Binder 2
Binder 2
DOI 10.1123/jsm.2019-0222
ISSN 0888-4773
ESSN 1543-270X
Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A
copy can be downloaded for personal non-commercial research or study, without prior permission
or charge. This item cannot be reproduced or quoted extensively from without first obtaining
permission in writing from the copyright holder(s). The content must not be changed in any way or
sold commercially in any format or medium without the formal permission of the copyright
holders.
This document is the author’s post-print version, incorporating any revisions agreed during the
peer-review process. Some differences between the published version and this version may
remain and you are advised to consult the published version if you wish to cite from it.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 1
34
ATHLETE BRANDS AND SELF-BRAND CONNECTION 2
35 The role of self-brand connection on the relationship between athlete brand image and
36 fan outcomes
37 Abstract
38 This research explores the role of athlete on-field and off-field brand image on consumer
39 commitment toward the athlete and associated team, preference by athlete’s sponsor, and the
41 collected from fans of soccer players through a cross-sectional survey promoted on social
42 media platforms. A partial least squares structural equation model examined the direct effects
43 of both athlete brand dimensions on athlete commitment, team commitment and athlete
44 sponsor preference, and the indirect effects mediated via self-brand connection. The results
45 indicate that athlete on-field image is significantly related to athlete sponsor preference, while
46 the off-field image influences athlete commitment and team commitment. Self-brand
47 connection is influenced by athlete off-field image, and mediates the relationship between
48 off-field image and athlete commitment. This study contributes to a better understanding of
49 how to manage athlete brands and linkages between fans, athletes and associated entities.
50
53 Introduction
54 Athletes enjoy increasing levels of social influence extending well beyond the sports
55 in which they compete (Arai, Ko, & Kaplanidou, 2013; Parmentier, 2011). Today’s athletes
56 often inspire their supporters (Kerr & Gladden, 2008) and many serve as multiplatform
57 promotional entities that draw attention from media and corporate sponsors (Arai et al., 2013;
58 Summers & Johnson Morgan, 2008). As illustrated by Forbes’ (2017) inclusion of athletes in
59 their yearly “Most Valuable Sports Brands” list, athletes themselves have become brands
60 (Chadwick & Burton, 2008). This trend has been exacerbated by the global expansion of
61 social media platforms, which has shifted away from team-following toward athlete-
62 following (Geurin-Eagleman & Burch, 2016). For example, in May 2019, Cristiano Ronaldo
63 (165 million) had more than double the number of Instagram followers than his former club
64 Real Madrid (71 million) and eight times the amount of his current club Juventus (26
65 million); and has signed a one-billion dollar lifetime endorsement deal with Nike in 2016
66 because of his image and reach (Badenhausen, 2016). Similarly, Ben Simmons (4.2 million)
67 had more than double the number of followers on Instagram than his team, the Philadelphia
68 76ers (1.8 million). These athletes often act as entrepreneurs of their own brands (Ratten,
69 2015), and require strategic growth management guidance to build their audience
71 Athlete branding has become a topic of academic inquiry, not just a trend in
72 marketing. Recent studies have focused on the conceptualization of athletes’ brand image
73 (e.g., Arai, Ko, & Ross, 2014; Parmentier & Fischer, 2012; Pegoraro, 2010). Arai et al.
74 (2014) proposed a model based on athletic performance (e.g., skills, performance), attractive
75 appearance (e.g., physical attractiveness, personal style), and marketable life style (e.g., off-
76 field life story, role model). Parmentier and Fisher (2012) suggested that athlete brand image
77 depends on professional image (e.g., playing opportunities) and his/her media persona (e.g.,
79 posts into two types: front stage (e.g., on-field performance) and backstage (e.g., personal
80 life). Although there is no consensus on the most relevant athlete brand associations, there
81 seems to be agreement among researchers that on-field image and off-field image are the two
82 key dimensions to properly understand athlete brands (e.g., Eagleman-Geurin & Burch,
83 2016). Following this reasoning, Hasaan, Kerem, Biscaia, and Agyemang (2018) have
84 proposed a framework based on on-field and off-field image, although the model has not
86 Previous studies have suggested that brand image influences attitudinal and
87 behavioral outcomes linked with the athlete and related entities, such as the athlete’s team
88 and sponsors (Carlson & Donavan, 2013; Yu, 2005; Williams, Kim, Agyemang, Martin,
89 2015). Research on sport brand architecture indicates that brands are evaluated in context,
90 rather than in isolation, and that image spill-over happens between brands in a portfolio (e.g.,
91 Cobbs, Groza, & Rich, 2015). Despite the common understanding of athletes as brands,
92 studies utilizing branding theories to expound the phenomena still require further
93 development to better understand fan reactions to athletes and related entities (Arai et al.,
94 2013; Hasaan et al., 2018). Star players are regularly cited as brand associations of teams
95 (e.g., Daniels, Kunkel, & Karg, 2019) and superstar athletes have the potential to impact the
96 gate revenue of their team (e.g., Shapiro, DeShriver, & Rasher, 2017). Similarly, high profile
97 athletes are regularly used as brand endorsers (Chanavat, Desbordes, & Dickson, 2016; Fink,
98 Parker, Cunningham, & Cuneen, 2012) and some have become entrepreneurs through
99 leveraging the value of their personal brand and social reach (Ratten, 2015). However, the
100 unique influence of athletes’ on-field and off-field brand image on fan’s attitude toward the
101 athletes, their teams and sponsors requires empirical testing (Summers & Johnson Morgan,
102 2008).
103 The recognition of athletes’ positive on-field and off-field image may not always
104 translate into favorable attitudes toward the athlete and related entities. For example, statistics
ATHLETE BRANDS AND SELF-BRAND CONNECTION 5
105 indicate that Cristiano Ronaldo and Lionel Messi have similar outstanding achievements
106 differentiating them from other soccer athletes, yet people have strong opinions (positive &
107 negative) toward them (The Guardian, 2017). This difference may be related with the on-field
108 and off-field brand image of the athlete, as well as fans’ self-brand connection, given that
109 extant branding research shows that consumers support brands that are congruent with their
110 self-brand image (Escalas & Bettman, 2003) and a strong self-brand connection often leads
111 consumers to develop and maintain a committed relationship with the brand (Fournier, 1998).
112 In a celebrity endorsement context, self-brand connection has been suggested to partly
113 mediate the relationship between the perceived credibility of an endorser and the subsequent
114 endorsed brand equity (Dwivedi, Johnson, & McDonald, 2015). Consequently, self-brand
115 connection may play a role on the relationship between athlete brand image and consumers’
116 commitment toward the athlete; (2) the athlete’s team; and (3) the athlete’s sponsor.
117 The purpose of this research was threefold. First, we examined the relationship
118 between athlete’s on-field and off-field brand image and consumers’ commitment to the
119 athlete. Second, we tested the impact of athlete brand image on consumer outcomes toward
120 related entities, such as his/her commitment to the team and sponsor preference. Third, we
121 examined whether these relationships were mediated by consumers’ self-brand connection to
122 the athletes. Data were collected from social media users who supported professional soccer
123 athletes. This research contributes to both sport marketing research and practice by
124 examining the specific brand associations that form the on-field and off-field brand image of
125 athletes, their impact on consumers’ reactions toward the athlete and related entities, and the
126 mediating role of self-brand connection to the athlete. The research findings provide insights
127 for agents and athletes building their own brands as entrepreneurs, as well as teams and
133 (Parmentier, 2011, p. 219). This means that the value of a brand lies in all descriptive and
134 evaluative information held in consumers’ memory. Information processing theories form the
135 basis for branding research (Allen, Fournier, & Miller, 2008). For example, Keller’s (1993)
136 seminal work of customer-based brand equity model was based on the associative network
137 memory model (Anderson, 1983), premising consumers’ various decision making largely
138 depends on their knowledge of the brand and the thoughts they link with the brand. When
139 applied to athletes, this brand knowledge consists of awareness of the athlete and his/her
140 brand image (e.g., Arai et al., 2014). As such, an athlete’s brand image refers to the set of
141 associations that people identify with a particular athlete (e.g., Parmentier, Fischer, & Reuber,
142 2013), and its development requires fans’ awareness of the athlete and an understanding of
143 how the athlete is different from other athletes (Montoya, 2002; Hasaan et al., 2018). While
144 athlete brand awareness is often generated through team/sport prominence and socializing
145 agents such as media and peer group influence (Hasaan et al., 2018), brand image generation
147 Athlete brand image is multi-faceted. While Keller’s original customer-based brand
148 equity model (1993) categorized brand associations into attributes (consumer associations of
149 the product’s descriptive characteristics), benefits (personal value consumer attach to the
150 product) and attitudes (overall consumer evaluation of the brand), brand attributes have been
151 suggested to represent controllable aspects of the athlete brand and important predictors of
152 consumers’ subsequent reactions to the athletes and associated brands (Arai et al., 2014).
