomous variable
(yes/no) was used to divide the soccer sample into two groups of fan
loyalty program participants and non-participants. (1) In each analysis,
a chis-quare difference test was used in order to compare a model in
which all hypothesized paths were constrained to be equal across the two
groups with an unconstrained model in which the hypothesized path to be
moderated was permitted to vary freely across the groups. If the
unconstrained model has a significantly lower chi-square value than the
constrained model, and if the impact is in the expected direction, the
moderating effect is evident. As Table 4 shows, the impact of team brand
equity on positive word-of-mouth was significantly stronger for fan
loyalty program participants (g = .26, p < .01) than for
non-participants ([gamma] = .04, n.s.; [DELTA][chi square] = 3.88,
[DELTA]df = 1, p < .05). Therefore, [H.sub.4d] was supported while
[H.sub.4a], [H.sub.4b], and [H.sub.4c] were rejected.
Discussion
Sport fan communities arise in numerous settings when sport
consumers participate in face-to-face, virtual, fan-initiated, or
team-initiated fan communities. In order to assess fans' collective
feelings of camaraderie and their actual behavior in sport fan
communities, we tested the proposed fan community identification model
using subjective consumer responses evoked by live experiences at two
professional sporting events in Japan. Since little effort has been made
to identify the outcome and moderator variables of fan community
identification, the current study makes a significant contribution to
the literature and practice in four different ways.
First, this study is one of the first attempts to apply the idea of
communal brand connection to the sport context. Examining the concept of
fan community identification in the sport setting extends the sport
marketing literature that has developed with self-brand connection
concepts such as team identification (Wann & Branscombe, 1993), team
attachment (Mahony et al., 2002), and team brand equity (Boyle &
Magnusson, 2007). Specifically, we conceptually differentiate fan
community identification from team identification and assume fan
community identification to be a more appropriate antecedent of the
extra-role prosocial behavior of fanatical fans: "The fanatical fan
engages in behavior that is beyond the normal devoted fan, yet the
behavior is accepted by significant others (family, friends, and other
fans) because it is considered supportive of the target - sport, team,
or player" (Hunt et al., 1999, p. 446). We proposed a conceptual
model of fan community identification and adopted the four-item scale of
communal brand connection developed by Keller (2003). The results in
both soccer and baseball settings provide evidence of construct
reliability and validity for the concept of fan community
identification.
A second contribution of the study is assessing the construct
validity of fan community-related consequences that are conceptually
relevant to the idea of fan-like behavior (Hunt et al., 1999). Although
recent qualitative studies provide evidence of descriptive validity for
the four dimensions of fan community engagement, customized product use,
member responsibility, and positive word-of-mouth (Schau et al., 2009;
Woolf et al., 2013), there has been no research to actually measure
these constructs. We developed scale items to measure the four
behavioral constructs and provided evidence of convergent and
discriminant validity in both soccer and baseball settings. Furthermore,
it is worth noting a contribution of this study to the literature is
evidence of the distinction between fan community identification and fan
community engagement. In the study of Algesheimer et al. (2005), both
community identification and community engagement were viewed as
attitudinal constructs, thereby failing to provide strong support for
discriminant validity between the two constructs. On the other hand, we
took a behavioral perspective on fan community engagement (Hunt et al.,
1999; Schau et al., 2009) and provide evidence of discriminant validity
across the two samples of soccer and baseball fans. These results
reinforce Yoshida and colleagues' (2014) recent study that found
fan engagement in nontransactional extra-role behavior should be
measured from a behavioral standpoint.
Third, the current study contributes to the sport marketing
literature by identifying the effects of fan community identification on
team brand equity and fan community-related consequences. A fundamental
assumption of the proposed relationships is that creating a sense of
friendship and camaraderie among fans (Keller, 2003) will enable the
team to increase consumers' involvement with the team brand in the
fan community (Oliver, 1999), develop strong team brand equity (Bagozzi
& Dholakia, 2006), attract more fans to the fan community (Katz
& Heere, 2013), and to receive more social support from the fans
(Rosenbaum, 2008). Based on our results, fan community identification
positively influenced team brand equity in both settings. Also, the
effects of fan community identification on fan community-related
consequences were positive, strong, and significant in both settings. On
the other hand, the direct effects of team brand equity on fan
community-related consequences were not significant in the soccer
setting and these effects were weak in the baseball setting. We suggest
from these findings that in the total sample, fans' unique social
behaviors are primarily a function of fan community identification, not
team brand equity. Furthermore, it should be noted the current study is
one of the first attempts to examine the impact of fan community
identification on team brand equity. While researchers have tested the
relationship between team identification and team brand equity (Boyle
& Magnusson, 2007), our results indicated that team brand equity
could be strengthened by consumers' social identification with the
fan community. From a practical standpoint, our results and the relevant
literature (Decrop & Derbaix, 2010; Hunt et al., 1999) suggest
fans' pre- and in-game activities such as anthems, rituals, fight
songs, group movements, and displays of team color can be used to foster
their collective feelings of enjoyment, friendship, and pride in fan
communities and eventually to receive social support from the members of
fan communities.