153 Consequently, Arai et al. (2013; 2014) proposed and empirically tested a model of athlete
154 brand image (MABI) consisting of 10 associations across three key dimensions: athletic
ATHLETE BRANDS AND SELF-BRAND CONNECTION 7
155 performance, attractive appearance and marketable lifestyle. These studies were instrumental
156 in understanding athletes as brands and highlighted how on-field performance (e.g., athletic
157 performance) does not capture every facet of an athlete’s brand, and that there are other
158 associations not related to on-field performance (e.g., marketable lifestyle or social
159 attractiveness) that contribute to creating an athlete brand in fans’ minds. Yu (2005) further
160 argued that a fan’s connection to a certain athlete is influenced by his/her successful career
161 and personal life’s appeal, and Arai et al. (2013) highlighted that athletes achieve their status
162 as brands through their outstanding performance on the field but also via their distinctive
163 lifestyle. For example, athletes such as David Beckham or Cristiano Ronaldo are strong
164 brands in the marketplace that have heavily relied on both on-field and off-field brand
166 It follows then that both on-field and off-field image should be considered when
167 examining an athlete’s brand (Parmentier & Fischer, 2012). Following Braunstein and Zhang
168 (2005) and Arai et al. (2013; 2014), Hasaan et al. (2018) conducted a literature review and
169 proposed a conceptual framework of athlete brand based on-field and off-field image. Given
170 that athletes primarily develop their brand status based on continued excellence and success
171 in their sport, on-field image attributes represent a crucial component of athlete brand image
172 (Arai et al., 2013). The on-field image attributes refer to performance-related characteristics
173 of an athlete (Arai et al., 2014) and in the current research include the dimensions of Fair
174 Play, Effort, Achievements, Style of Play, Impact, and Skills (Arai et al., 2013; Chadwick &
175 Burton, 2008; Hasaan et al., 2018). Additionally, the public persona of an athlete is important
176 to brand status (Walsh & Williams, 2017), given how off-field activities tend to contribute to
177 an athlete’s broad public perception (Arai et al., 2014; Summers & Johnson Morgan, 2008).
178 Parmentier and Fischer (2012) further posited that, unlike products and services that
179 are produced purely to serve markets, person brands have purposes beyond bringing profit to
180 themselves or the organization they work in. Similarly, Thomson (2006) refers that
ATHLETE BRANDS AND SELF-BRAND CONNECTION 8
181 connections with human brands imply a bona fine human being (e.g., David Beckham) and
182 not an inanimate object (e.g., book). To this end, the off-field image of an athlete is related to
183 his/her life beyond the sport activity, and are conceptualized in the current research through
184 the dimensions of Physical Attraction, Body Conditioning, Lifestyle, Personality, Cultural
185 Background, Social Responsibility, and Role Model (e.g., Arai et al., 2014; Hasaan et al,
186 2018; Parmentier & Fischer, 2012). The definitions of the on-field and off-field image used in
187 this study, theoretical support, and examples for each association are provided in Table 1.
190 Researchers have proposed that both the on-field and off-field image of athletes are
191 important parts of establishing and sustaining a connection with fans (e.g., Arai et al., 2013;
192 Parmentier & Fischer, 2012). However, there is a lack of empirical evidence demonstrating
193 the impact of athlete associations on fans’ loyalty toward the athlete. Examining sport brand
194 elements that drive fans' reactions is paramount to increase brand health (Biscaia et al., 2016),
195 which suggests the importance of understanding how consumer perceptions of an athlete’s
196 on-field and off-field image influence their commitment to support that athlete. Building on
197 the fact consumer commitment is a vital component of loyalty toward brands (Arai et al.,
198 2014; Oliver, 1999; Tsiotsou, 2013; Yoo, Donthu, & Lee, 2000), the current study measures
199 consumers’ psychological commitment to an athlete (i.e. athlete commitment) and examines
200 how it is impacted by both the on-field and off-field image of that athlete.
201 Thomson (2006) posits human brands (e.g., athletes) are often perceived as sources of
202 attractiveness (familiarity, likeability, and/or similarity) and credibility (expertise and
203 trustworthiness), and evidence from daily life indicates that attractiveness and lifestyle are
204 often the subject of great fan interest (The Guardian, 2016). Source credibility and
205 attractiveness models (McGuire, 1985) can be used to explain the impact of athlete image on
206 consumer perceptions of the athlete. That is, athletes who are well known and liked by
ATHLETE BRANDS AND SELF-BRAND CONNECTION 9
207 individuals are likely to be the object of strong interest and followership (Geurin-Eagleman &
208 Burch, 2016; Väätäinen & Dickenson, 2018). Thomson (2006)’s study indicates consumers
209 can develop committed relationship with human brands (e.g., athletes) that are similar to real
210 interpersonal relationship. By repeatedly being exposed to the image of a human brand,
211 consumers feel emotional security and fulfilment of fundamental needs (i.e., autonomy,
212 relatedness, and competence) provided by the human brand, and subsequently develop
213 satisfied, trusting, and committed relationships. This process provides initial evidence how
214 athlete image can affect consumers’ commitment toward that athlete.
215 Additionally, findings from brand research demonstrate that team brand associations
216 affect consumer reactions (e.g., Biscaia et al., 2016; Gladden & Funk, 2001; Kunkel, Doyle,
217 Funk, Du, & McDonald, 2016; Moore & Homer, 2008; Watkins, 2014), and the Fan Attitude
218 Network (FAN) model (Funk & James, 2004) posits that consumers develop a positive
219 attitude toward a sport brand when they perceive that the brand has attractive attributes (e.g.,
220 Funk, Beaton & Alexandris, 2012). For instance, Lunardo, Gergaud, and Livat (2015) found
221 that personality dimensions (i.e., sophistication, sincerity, competence, excitement, and
222 appeal) have a positive impact on a celebrity’s appeal. Moreover, consumer commitment with
223 athletes and teams throughout time is an important pillar for the sport industry to succeed
224 (e.g., Wang, Zhang, & Tsuji, 2011; Wu, Tsai, & Hung, 2012; Gladden & Funk, 2001) and,
225 regardless of the focal entity, brand associations are often suggested to be paramount to
226 understand the level of commitment toward a brand (Arai et al., 2014; Kunkel et al., 2016;
227 Yoo et al., 2000). As such, Summers and Johnson Morgan (2008) indicate consumers expect
228 exemplary behavior both on- and off-field from athletes. However, sports fans can generally
229 separate the on- and off-field behaviors of athletes (Lee & Kwak, 2016; Summer & Johnson
230 Morgan, 2008) and are willing to manage their expectations accordingly. Consequently, the
231 following hypotheses were developed to examine how an athlete’s on-field and off-field
233 Hypothesis 1a: Consumers’ perceptions of an athlete’s on-field image are positively
235 Hypothesis 1b: Consumers’ perceptions of an athlete’s off-field image are positively
238 indicates that entities at all levels in the sport industry are connected (Cobbs et al., 2015;
239 Williams et al., 2015). For example, in the context of professional spectator sport, a league
240 often represents the master brand that provides the framework for teams to compete, while
241 teams represent sub-brands that provide the core product within the portfolio of the league
242 (Kunkel, Funk, & King, 2014). Similarly, it has been proposed that athletes are integrated
243 into the brand portfolio of their teams (Williams et al., 2015). Leagues and teams both
244 influence consumer involvement (Stevens & Rosenberger, 2010), because they are in a
245 mixed-branding brand architecture where they are visibly connected, and consumer
246 perceptions of the team influence consumer perceptions of the league (Kunkel, Funk, &
247 Lock, 2017). Studies have demonstrated that sub-brands can affect the evaluation of the
248 master brand by either diluting or enhancing its corporate brand image (Balachander &
249 Ghose, 2003). Consequently, it is important to investigate the impact of athletes on their
251 Researchers have highlighted the impact of athletes on their teams (e.g., Brandes,
252 Franck, & Nüesch, 2008), leagues (e.g., Shapiro et al., 2017), and sponsors (e.g., Fink et al.,
253 2012). Drawing on the meaning transfer model (McCracken, 1986), which posits that
254 meaning is transferred in the mind of consumers between related brands, we extend Kunkel et
255 al.’s (2014) conceptualization of sport brand architecture to athletes. That is, athletes can be
256 considered as sub-brands that are visibly connected to their respective teams, which represent
257 the master brand from a structural perspective (i.e., the team governs the athlete). For
258 example, the transfer of David Beckham to Major League Soccer (MLS) contributed to the
ATHLETE BRANDS AND SELF-BRAND CONNECTION 11
259 brand image and revenue of the MLS as a whole and his new team, LA Galaxy (Shapiro et
260 al., 2017). Following this rationale, understanding how perceptions of an athlete’s on- and
261 off-field brand image affects consumers’ psychological commitment to the associated team
262 (i.e., team commitment) represents an important step when managing the brands within the
264 Hypothesis 2a: Consumers’ perceptions of an athlete’s on-field image are positively
266 Hypothesis 2b: Consumers’ perceptions of an athlete’s off-field image are positively
268 In a similar way, research has also indicated that athletes influence consumers’
269 reactions of their sponsors. Like with athletes, teams and leagues, consumers transfer the
270 meanings associated with an endorser to a brand when an associative link is established
271 between them (Amos, Holmes, & Strutton, 2008; Fink et al., 2012). This relationship has also
272 been demonstrated for leagues and their sponsors (e.g., Farrelly & Quester, 2005), teams and
273 their sponsors (e.g., Biscaia, Correia, Ross, Rosado, & Marôco, 2013), and athletes and their
274 sponsors (e.g., Yu, 2005). Given the mixed-branding architecture of sport brands, athlete
275 brand image has been proposed to transfer to their sponsors (e.g., Chanavat et al., 2016,
276 Chanavat, Martinent, & Ferrand, 2009). The social attractiveness and professional
277 trustworthiness of the athletes often make them marketable for sponsorship deals (Ratten,
278 2015; Summer & Johnson Morgan, 2008), and the ultimate goal of sponsors is to orient
279 consumer preferences toward their products (Barros & Silvestre, 2006), as sponsors expect
280 return on the investments on the athletes (Yu, 2005). Thus, examining how athlete’s on-field
281 and off-field image influences athlete sponsor preference (i.e., consumers’ positive
282 evaluations of athlete sponsors and willingness to purchase associated products) is critical to
285 Hypothesis 3a: Consumers’ perceptions of an athlete’s on-field image are positively
287 Hypothesis 3b: Consumers’ perceptions of an athlete’s off-field image are positively
289 Fans create connections with brands (Escalas & Bettman, 2017; Geurin-Eagleman &
290 Burch, 2016). For example, action sports athletes (e.g., snowboarders) frame their athletic life
291 under an extreme environment allowing them to actively communicate a symbolic meaning
292 that fosters a sense of identification and attract loyal audiences. Consistent with this view,
293 self-brand connections are formed when consumers engage in a matching process to identify
294 brands that are congruent with their self-images (Chaplin & John, 2005). The current study
295 theorizes that the meaning linked to athlete brands is vital for consumers to develop a self-
297 Self-brand connection is defined as the degree to which a brand delivers on important
298 identity concerns, tasks or themes, thereby expressing a significant aspect of self, including to
299 past (nostalgic), current and future (possible or desired) selves (Fournier, 1998). Fournier
300 (1998) argued that self-brand connection and commitment are two related but different
301 aspects contributing to strengthen the ties between consumers and brands. While self-brand
302 connection refers to fans’ perception of how closely the athlete represents themselves (e.g.,
303 Escalas, 2004), fans’ commitment toward the athlete is a subsequent response indicating their
304 promise to the relationship with the athlete, as described for teams in the FAN model (Funk
305 & James, 2004). The self-brand connection framework contends that people use brands to
306 create and represent their (desired) self-concept and to present their self-concept internally
307 and externally (Chaplin & John, 2005; Escalas, 2004). Escalas and Bettman (2015) further
308 argued that the set of brand associations is more meaningful for consumers when it is closely
309 linked to their self-concept because it can be used to construct their self-image. Furthermore,
310 according to the self-expansion theory (Aron, Aron, & Smollan, 1992), people possess an
ATHLETE BRANDS AND SELF-BRAND CONNECTION 13
311 inherent motivation to incorporate others (i.e. brands) into their self-concept, and consumers
312 for whom self-brand connection is high are likely to commit to a relationship with a brand
313 leading to greater loyalty (Parks, MacInnis, Eisingerich, & Iacobucci, 2010). For example, a
314 fan who feels strongly self-connected to Stephen Curry is likely to become loyal to him.
315 Thus, consumers matching their self-image with the athlete is a key aspect preceding a
317 Moreover, previous studies considered sports celebrities as product endorsers and
318 examined how consumers form self-brand connections through the symbolic brand meaning
319 derived from the celebrity endorser (Dwivedi et al., 2015; Escalas & Bettman, 2015, 2017).
320 Based on the meaning transfer model (McCracken, 1986), symbolic properties are first
321 associated with the brands the celebrity endorses and the symbolic meanings are transferred from
322 the celebrity to consumers (Escalas & Bettman, 2009). When the symbolic meaning associated
323 with the celebrities is used to communicate the self-concept to others, the meaning is internalized
324 to consumers and a self-brand connection is formed. By applying the rationale to the athlete
325 brand context, the current study argues that on-field and off-field image of an athlete may
326 convey a symbolic meaning to consumers, who will likely form a self-brand connection with
327 that athlete if they identify with those symbolic properties. Dwivedi et al. (2015) further
328 investigated the impact of celebrity endorsers’ image on endorsed brand outcomes and found that
329 self-brand connection partly mediates the relationship between the endorser and associated
330 products. In addition, self-brand connection has been suggested to increase brand relationship
331 durability regardless of the circumstances (Fournier, 1998; Swaminathan, Page, & Gürhan-Canli,
332 2007). These studies provide support to the idea that celebrities are a direct source of symbolic
333 meaning for consumers and that self-brand connections influence reactions toward the athlete
334 brand and endorsed brands. Thus, the following hypotheses were developed, and all
336 Hypothesis 4a: Consumers’ perceptions of an athlete’s on-field image are positively
ATHLETE BRANDS AND SELF-BRAND CONNECTION 14
338 Hypothesis 4b: Consumers’ perceptions of an athlete’s off-field image are positively
342 Hypothesis 6: Self-brand connection partially mediates the relationship between on-field
343 image and a) athlete commitment; b) team commitment; c) athlete sponsor preference.