Fourth, this study advances our understanding of the role that fan
loyalty programs play in fostering positive word-of-mouth in fan
communities. The results of the moderator analyses indicated fan loyalty
program participants were more likely to tell others about their
positive impressions of the team itself, the other fans, and the fan
community. Within a fan base, there is a distinction between leaders who
have a strong emotional connection to a sport team and followers who
have a low emotional connection to a sport team but are strongly
attached to other fans (Katz & Heere, 2013). Fan community members
can serve as "brand advocates" who communicate with one
another and with followers and become strong referents for promoting the
social desirability of engaging in fan community-related behaviors.
Limitations and Directions for Future Research
Several limitations may influence the results of this study. First,
it is unclear whether these results would be replicated in other
cultural settings. The proposed framework was tested in the context of
Japanese professional sport, and it should be acknowledged that Japan
has a collectivist culture (Hofstede, 2001). This collectivism might
enhance the relationships between the proposed constructs. In
particular, this cultural characteristic may have inflated the
predictive power of fan community identification for behavioral
consequences. The emphasis here is on the word "may." Based on
the experiences of the authors with American, European, and New Zealand
fan communities, it might be questionable whether these relations arise
even in a less collectivist culture. Nevertheless, it will be
interesting to replicate this study in a different cultural context.
Second, the SEM results indicated the proposed model was a mediocre fit
to the data. Additional effort should be made to validate whether the
factor structure can be observed in other settings empirically. Finally,
this study only examined the moderating effect of fan loyalty program
participation on the proposed framework. A suggestion for future
research is to examine the impact of other moderators (e.g., team
identification and involvement) on the proposed framework (Seiders et
al., 2005).
The developed fan community identification model serves to advance
the study of sport marketing by examining the impact of self-team
connection (team brand equity) and communal-fan connection (fan
community identification) on extra-role fan behavior. The proposed model
and recommendations for future research provide numerous opportunities
to continue advancing our knowledge of sport fan communities.
Masayuki Yoshida, PhD, is an associate professor in the Department
of Sport Science at Biwako Seikei Sport College. His research interests
include consumer satisfaction, fan loyalty, and engagement behavior in
the sport context.
Brian Gordon, PhD, is an assistant professor in the Department of
Health, Sport, and Exercise Sciences at the University of Kansas. His
research interests include consumer behavior, brand management, and fan
loyalty.
Bob Heere, PhD, is an associate professor and PhD program director
in the Department of Sport and Entertainment at the University of South
Carolina. His research interests include social identity theory and
community development.
Jeffrey D. James, PhD, is a professor and chair of the Sport
Management Department at Florida State University. His research
interests include sport consumer psychology, sport consumer behavior,
and sport sponsorship.
References
Ahearne, M., Bhattacharya, C. B., & Gruen, T. (2005).
Antecedents and consequences of customer-company identification:
Expanding the role of relationship marketing. Journal of Applied
Psychology, 90, 574-585.
Ajzen, I. (1991). The theory of planned behavior. Organizational
Behavior & Human Decision Processes, 50, 179-211.
Algesheimer, R., Dholakia, U .M., & Hermann, A. (2005). The
social influence of brand community: Evidence from European car clubs.
Journal of Marketing, 69(3), 19-34.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equation
modeling in practice: A review and recommended two-step approach.
Psychological Bulletin, 103, 411-423.
Ashforth, B. E., & Mael, F. A. (1989). Social identity theory
and the organization. Academy of Management Review, 14, 20-39.
Bagozzi, R. P., & Dholakia, U. M. (2006). Antecedents and
purchase consequences of customer participation in small group brand
communities. International Journal of Research in Marketing, 23, 45-61.
Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of
structural equation models. Journal of the Academy of Marketing Science,
16, 74-94. Bettencourt, L. A. (1997). Customer voluntary performance:
Customers as partners in service delivery. Journal of Retailing, 73,
383-406.
Boyle, B., & Magnusson, P. (2007). Social identity and brand
equity formation: A comparative study of collegiate sports fans. Journal
of Sport Management, 21, 497-520.
Brady, M. K, Cronin, J. J., Fox, G. L., & Roehm, M. L. (2008).
Strategies to offset performance failures: The role of brand equity.
Journal of Retailing, 84, 151-164.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of
assessing model fit. In K. A. Bollen & J. S. Long (Eds.), Testing
structural equation models (pp. 136-162). Newbury Park, CA: Sage.
Carlson, B. D., Suter, T. A., & Brown, T. J. (2008) Social
versus psychological brand community: The role of psychological sense of
brand community. Journal of Business Research, 61, 284-291.
Decrop, A., & Derbaix, C. (2010). Pride in contemporary sport
consumption: A marketing perspective. Journal of the Academy of
Marketing Science, 38, 586-603.
Ellemers, N., Kortekaas, P., & Ouwerkerk, J. W. (1999).
Selfcategorization, commitment to the group, and group self-esteem as
related but distinct aspects of social identity. European Journal of
Social Psychology, 29, 371389.
Evanschitzky, H., & Wunderlich, M. (2006). An examination of
moderator effects in the four-stage loyalty model. Journal of Service
Research, 8, 330-345.
Farquhar, P. H. (1989). Managing brand equity. Marketing Research,
24-34.
Fornell, C., & Larcker, D.F. (1981). Evaluating structural
equation models with unobservable variables and measurement error.
Journal of Marketing Research, 18(1), 39-50.
Fuller, J., Matzler, K., & Hoppe, M. (2008). Brand community
members as a source of innovation. Journal of Product Innovation
Management, 25, 608-619.
Grant, N., Heere, B., & Dickson, G. (2011). New sport teams and
the development of brand community. European Sport Management Quarterly,
11, 35-54.
Gwinner, K., & Swanson, S. (2003). A model of fan
identification: Antecedents and sponsorship outcomes. Journal of
Services Marketing, 17, 275-294.
Hair, J. F., Black, W., Babin, B., Anderson, R. E., & Tatham,
R. L. (2006). Multivariate data analysis (5th ed.). Upper Saddle River,
NJ: Prentice Hall.
Hofstede, G. (2001). Culture's consequences: Comparing values,
behaviors, institutions, and organizations across nations (2nd ed.).
Beverley Hills, CA: Sage Publications.
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit
indexes in covariance structure analysis: Conventional criteria versus
new alternatives. Structural Equation Modeling, 6, 1-55.
Hunt, K. A., Bristol, T., & Bashaw, R. E. (1999). A conceptual
approach to classifying sports fans. Journal of Services Marketing, 13,
439-452.
J. League (2013). J. League fan survey 2013 summary report. Tokyo,
Japan: Japan Professional Football League.
Jones, M. A., & Reynolds, K. E. (2006). The role of retailer
interest on shopping behavior. Journal of Retailing, 82, 115-126.
Jones, T., Taylor, S. F., & Bansal, H. S. (2008). Commitment to
a friend, a service provider, or a service company: Are they
distinctions worth making? Journal of the Academy of Marketing Science,
36, 473-487.
Katz, M., & Heere, B. (2013). Leaders and followers: An
exploration of the notion of scale-free networks within a new brand
community. Journal of Sport Management, 27, 271-287.
Keller, K. L. (2003). Strategic brand management: Building,
measuring and managing brand equity (2nd ed.). Englewood Cliffs, NJ:
Prentice-Hall.
Mael, F. A., & Ashforth, B. E. (2001). Identification in work,
war, sports, and religion: Contrasting the benefits and risks. Journal
for the Theory of Social Behaviour, 31, 197-222.
Mahony, D. F., Nakazawa, M., Funk, D. C., James, J. D., &
Gladden, J. M. (2002). Motivational factors influencing the behavior of
J. League spectators. Sport Management Review, 5, 1-24.
McAlexander, J. H., Schouten, J. W., & Koenig, H. F. (2002).
Building brand community. Journal of Marketing, 66, 38-54.