344 Hypothesis 7: Self-brand connection partially mediates the relationship between off-field
345 image and a) athlete commitment; b) team commitment; c) athlete sponsor preference.
347 Method
348 A cross-sectional survey design was used to examine the relationships between athlete
349 brand image (conceptualized based on both on-field and off-field associations), self-brand
350 connection, athlete commitment, team commitment and athlete sponsor preference. We
351 choose one sport to eliminate sport-specific differences, focusing on soccer because of its
352 global prominence. Because consumers’ perceptions of athlete brands are socially
353 constructed and dependent on various aspects, such as culture and context (e.g., Aaker,
354 Benet-Martinez, & Garolera, 2001), we focused on measuring those brand associations that
357 Quantitative data were collected to test the proposed hypotheses. Participants were
358 recruited via Twitter and Facebook through a video created by the authors titled “Who is your
359 favorite soccer player?” The video description included hashtags relevant to the topic, such as
360 #football, #soccer, #FIFA, #MLS and #EPL, and contained a link to a survey hosted by
361 Qualtrics. The video was promoted to individuals who ‘liked’ soccer-specific Facebook
362 pages, were in soccer-specific groups, or followed high-profile soccer athletes. Respondents
ATHLETE BRANDS AND SELF-BRAND CONNECTION 15
363 had a chance to win prizes, such as $100 vouchers for Amazon. A total of $200 was spent on
365 A total of 455 respondents started the questionnaire during a period of two weeks.
366 After data cleaning, 279 responses were eliminated because they either completed the
367 questionnaire in an unrealistically short time, marked the same answer for every question,
368 provided inconsistent combinations, failed the attention check (i.e., “Click strongly disagree
369 to demonstrate that you are paying attention to the questions.”), had duplicate IP addresses, or
370 did not answer the sponsorship question correctly. Subsequently, a total of 177 respondents,
371 representing a useable response rate of 38.9%, were included in the data analysis. Despite the
372 response rate and the use of a convenience sample, in a review of survey research best
373 practices, Hulland, Baumgartner and Smith (2018) highlighted that low response rate do not
374 necessarily represent a problem for theory testing, and that the use of convenience samples
375 suffice when the aim is to test the veracity of proposed relationships. It is also important to
376 note that although the sample size is slightly below the targeted 200 participants (Hair,
377 Anderson, Tatham, & Black, 2005), there is no absolute standard regarding adequate sample
378 size (Muthen & Muthen, 2002) and this rule of thumb has been suggested by some to be
379 simplistic (Wolf, Harrington, Clark, & Miller, 2013). Following Hinkin’s (1995)
380 recommendations for an item-to-response ratio ranging from 1:4 to 1:10, an acceptable
381 sample size in the current study would be between 88 and 200 subjects. Also, through a
382 power analysis program with an anticipated effect size of .20 at a probability level of .05 and
383 at a statistical power level of .80 (Westland, 2010), the researchers concluded that 123
384 respondents would suffice for the current study. Additionally, the PLS algorithm that was
385 used to analyze data in the current research is not sensitive to small sample sizes (Hair,
386 Ringle, & Sarstedt, 2011). Thus, the current sample was deemed suitable to test the research
388 Respondents were between 18 and 67 years old (M = 27.11; SD = 9.27). The majority
389 of the sample was male (59.4%) and had a university degree (70%). Most respondents were
390 from the United States (42.4%), Mexico (24.3%), and the United Kingdom (8.5%). A total of
391 86 different soccer players were mentioned as participants’ favorite players, of which 90.7%
392 played in one of the top five European leagues (i.e. England, Germany, Spain, Italy and
393 France) and only 2.8% were female. The five most mentioned players were Lionel Messi
394 (12.9%), Cristiano Ronaldo (6.7%), Javier “Chicharito’ Hernandez (6.2%), Mesut Ozil
395 (5.6%), and Thomas Mueller (3.4%). The wide range of mentioned athletes made it possible
396 to have a better understanding of athlete brands because it was not focused on one specific
398 Materials
399 The questionnaire contained items measuring athlete brand image, consumer self-
400 brand connection, athlete commitment, team commitment and athlete sponsor preference.
401 These measures were adapted from existing literature (e.g., Arai et al., 2013; Biscaia et al.,
402 2013; Escalas & Bettman, 2003; Ross, Russell, & Bang, 2008; Tsiotsou, 2013) and a process
403 of content and face validity. To capture athlete on-field and off-field associations, a list of
404 items influencing athlete brand image was developed based on Arai et al. (2013, 2014),
405 Hasaan et al. (2018) and Ross et al. (2008). The content and face validity of the items was
406 assessed by a panel of experts. Eight sport management academics were provided with
407 detailed information about the purpose of this study, a list of constructs containing the
408 associated definition and items. These academics were asked to rate each item on a 5-point
409 Likert scale (1 = does not reflect construct at all; 5 = reflects construct very well) in terms of
410 each criteria; relevance, representativeness and clarity. They were also asked to provide
411 qualitative feedback to improve the face validity of each item. Items with an average rating
412 below four (80% threshold; Polit & Beck, 2006) were removed. The item most accurately
413 measuring the core of each brand association was selected to be included in the questionnaire.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 17
414 The use of single-item measures followed previous brand association research (e.g., Kunkel
415 et al., 2014; 2017), and was deemed appropriate because they captured the core of each brand
416 association (for a review see Rossiter, 2002) as determined in the expert feedback.
417 Additionally, single-item measures have been deemed preferable to multi-item measures for
418 online data collection as respondent fatigue was reduced and respondent frustration for
419 addressing similar questions about the same construct, such as asking three times whether an
420 athlete is attractive, could be avoided (cf., Bergkvist & Rossiter, 2007). Thus, athlete on-field
421 image was measured with six items and athlete off-field image was measured with seven
422 items.
423 Consumer self-brand connection, athlete commitment, team commitment and athlete
424 sponsor preference were measured with multi-item constructs as they represented complex
425 psychographic constructs (cf., Rossiter, 2002). Consumer self-brand connection was
426 measured with three items adapted from Escalas and Bettman (2003) and Dwivedi et al.
427 (2015). Athlete commitment was measured with two items adapted from Tsiotsou (2013) and
428 Sumino and Harada (2004), while two items were also used to capture team commitment that
429 were derived from Wu et al. (2012) and Sumino and Harada (2004). In turn, athlete sponsor
430 preference was measured with two items derived from Biscaia et al. (2013). The
431 questionnaire was optimized for mobile devices and 5-point Likert scale items anchored from
432 [1] = strongly disagree to [5] = strongly agree were used. At the beginning of the
433 questionnaire, participants were asked who their favorite player was. Then, before being
434 exposed to the sponsor-related questions, they were asked “Which of the following brands
435 sponsor the athlete” with leading sport equipment manufacturers being listed (i.e., Nike,
436 Adidas, Reebok, Umbro, Under Armour). The sponsorship questions were then linked to the
437 sponsor that the respondent indicated and only respondents who correctly indicated the actual
439 Given the study is cross-sectional with both independent and dependent variables
440 being collected from the same source at one moment in time, procedural remedies to alleviate
441 concerns about common method variance (CMV) bias were adopted (Hulland et al., 2018;
442 Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). First, the draft questionnaire was subjected
443 to a content and face validity procedure by an expert panel (Polit & Beck, 2006). Then, the
444 final items were randomized with a separation of dependent (athlete commitment, team
445 commitment, sponsor preference) and independent variables (on-field attributes, off-field
446 attributes, and self-brand connection) into different sections of the questionnaire (Hulland et
449 Data were analyzed via SPSS version 24 and SmartPLS version 3. Skewness and
450 Kurtosis were examined to evaluate data distribution. The proposed hypotheses were tested
451 using structural equation modeling (SEM). Data analysis was guided by considerations of the
452 direction of causality between a construct and its measures to avoid inaccurate conclusions
453 about the structural relationships between constructs (Jarvis, MacKenzie, Podsakoff, 2003).
455 formative measures (i.e. direction of causality is from items to construct) where the meaning
456 of the brand was derived from the cumulative effect of the unique brand associations (for a
457 detailed review, see Kunkel et al., 2017). Following a formative approach, the items form the
458 latent constructs with arrows pointing from the item to the construct. Thus, unique brand
459 associations did not have to be conceptually interchangeable, did not have to covary, and
460 could have different antecedents (cf., Finn & Wang, 2014; Jarvis et al., 2003). This approach
461 is consistent with previous literature on sport consumers (e.g. Uhrich & Benkenstein, 2012)
462 and followed Kunkel et al.’s (2017) recommendation that formative measures are preferred to
463 reflective measures when analyzing sport brand associations. For example, Kunkel and
464 colleagues treated consumers’ evaluation of the logo of a sport league and whether the league
ATHLETE BRANDS AND SELF-BRAND CONNECTION 19
465 has star players as factors forming the brand image of the sport league. Similarly, an athlete
466 could have an interesting life story but not be physically attractive, yet, these associations are
467 both related to off-field image. Psychographic constructs such as consumer self-brand
468 connection, athlete commitment, team commitment and athlete sponsor preference were
470 internal consistency, average variance extracted (AVE) and squared-correlation tests of
472 Partial Least Squares (PLS) SEM regression analysis with bootstrapping was
473 employed to analyze the hypothesized model. PLS-SEM analysis was selected over
474 covariance-based SEM, as it allowed for the inclusion of formative and reflective measures
475 and can handle small sample sizes (Hair et al., 2011). Guidelines recommended by Hair et al.
476 (2011) were followed to assess the model. To evaluate formative constructs, tests for multi-
477 collinearity examined whether each indicator’s variance inflation factor (VIF) was less than
478 the recommended threshold of five [5], and tests for validity examined whether the parameter
479 estimates for each indicator was statistically significant (at the 5% level) after a
480 nonparametric bootstrapping procedure of 5,000 resamples (Hair et al., 2011). These tests
481 replace standard tests of covariance-based SEM, such as the average variance extracted of the
482 latent formative construct or factor loading thresholds of over .60 (for a detailed review, see
483 Hair et al., 2011). This process represents the preferred method for examining models with
484 mediating variables (Cheung & Lau, 2008) and followed previous sport management research
485 (cf., Magnusen, Kim, & Kim, 2012). Mediation effects examined whether the confidence
486 intervals for bootstrapping procedures were significantly different from zero. These
487 bootstrapping tests provided a robustness check on the standard error and corresponding
488 statistical significance of the proposed mediation effects. The direct, indirect, and total effects
489 of the proposed model were tested via PLS-SEM regression analysis.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 20
490 Results
492 Descriptive statistics are presented in Table 2. Data were positively skewed with
493 mean scores significantly above three (3), which represents the mid-point of the 5-point
494 Likert scale items, for all athlete brand image associations. However, skewness (values
495 smaller than -1.779) and kurtosis (values smaller than 3.364) indicated that data distribution,
496 and consequently multi-collinearity, was not an issue (Hair et al., 2011). The validity
497 assessment of on-field and off-field image associations showed that Fair Play, Impact, and
498 Skill had no statistically significant relationship with on-field brand image associations; yet,
499 following Jarvis et al.’s (2003) recommendations for formative models, these items were
500 retained to avoid omitting unique parts of the composite variable and prevent restricting the
501 theoretical domain of the construct. The mean score for on-field image (M = 4.54, SD = .779)
502 was higher than for off-field image (M = 3.96, SD = .959), while athlete commitment was the
503 outcome variable with the highest mean score (M = 4.23, SD =.923). In addition, Cronbach α
504 scores and composite reliability (CR) scores of the reflective constructs exceeded the
505 recommended threshold of .70 (Nunnally & Bernstein, 1994), providing support for the
508 The correlation matrix for the constructs and average variance extracted (AVE) tests
509 of discriminant validity are presented in Table 3. The AVE scores varied from .655 (self-
510 brand connection) to .899 (athlete sponsor preference), exceeding the recommended threshold
511 of .50, and provided evidence of convergent validity. In addition, evidence of discriminant
512 validity was accepted given that the correlation coefficients were lower than the suggested
513 criterion of .85 (Kline, 2005) and none of the squared correlations exceeded the AVE values
514 for each associated construct (Hair et al., 2011). Both formative constructs (on-field and off-
515 field brand image) and reflective constructs (self-brand connection, athlete commitment, team
ATHLETE BRANDS AND SELF-BRAND CONNECTION 21
516 commitment and athlete sponsor preference) showed good psychometric properties.
520 The results of the structural model are pictorially presented in Figure 2 and
521 bootstrapping results are presented in Table 4. The model explained 39.2% of self-brand
522 connection, 46.2% athlete commitment, 11.9% team commitment, and 11.3% athlete sponsor
523 preference. The path coefficients from on-field image to athlete commitment and team
524 commitment were not significant (p > .05). As such, H1a and H2a were not supported; yet,
525 there was a significant positive effect of on-field image on athlete sponsor preference (β =
526 .251, p < .05) supporting H3a. In turn, off-field image was positively related to athlete
527 commitment (β = .313, p < .05) and team commitment (β = .326, p < .05), but not to athlete
528 sponsor preference (p > .05). Therefore, H1b and H2b were supported while H3b was not
529 supported. The relationship between athlete image and self-brand connection was only
530 significant for on-field image (β = .602), thus supporting H4a, but not H4b. In turn, self-
531 brand connection showed a significant positive effect on athlete commitment (β = .490), but
532 not on neither team commitment nor athlete sponsor preference (p > .05). Therefore, H5a was
533 supported while H5b and H5c were not supported. Self-brand connection did not show a
534 significant mediating effect on the relationship between on-field image with athlete
535 commitment, team commitment and athlete sponsor preference (p > .05) not supporting H6a,
536 H6b and H6c. Finally, self-brand connection partially mediated the relationship between off-
537 field image and athlete commitment (β = .295), supporting H7a, but the effect was not
538 significant for team commitment and athlete sponsor preference. Therefore, H7b and H7c
541 Discussion
542 The purpose of this study was to explore the role of athlete brand image dimensions
543 (on-field and off-field) on athlete commitment, team commitment and athlete sponsorship
544 preference, as well as the mediating effects of self-brand connection on these relationships.
545 The current study extends previous research by conceptualizing and measuring on-field and
546 off-field associations that contribute to athlete brand image, as well as by examining the role
547 of these associations on fan outcomes to different brands in the sport ecosystem. More
548 specifically, it extends the body of knowledge related to athlete branding by (1) empirically
549 testing the associations that contribute to athletes’ on-field and off-field brand image, (2)
550 exploring the impact of athlete brand image on consumer commitment toward the athlete and
551 associated team and preference of athlete-related sponsors, and (3) examining the mediating
552 role of self-brand connection on the relationship between athlete brand image and other
553 outcomes (i.e., athlete commitment, team commitment, and athlete sponsor preference. We
554 extend previous studies focusing on athlete brand outcomes that have either not incorporated
555 fans’ view (e.g. Hasaan et al., 2018; Parmentier & Fischer, 2012) or simply examined fan
556 reactions to athletes (Arai et al., 2013) by examining reactions to teams and sponsors. Also,
557 the current study extends previous research by conceptualizing and measuring on-field and
558 off-field associations that contribute to athlete brand image, as well as by examining the
559 effect of these associations on the athlete’s team and sponsors. This is of vital importance
560 because brands do not act in isolation and the sport ecosystem is composed by a variety of
561 brands with potential to influence each other (Chanavat et al., 2016).