Melnick, M. J. (1993). Searching for sociability in the stands: A
theory of sports spectating. Journal of Sport Management, 7, 44-60.
Muniz, A. M., & O'Guinn, T.C. (2001). Brand community.
Journal of Consumer Research, 27, 412-432.
Nakazawa, M., Mahony, D., Funk, D., & Hirakawa, S. (1999).
Segmenting J. League spectators based on length of time as a fan. Sport
Marketing Quarterly, 8, 55-65.
Oliver, R. L. (1999). Whence consumer loyalty? Journal of
Marketing, 63(5), 33-44.
Palmatier, R. W., Scheer, L. K., & Steenkamp, J. E. M. (2007).
Customer loyalty to whom? Managing the benefits and risks of
salesperson-owned loyalty. Journal of Marketing Research, 44, 185-199.
Rindfleisch, A., Burroughs, J. & Wong, N. (2008). The safety of
objects: Materialism, existential insecurity, and brand connection.
Journal of Consumer Research, 36, 1-16.
Rosenbaum, M. S. (2008). Return on community for consumers and
service establishments. Journal of Service Research, 11, 179-196.
Rust, R. T., Zeithaml, V. A., & Lemon, K. N. (2000). Driving
customer equity: How customer lifetime value is reshaping corporate
strategy. New York, NY: Free Press.
Schau, H. J., Muniz, A. M., & Arnold, E. J. (2009) How brand
community practices create value. Journal of Marketing, 73(5), 30-51.
Schouten, J. W., & McAlexander, J. H. (1995). Subcultures of
consumption: An ethnography of the new bikers. Journal of Consumer
Research, 22, 43-61.
Seiders, K., Voss, G. B., Grewal, D., & Godfrey, A. L. (2005).
Do satisfied customers buy more? Examining moderating influences in a
retailing context. Journal of Marketing, 69(4), 26-43.
Steenkamp, J. B. E., & Baumgartner, H. (1998). Assessing
measurement invariance in cross-national consumer research. Journal of
Consumer Research, 25(1), 78-107.
Sutton, W. A., McDonald, M. A., Milne, G. R., & Cimperman, J.
(1997). Creating and fostering fan identification in professional sport.
Sport Marketing Quarterly, 6(1), 15-22.
Tajfel, H., & Turner, J. C. (1985). The social identity theory
of intergroup behavior. In S. Worchel & W. G. Austin (Eds.),
Psychology of intergroup relations (2nd ed., pp. 7-23). Chicago, IL:
Nelson-Hall.
Tian, K. T., Bearden, W. O., & Hunter, G. L. (2001).
Consumers' need for uniqueness: Scale development and validation.
Journal of Consumer Research, 28(2), 50-66.
Wann, D. L., & Branscombe, N. R. (1993). Sports fans: Measuring
degree of identification with their team. International Journal of Sport
Psychology, 24(1), 1-17.
Woolf, J., Heere, B., & Walker, M. (2013). Do charity sport
events function as "brandfests" in the development of brand
community? Journal of Sport Management, 27, 95-107.
Yoshida, M., & James, J.D. (2010). Customer satisfaction with
game and service experiences: Antecedents and consequences. Journal of
Sport Management, 24, 338-361.
Yoshida, M., & Gordon, B. (2012). Who is more influenced by
customer equity drivers? A moderator analysis in a professional soccer
context. Sport Management Review, 15, 389-403.
Yoshida, M., Gordon, B. S., 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. S. (in press). 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. Retrieved from
http://dx.doi.org/10.1123/jsm.2013-0306
Endnote
(1) The baseball sample was eliminated from the moderator analysis
because (1) the sample included unpaid loyalty program customers with no
membership fees and (2) the sample size of non-participants was too
small for factor analysis (n = 83).
Acknowledgments
This study was partially supported by a grant from the Japan
Professional Football League.
Table 1
CFA Results
Construct Item
Fan community identification ([CR.sub.soccer] = .86,
[AVE.sub.soccer] = .61; [CR.sub.baseball] = .89,
[AVE.sub.baseball] = .67)
FCI1. I really identify with people who follow
(team name).
FCI2. I feel like I belong to a club with other
fans of (team name).
FCI3. (team name) is supported by people like me.
FCI4. I feel a deep connection with others who
follow (team name).