563 The results show that respondents evaluated on-field brand associations positively
564 with all mean scores above the mid-point. The positive evaluation is related to the fact that
565 the mentioned favorite athletes competed at the highest level, with many having global star
ATHLETE BRANDS AND SELF-BRAND CONNECTION 23
566 status. The brand associations of Effort, Achievement, and Style of Play had a significant
567 positive relationship with athletes’ on-field brand image, whereas the brand associations Fair
568 Play, Impact, and Skill did not show a significant positive relationship with athletes’ on-field
569 brand image. These findings indicate that while consumers acknowledged the Skill and
570 Impact of a player on the field, supporting propositions by previous athlete brand research
571 (e.g., Arai et al., 2013, 2014; Braunstein & Zhang, 2005), these associations did not
572 distinguish the athlete brand from other athletes who also have a high level of skill and are
573 impactful on the field. This may have been related to the fact the athletes in our sample have
574 all a high-profile. Nevertheless, there may be instances where athletes have built their brand
575 on unique associations that are not-significant for the majority of athletes. Brand associations
576 such as Skill may represent the foundation to generate brand awareness which together with
577 brand image form brand equity. Conversely, the athlete’s individual Style of Play and Effort
578 on the field, as well as personal Achievements, seem to build the on-field brand more
579 effectively, providing initial empirical support for recent athlete brand research (e.g., Arai et
580 al., 2013, 2014; Hasaan et al., 2018; Väätäinen & Dickenson, 2018).
581 The examined off-field brand associations were also evaluated positively with all
582 mean scores above the mid-point, and all examined associations showed a significant positive
583 relationship with athletes’ on-field brand image. The findings related to the associations Body
584 Fitness and Physical Appearance support Arai et al.’s (2013, 2014) propositions that physical
585 attractiveness contributes to athletes’ brand image. Similarly, results generalize Geurin-
586 Eagleman and Burch’s (2016) findings that posting sexually suggestive photos generates
587 higher engagement than other content on athletes’ social media profiles does. It is also in line
588 with sponsorship effectiveness research suggesting that physical appeal moderates the
589 effectiveness of the endorser (e.g., Till & Busler, 2000; Yu, 2005). Similarly, the significant
590 effects of Lifestyle, Personality, Social Responsibility, and Role Model support Arai et al.’s
ATHLETE BRANDS AND SELF-BRAND CONNECTION 24
591 (2013, 2014) propositions that a marketable lifestyle adds to an athlete’s brand image; while
592 findings for Personality and Culture are consistent with Hassan et al.’s (2018) propositions.
593 Overall, the formative approach to measuring brand image revealed the brand
594 associations that significantly influenced athlete brand image, and indicates that there are
595 several aspects that contribute to an athlete’s brand image. For example, Cristiano Ronaldo
596 has on-field achievements with his current (i.e., Juventus) and former clubs (e.g., Manchester
597 United and Real Madrid), and the Portuguese National team, while his off-field brand can be
598 linked to being a father, sex-symbol, a model, and entrepreneur (The Guardian, 2016). As
599 such, a valuable athlete image might consist both of sport-related performance and personal
600 life (Geurin-Eagleman & Burch, 2016; Parmentier et al., 2013; Parmentier & Fischer, 2012).
601 This is evident in the current study through the significant and positive correlation between
602 on-field image and off-field image (.468; Table 3). In this sense, one can argue that athletes
603 have the potential to become recognizable personalities outside of the sports environment,
604 and the combination of their athletic performances and public personas is vital for the
605 development of a strong brand (Walsh & Williams, 2017). The current study extends
606 previous research by conceptualizing and measuring on-field and off-field associations that
607 contribute to athlete brand image and its subsequent effects on associated brands.
609 The two athlete brand image dimensions influenced outcome variables differently.
610 The current study broaden research findings on team brand associations (e.g., Kunkel et al.,
611 2016) and the FAN model (Funk & James, 2004) to athletes, indicating that consumers’
612 perceptions of the on- and off-field athlete brand impact their connection to the athlete (Arai
613 et al., 2013, 2014; Hassan et al., 2018) and brands that are related to the athlete brand (Yu,
614 2005). Consequently, the findings provide empirical support to the theoretical propositions by
615 Williams et al. (2015) that athletes hold various roles in a sport organization's brand
ATHLETE BRANDS AND SELF-BRAND CONNECTION 25
616 architecture. The spill-over effect observed from teams to their league (Kunkel et al., 2017)
617 also applies for athletes and their teams extending knowledge from the associative network
618 memory model (Anderson, 1983) and meaning transfer model (McCracken, 1986) to the
620 Spill-over effects from athlete brand image to related entities differ by entity, which
621 contributes to deepen the understanding of how to manage an athlete’s brand image. Athlete
622 sponsor preference was influenced by athletes’ on-field image. As Carlson and Donovan
623 (2013, p. 193) stated, “numerous firms tie their brands to successful athletes with the
624 expectation that doing so will transfer the athlete’s positive attributes onto the brand”.
625 Likewise, the findings of this research indicate that on-field associations are more important
626 for sponsors than off-field associations. This suggests that sponsorship success follows on-
627 field success (Biscaia, Trail, Ross, & Yoshida, 2017). To this respect, footwear and apparel
628 maker Under Armour signing NBA star Stephen Curry has led to a strong growth of the
629 company’s basketball footwear (Business Insider, 2017). Conversely, athlete commitment
630 and team commitment were influenced by the athlete’s off-field image, highlighting the
631 importance for teams to hire athletes with strong off-field brand images. These findings
632 suggest that athletes are hybrid brands (Cortsen, 2013) and that the way their image is
633 managed outside the field of play is vital to increase followership for them and their teams
634 (Agyemang et al., 2015; Guerin-Eagleman & Burch, 2016). For example, David Beckham
635 has been commonly acknowledged as an example of how an athlete’s off-field image may
636 play a significant role to increase commitment toward their favorite team (Kerr & Gladden,
637 2008).
638 Consequently, these findings extend previous research on athlete brand by empirically
639 supporting the notion that both on-field and off-field brand image is important (Pegoraro &
640 Jinnah, 2012; Parmentier & Fischer, 2012) to invoke positive fan reactions. While previous
641 studies have suggested the existence of spill-over effects among brands within the sport
ATHLETE BRANDS AND SELF-BRAND CONNECTION 26
642 ecosystem (e.g. Biscaia et al., 2017; Kunkel et al., 2017), athlete’s on-field and off-field
643 brand image were not equally effective at predicting athlete commitment, team commitment
644 and athlete sponsor preference. Based on the current findings, performance-related
645 characteristics are more effective at promoting positive reactions toward athletes’ associated
646 sponsors, while life beyond the sport activity represents a more valuable component to
647 promote commitment toward the athlete and his/her team. Still, the athletes examined in this
648 study were among the top soccer players worldwide (e.g. Messi and Cristiano Ronaldo) and
649 the lack of variability in participants’ responses regarding on-field image may have limited a
650 deeper understanding of its role on fans’ reactions. However, the findings indicate that for
651 athletes at the top level, off-field associations rather than on-field associations influence fans’
654 The current study illuminates the role of self-brand connection on the relationship
655 fans establish with their favorite athletes and associated teams and sponsors. As the findings
656 indicate, self-brand connection played a role in the relationship between athlete brand image
657 and consumer’s commitment with the athlete. Notably, self-brand connection was not
658 significantly related to team commitment and athlete sponsor preference. These findings
659 suggest that fans’ personal feelings toward an athlete do not necessarily translate into positive
660 outcomes for related entities. Indeed, a strong athlete brand image is more important than
661 athletes who are relatable. However, findings show that self-brand connection directly
662 affected athlete commitment and mediated the relationship between off-field image and
663 athlete commitment. These findings support the assumption that people use brands to
664 represent self-images, and that this representation of the self is an important element in the
665 development of consumer loyalty toward athlete brands (Escalas & Bettman, 2003). For
ATHLETE BRANDS AND SELF-BRAND CONNECTION 27
666 example, Steph Curry has been described as “the rare superstar where an 8-year-old kid can
667 go, ‘He looks just like me. Maybe I can do it.’” (Pandian, 2017).
668 On-field brand image was not significantly related with self-brand connection
669 suggesting that participants did not perceive their self-concept to be reflected by the athlete
670 professional expertise. As noted by Morhart, Malär, Guèvremont, Girardin, and Grohmann
671 (2015), brands play an important role in consumers' identity, given that they tend to rely on
672 brands to express themselves. In the context of professional athletes, the non-significant
673 relationship between on-field image and self-brand connection may indicate difficulties for
674 fans to associate themselves with top performers. Escalas and Bettman (2003) stated that
675 consumers often select brands with meanings congruent with an aspect of their current self-
676 concept or possible self, while Pegoraro and Jinnah (2012) noted that fans mainly consume
677 sport brands loaded with meaning to which they can relate. These ideas are particularly
678 relevant in this case given that the professional soccer players in this study are some of the
679 top performers in the world and their on-field image was ranked very high. Consequently,
680 this high level of on-field performance may be perceived by fans as personally unachievable,
681 which may limit the development of a self-brand connection. In this instance, the aspirational
682 self (Fournier, 1998), not the perceived self, may play a bigger role. While fans may have
683 perceived some athletes to have unattainable on-field image, the significant role of off-field
684 image indicates that the brand associations related to the athlete’s personal life (e.g.,
685 personality, role model, lifestyle, culture) may also be used to create and define a consumer's
686 self-concept (Chaplin & John, 2005; Escalas, 2004). Further research that considers a more
687 diverse range of athletes would be helpful to understand better the relationships between fans,
690 The findings have several managerial implications germane to athlete brand
691 management. Considering the importance of the Effort association for on-field image,
692 athletes and managers should consider being more intentional in demonstrating their
693 commitment to the field of play in every competition. Such exertion often ingratiates the
694 athlete to the fan and can lead to an overall positive perception of the athlete brand. For
695 instance, consider a basketball player who dives for a loose ball or an attacking player in
696 soccer who sacrifices for his/her team and “tracks back” to deter the opposing team from
697 scoring. Such instances are often applauded and can help the player win over fans, and brand
698 managers should encourage this activity. Similarly, the associations of Style of Play and
699 Achievement suggest that athletes’ self-presentation to fans (e.g., social media) should take
700 the athletic action (Guerin-Eagleman & Burch, 2016) into consideration by highlighting
701 important career achievements and their distinctive style. For example, Cristiano Ronaldo’s
702 posts about match performances are often shared by thousands of followers contributing to
704 In addition, our findings support the idea that the effectiveness of endorsement is
705 moderated by their athletic performance (Yu, 2005). Therefore, athletes must create, maintain
706 and protect an image of success on the field. To this end, the use of social media seems to
707 offer strong potential for athletes and their associated sponsors (Pegoraro & Jinnah, 2012), as
708 the importance of celebrity endorsement has been suggested to strengthen over time (Spry,
709 Pappu, & Cornwell, 2011). Athletes should also pursue opportunities that would allow
710 individual and collective achievements. The example of Kevin Durant (NBA player) moving
711 from the Oklahoma City Thunder to the Golden State Warriors created various opportunities
712 for sponsoring brands, resulting in Durant having the third highest endorsement earnings of
713 all NBA players (Forbes, 2018). Indeed, his move was not solely based on basketball but to
714 capitalize on the technology investment opportunities in the Bay Area (Rovell, 2018). With
ATHLETE BRANDS AND SELF-BRAND CONNECTION 29
715 this in mind, brand managers should consider destinations and opportunities beyond merely
716 the sport an athlete plays, given that athletes have become valuable personal brands with
718 There are also implications tied to off-field brand image, particularly in reference to
719 its role on athlete commitment and team commitment, as well as self-brand connection by
720 fans. The importance of athlete off-field image at predicting fan outcomes extends Guerin-
721 Eagleman and Burch’s (2016) research highlighting how fans value athletes’ personal life
722 (i.e., life outside of sports). The findings also illustrate that Attractiveness is important. Thus,
723 conventional wisdom would hold that it is in athletes’ best interest to manage their body and
724 overall look to the best of their ability (e.g., Geurin-Eagleman & Burch, 2016). Moreover, the
725 role of Personality enforces the idea that brand personality regularly optimizes a consumer’s
726 connection with a brand (e.g., Aaker, 1997). It is thus beneficial for athletes to demonstrate
727 their unique personality, which can be done via social media by giving consumers first-hand
728 accounts of them engaging in activities aimed at highlighting their personality and provide
729 insights on their daily life (e.g., Geurin & Burch, 2017). Such glimpses keep fans entertained
730 and allow the athlete to stand out and create strong connections with the target audience.
731 Furthermore, athletes and teams should also take notice of the cultural background
732 finding. As sport becomes more globalized, leagues are attracting players from previously
733 unchartered territory. Teams have begun to notice the benefit of having athletes from various
734 backgrounds. For instance, Mohamed Salah’s athletic performance and actions off the field
735 have contributed to increase fans’ sense of connection with him (The Guardian, 2018) and
736 brought an Egyptian following to Liverpool Football Club. Lastly, in the same vein as people
737 like socially responsible companies, they are attracted to athletes who are good role models
738 that “do good” within their communities. One might argue that this spills-over onto teams,
739 and that it would be advisable for athletes to seek opportunities to engage in athlete
ATHLETE BRANDS AND SELF-BRAND CONNECTION 30
740 citizenship (Agyemang, 2014), whereby they influence society in a positive way and utilize it
741 strategically to build their personal brand (e.g., Kunkel, Scott, & Beaton, 2016).
743 As with any research, there are limitations in this study that may have influenced the
744 results and provide opportunities for future research. First, data were collected only via two
745 social media platforms (i.e., Facebook and Twitter), and although the survey link was widely
746 distributed using various related hashtags, the study only targeted soccer fans. Also, the
747 results were derived from a convenience sample of individuals who may have taken the
748 survey because of their interest in soccer, followership of specific online groups and high
749 profile-athletes, or even the potential incentives that were provided. The collection of a
750 larger, representative sample including fans and athletes from different sports and using
751 different methods (e.g., online surveys and paper-and-pencil) would help future studies to
752 further investigate the accuracy of the relationships tested in the current research and increase
753 the generalizability of the results. As the current study focused on top athletes, which may
754 have been the reason why Skill was not a statistical differentiator to drive their brand image,
755 we recommend future research to include brand awareness and use a more diverse pool of
756 athletes, including second tier athletes and more female athletes, to further understand the
757 impact of athlete image dimensions on fans reactions toward those athletes and related
758 entities.
759 Second, athletes with different brand images were combined into the same model.
760 Although the free choice of athlete maximized the opportunity to capture participants with
761 various levels of self-brand connection, this method sacrificed certain control over the target
762 athletes. For example, the study could not exclude confounding variables such as likability.