Team brand equity ([CR.sub.soccer] .79,
[AVE.sub.soccer] .50; [CR.sub.baseball]
.81, [AVE.sub.baseball] .52)
TBE1. How loyal are you to (team name)? (not at all
loyal [1] to very loyal [7])
TBE2. What kind of attitude do you have about
(team name)? (negative attitude [1] to
positive attitude [7])
TBE3. What kind of image do you have about (team
name)? (negative image [1] to positive
image [7])
TBE4. How would you rate the event quality delivered
by your team? (low quality [1] to
high quality [7])
Fan community engagement ([CR.sub.soccer] = .81,
[AVE.sub.soccer] = .52; [CR.sub.baseball] = .89,
[AVE.sub.baseball] = .67)
FCE1. I often buy memorabilia to represent
memorable games.
FCE2. I often buy apparel which represents the fans
of (team name).
FCE3. In order to share a sense of belonging with
(team name)'s fan group, I often wear clothing
that displays the logo of (team name).
FCE4. I often talk to others or blog about my unique
experiences shared with other fans of (team name).
Customized product use ([CR.sub.soccer] = .86,
[AVE.sub.soccer] = .67; [CR.sub.baseball] =
.92, [AVE.sub.baseball] = .79)
CPU1. I love to show my customized products to other
fans of (team name).
CPU2. I often design spectator products in order to
fit the unique concept of (team name)
CPU3. The extensive use of my customized items enables
me to guide other fans of (team name).
Member responsibility ([CR.sub.soccer] .84,
[AVE.sub.soccer] .64; [CR.sub.baseball] .94,
[AVE.sub.baseball] .83)
MR1. I have a sense of duty to attract new fans
of (team name).
MR2. I have a sense of obligation for keeping the
current fans of (team name).
MR3. I am obligated to provide team-related
information to other fans.
Positive word of mouth ([CR.sub.soccer]
.84, [AVE.sub.soccer] .64; [CR.sub.baseball]
.91, [AVE.sub.baseball] .78)
WOM1. When I talk with my friends, I give them a
good impression of (team name).
WOM2. I often say positive things to a friend about
enthusiastic fans of (team name).
WOM3. When I talk with my friends, I give them a
good impression of the fan community of
(team name).
Fit Indices [chi square]
df
[chi square] / df
CFI
NNFI
RMSEA
Factor loading ([lambda])
Construct Soccer Baseball
(n = 401) (n = 326)
Fan community identification ([CR.sub.soccer] = .86,
[AVE.sub.soccer] = .61; [CR.sub.baseball] = .89,
[AVE.sub.baseball] = .67)
FCI1. .80 .83
FCI2. .82 .88
FCI3. .75 .73
FCI4. .75 .80
Team brand equity ([CR.sub.soccer] .79,
[AVE.sub.soccer] .50; [CR.sub.baseball]
.81, [AVE.sub.baseball] .52)
TBE1. .53 .60
TBE2. .84 .82
TBE3. .85 .83
TBE4. .55 .61
Fan community engagement ([CR.sub.soccer] = .81,
[AVE.sub.soccer] = .52; [CR.sub.baseball] = .89,
[AVE.sub.baseball] = .67)
FCE1. .74 .86
FCE2. .85 .91
FCE3. .70 .79
FCE4. .58 .70
Customized product use ([CR.sub.soccer] = .86,
[AVE.sub.soccer] = .67; [CR.sub.baseball] =
.92, [AVE.sub.baseball] = .79)
CPU1. .76 .86
CPU2. .88 .93
CPU3. .82 .86
Member responsibility ([CR.sub.soccer] .84,
[AVE.sub.soccer] .64; [CR.sub.baseball] .94,
[AVE.sub.baseball] .83)
MR1. .82 .91
MR2. .85 .96
MR3. .72 .86
Positive word of mouth ([CR.sub.soccer]
.84, [AVE.sub.soccer] .64; [CR.sub.baseball]
.91, [AVE.sub.baseball] .78)
WOM1. .83 .85
WOM2. .77 .93
WOM3. .81 .86
Fit Indices
[chi square] 484.31 477.79
df 174 174
[chi square] / df 2.78 2.75
CFI .98 .98
NNFI .97 .97
RMSEA .067 .073
Note. CR = composite reliability; AVE = average variance extracted
Table 2
Descriptive Statistics [phi] Matrix, and AVE Values (a)
[phi] matrix (b) ([n.sub.
soccer] = 401)
Construct 1 2 3
1. Fan community identification .61 .26 .40
2. Team brand equity .51 ** .50 .18
3. Fan community engagement .64 ** .43 ** .52
4. Customized product use .48 ** .36 ** .59 **
5. Member responsibility .64 ** .43 ** .78 **
6. Positive word-of-mouth .76 ** .52 ** .80 **
7. Attendance frequency .13 * .11 * .24 **
8. Length of time as a fan .10 .07 .00
[Mean.sub.soccer] (c) 4.64 4.99 3.78
Standard [deviation.sub. 1.28 1.17 1.52
soccer] (c)
[phi] matrix (b) ([n.sub.
baseball] = 326)
Construct 1 2 3
1. Fan community identification .67 .44 .41
2. Team brand equity .67 ** .52 .30
3. Fan community engagement .64 ** .55 ** .67
4. Customized product use .34 ** .30 ** .52 **
5. Member responsibility .62 ** .43 ** .60 **
6. Positive word-of-mouth .73 ** .68 ** .68 **
7. Attendance frequency .19 ** .03 .21 **
8. Length of time as a fan .16 ** .09 .15 **
Mean.sub.baseball] (c) 5.38 5.91 4.38
Standard deviatio[n.sub. 1.33 .98 1.67
baseball] (c)
[phi] matrix (b)
([n.sub.soccer] = 401)
Construct 4 5 6
1. Fan community identification .23 .41 .58
2. Team brand equity .13 .19 .27
3. Fan community engagement .34 .61 .64
4. Customized product use .67 .39 .20
5. Member responsibility .63 ** .64 .65
6. Positive word-of-mouth .45 ** .80 ** .64
7. Attendance frequency -.01 .08 .23 **
8. Length of time as a fan .00 .00 .07
[Mean.sub.soccer] (c) 2.76 3.70 4.44
Standard [deviation.sub. 1.48 1.54 1.51
soccer] (c)
[phi] matrix (b)
([n.sub.baseball] = 326)
Construct 4 5 6
1. Fan community identification .12 .38 .53
2. Team brand equity .09 .18 .47
3. Fan community engagement .27 .36 .46
4. Customized product use .79 .32 .16
5. Member responsibility .56 ** .83 .42
6. Positive word-of-mouth .40 ** .65 ** .78
7. Attendance frequency .15 ** .11 .17 **
8. Length of time as a fan .02 .10 .15 *
Mean.sub.baseball] (c) 2.99 3.86 5.03
Standard deviatio[n.sub. 1.76 1.76 1.61
baseball] (c)
[phi] matrix (b)
([n.sub.soccer]
= 401)
Construct 7 8
1. Fan community identification .02 .01
2. Team brand equity .01 .00
3. Fan community engagement .06 .00
4. Customized product use .00 .00
5. Member responsibility .01 .00
6. Positive word-of-mouth .05 .00
7. Attendance frequency N.A. .10
8. Length of time as a fan .32 ** N.A.
[Mean.sub.soccer] (c) 8.70 6.81
Standard [deviation.sub. 4.96 4.44
soccer] (c)
[phi] matrix
(b) ([n.sub.
baseball] = 326)
Construct 7 8
1. Fan community identification .04 .03
2. Team brand equity .00 .01
3. Fan community engagement .04 .02
4. Customized product use .02 .00
5. Member responsibility .01 .01
6. Positive word-of-mouth .03 .02
7. Attendance frequency N.A. .06
8. Length of time as a fan .25 ** N.A.
Mean.sub.baseball] (c) 11.77 10.05
Standard deviatio[n.sub. 11.41 8.23
baseball] (c)
(a) The AVE value for each construct is shown in boldface
italic on the diagonal.
(b) Correlations are taken from f matrix using LISREL 8.8 and are
reported in the lower triangle of the f matrix; Squared correlations
are depicted in the upper triangle of the f matrix.
(c) The mean scores and standard deviations for the eight
constructs are calculated using IBM SPSS statistics 20.0.
* p < .05; ** p < .01; N.A. = not applicable.
Table 3
Standardized Parameter Estimates (t-value) and Hypothesis Testing
Path Hypothesis Soccer (n =
401) Path
Fan community identification coefficient
[right arrow] Team brand equity [H.sub.1] .53 **(6.97)
[right arrow] Fan community [H.sub.2a] .75 **(11.18)
engagement
[right arrow] Customized [H.sub.2b] .55 **(8.37)
product use
[right arrow] Member [H.sub.2c] .79 **(12.27)
responsibility
[right arrow] Positive word-of- [H.sub.2d] .81 **(13.08)
mouth
Team brand equity
[right arrow] Fan community [H.sub.3a] .02(.28)
engagement
[right arrow] Customized [H.sub.3b] .07(1.17)
product use
[right arrow] Member [H.sub.3c] .02(.42)
responsibility
[right arrow] Positive word-of- [H.sub.3d] .08(1.50)
mouth
Attendance frequency
[right arrow] Team brand equity Control .04(.72)
[right arrow] Fan community Control .18 **(4.09)
engagement
[right arrow] Customized Control -.08(-1.52)
product use
[right arrow] Member Control .000(.001)
responsibility
[right arrow] Positive word-of- Control .14 **(3.17)
mouth
Length of time as a fan
[right arrow] Team brand equity Control -.01(-.09)
[right arrow] Fan community Control -.14 **(-3.29)
engagement
[right arrow] Customized Control -.04(-.81)
product use
[right arrow] Member Control -.08(-1.88)
responsibility
[right arrow] Positive word-of- Control -.07(-1.76)
mouth
[R.sup.2] .29
Team brand equity
Fan community engagement .62
Customized product use .35
Member responsibility .63
Positive word-of-mouth .77
Fit indices
[chi square] (df) 863.51(210)
[chi square] (df) 4.11
CFI .95
NNFI .95
RMSEA .088
Path Baseball (n = 326)
Path coefficient
Fan community identification
[right arrow] Team brand equity .69 **(8.45)
[right arrow] Fan community .53 **(7.05)
engagement
[right arrow] Customized .34 **(3.92)
product use
[right arrow] Member .67 **(8.39)
responsibility
[right arrow] Positive word-of- .52 **(7.64)
mouth
Team brand equity
[right arrow] Fan community .22 **(2.93)
engagement
[right arrow] Customized .10(1.14)
product use
[right arrow] Member .02(.21)
responsibility
[right arrow] Positive word-of- .35 **(4.85)
mouth
Attendance frequency
[right arrow] Team brand equity .10 *(-1.96)
[right arrow] Fan community .09(1.90)
engagement
[right arrow] Customized .09(1.67)
product use
[right arrow] Member -.02(-.43)
responsibility
[right arrow] Positive word-of- .05(1.21)
mouth
Length of time as a fan
[right arrow] Team brand equity .01(.20)
[right arrow] Fan community .02(.51)
engagement
[right arrow] Customized -.07(-1.18)
product use
[right arrow] Member .01(-.14)
responsibility
[right arrow] Positive word-of- .02(.40)
mouth
[R.sup.2] .46
Team brand equity
Fan community engagement .51
Customized product use .19
Member responsibility .46
Positive word-of-mouth .66
Fit indices
[chi square] (df) 831.71(210)
[chi square] (df) 3.96
CFI .96
NNFI .95
RMSEA .095
* p < .05; ** p < .01
Table 4
Moderating Effects of Fan Loyalty Program Participation:
Results of the Soccer Sample
[gamma]
Estimates of
Free Model
(t-Value)
Moderating Effect of Fan Loyalty Participants
Program Participation on (n = 183)
Team brand equity
[right arrow] Fan community [H.sub.4a] .09(1.32)
engagement
Team brand equity
[right arrow] Customized [H.sub.4b] .12(1.66)
product use
Team brand equity
[right arrow] Member [H.sub.4c] .16 *(2.23)
responsibility
Team brand equity
[right arrow] Positive [H.sub.4d] .26 **(3.44)
word-of-mouth
[gamma]
Estimates of
Free Model
(t-Value)
Moderating Effect of Fan Loyalty Non-participants [DELTA][chi
Program Participation on (n= 191) square]
([DELTA]df)
Team brand equity
[right arrow] Fan community .11(1.43) .18(1)
engagement
Team brand equity
[right arrow] Customized .15(1.86) .22(1)
product use
Team brand equity
[right arrow] Member .02(.22) 1.77(1)
responsibility
Team brand equity
[right arrow] Positive .04(.59) 3.88 *(1)
word-of-mouth
Note. * p < .05;** p < .01; The critical values for [DELTA]
[chi square] with df = 1 are 3.84 at the .05 level and 6.64
at the .01 level.