763 Considering that fans tend to react to brands in the sport environment in different ways
764 (Biscaia et al., 2017), additional research could compare the current model across specific
765 athletes to better understand the relationships between athlete image dimensions, self-brand
ATHLETE BRANDS AND SELF-BRAND CONNECTION 31
766 connection and fan reactions to the athlete and related entities. Given that sport athletes tend
767 to be subjected to extensive scrutiny and high fan expectations (Summers & Johnson Morgan,
768 2008), of particular interest should be cases in which on-field and off-field image are
769 incongruent. For example, when excellent soccer players are involved in major incidents,
770 such as Lionel Messi’s tax fraud, or off-court abuse, such as Cristiano Ronaldo’s rape
771 allegations. The impact of athlete transgression on different entities linked to the athlete
772 would be both theoretically and practically relevant. This would likely allow to extend Lee
773 and Kwak’s (2016) research, which demonstrates consumers can decouple more from the
774 transgression related to the job performance (i.e., doping) than transgressions not related to
775 the job performance (i.e., fraud). By comparing different athletes, control variables such as
777 A third limitation and research opportunity involves the inclusion of self-brand
778 connection measures allowing a distinction between perceived self and desired self. Previous
779 studies have suggested that the usage of products and brands by individuals is often linked to
780 desired self-images (Escalas, 2004; Cătălin & Andreea, 2014). Thus, measuring both
781 consumer actual self-image and desired self-image may contribute to our understanding of
782 the importance of on-field and off-field brand image dimensions and their role in fan
784 Fourth, the role of different types of media platforms (e.g., athletes mainly promoted
785 through traditional media, reality shows, social media; Escalas & Bettman, 2017) on the
787 consumers develop different connections with brands within the sport environment, and
788 further outcomes could be included in the current model. For example, self-brand connection
789 with an athlete brand may lead consumers to appropriate social needs (Escalas and Bettman,
790 2015), and recent studies have highlighted the importance of sports to increase individuals’
791 well-being (e.g., Inoue, Sato, Du, & Funk, 2017). Therefore, examining how self-brand
ATHLETE BRANDS AND SELF-BRAND CONNECTION 32
792 connections with athletes relates to sport fans’ well-being may help expand our knowledge
794 Fifth, we only focused on sponsors in the traditional sense of companies sponsoring
795 athletes to endorse their brand. However, an increasing number of athletes are becoming
796 entrepreneurs (Ratten, 2015) by launching their own brands, such as Cristiano Ronaldo
797 selling jeans and boxer shorts with his CR7 brand, or Zlatan Ibrahimovic selling hoodies and
798 caps with his A-Z brand. Thus, future research should investigate the role of athlete brand
799 associations and self-brand connection on fans’ purchase behavior toward the athlete branded
800 merchandise vs. traditional sponsored merchandise. This would likely contribute to a better
801 understanding of athlete entrepreneurship beyond the influencer status. In addition, the
802 examination of spill-over effects between athlete sponsors and team sponsors may also prove
803 to be beneficial to extend previous research on sport brand architecture (e.g., Kunkel et al.,
804 2017).
805 Conclusion
806 In summary, the current study represents an initial effort to understand how
807 consumers' perception of athlete brand image influence their commitment toward the athlete,
808 his/her team and associated sponsors, as well as how these relationships are mediated by their
809 self-brand connection. It provides a deeper understanding of the aspects that contribute to the
810 development of fan relationships with not just their favorite athlete but also other brands in
811 the sport ecosystem such as teams and sponsors. In particular, the findings indicate that on-
812 field and off-field image impact outcome variables differently. On-field athlete image is
813 important to increase athlete sponsor preference, while off-field image leads to an improved
814 athlete commitment and team commitment. In addition, self-brand connection mediates the
815 relationship between off-field image and athlete commitment. Taken together, the findings
816 from the current research suggest a new perspective on athlete branding and shed light on the
ATHLETE BRANDS AND SELF-BRAND CONNECTION 33
817 importance of athlete brand image dimensions on fan outcomes to brands involved in the
819
ATHLETE BRANDS AND SELF-BRAND CONNECTION 34
820 References
821 Aaker, J. (1997). Dimensions of brand personality. Journal of Marketing Research, 34(3), 347-
822 356.
823 Aaker, J. Benet-Martinez, V., & Garolera, J. (2001). Consumption symbols as carriers of
824 culture: A study of Japanese and Spanish brand personality constructs. Journal of
825 Personality and Social Psychology, 81(3), 492-508.
826 Agyemang, K. J. A. (2014). Toward a framework of “athlete citizenship” in professional sport
827 through authentic community stakeholder engagement. Sport, Business and Management:
828 An International Journal, 4(1), 26-27.
829 Agyemang, K. J. A., Williams, A. S., & Kim, D. Y. (2015). “Scandalous!”: Reputation,
830 impression management, and employee assistance programs (EAPs) in the NBA. Sport
831 Management Review, 18(4), 609-617.
832 Allen, C. T., Fournier, S., & Miller, F. (2008). Brands and their meaning makers. In C. P.
833 Haugtvedt, P. Herr, F. R. Kardes (Ed.) Handbook of Consumer Psychology (pp. 781-822).
834 New York: Lawrence Erlbaum.
835 Amos, C., Holmes, G., & Strutton, D. (2008). Exploring the relationship between celebrity
836 endorser effects and advertising effectiveness. International Journal of Advertising, 27(2),
837 209-234.
838 Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University
839 Press.
840 Arai, A., Ko, Y. J., Kaplanidou, K. (2013). Athlete brand image: scale development and model
841 test. European Sport Management Quarterly,13(4), 383-403.
842 Arai, A., Ko, Y. J., & Ross, S. (2014). Branding athletes: exploration and conceptualization of
843 athlete brand image. Sport Management Review, 17(2), 97-106.
844 Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of other in the self scale and the
845 structure of interpersonal closeness. Journal of Personality and Social Psychology, 63(4),
846 596-612.
847 Badenhausen, K. (2016). Why Cristiano Ronaldo's $1 billion Nike deal may be a bargain for
848 sportswear giant. Accessed May 26th, 2019, from
849 https://www.forbes.com/sites/kurtbadenhausen/2016/12/02/cristiano-ronaldos-1-billion-
850 nike-deal-is-a-bargain-for-sportswear-giant/#71fbfbd5a918
851 Balachander, S., & Ghose, S. (2003). Reciprocal spillover effects: A strategic benefit of
852 brand extensions. Journal of Marketing, 67(1), 4-13.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 35
853 Barros, C. P., & Silvestre, A. L. (2006). An evaluation of the sponsorship of Euro 2004.
854 International Journal of Sports Marketing & Sponsorship, 7(3), 192–212.
855 Bergkvist, L., & Rossiter, J. R. (2007). The predictive validity of multiple-item versus single-
856 item measures of the same constructs. Journal of Marketing Research, 44, 175-184.
857 Biscaia, R., Correia, A., Ross, S., Rosado, A, & Marôco, J. (2013). Sport sponsorship: The
858 relationship between team loyalty, sponsorship awareness, attitude toward the sponsor and
859 purchase intentions. Journal of Sport Management, 27(3), 288-302.
860 Biscaia, R., Ross, S., Yoshida, M., Correia, A., Rosado, A., & Marôco, J. (2016). Investigating
861 the role of fan membership on perceptions of team brand equity in football. Sport
862 Management Review, 19(2), 157–170.
863 Biscaia, R., Trail, G., Ross, S., & Yoshida, M. (2017). A model bridging team brand experience
864 and sponsorship brand experience. International Journal of Sports Marketing &
865 Sponsorship, 18(4), 380-399.
866 Brandes, L., Franck, E., & Nüesch, S. (2008). Local heroes and superstars: An empirical
867 analysis of star attraction in German soccer. Journal of Sports Economics, 9(3), 266-286.
868 Braunstein, J. R., & Zhang, J. J. (2005). Dimensions of athletic star power associated with
869 Generation Y sports consumption. International Journal of Sports Marketing and
870 Sponsorship,6(4), 37-62.
871 Business Insider (2017). Under Armour hit the jackpot with its Stephen Curry bet. Assessed
872 21 October at: https://www.businessinsider.com/under-armour-hit-the-jackpot-with-
873 stephen-curry-bet-2015-5
874 Carlson, B. D., & Donavan, D. T. (2013). Human brands in sport: athlete brand personality and
875 identification. Journal of Sport Management, 27(3), 193-206.
876 Cătălin, M. C., & Andreea. P. (2014). Brands as a mean of consumer self-expression and
877 desired personal lifestyle. Procedia – Social and Behavioral Sciences, 109(8), 103-107.
878 Chanavat, N., Desbordes, M., & Dickson, G. (2016). Sponsorship networks: toward an
879 innovative model. Sport, Business and Management: An International Journal, 6(4), 424-
880 439.
881 Chanavat, N., Martinent, G., & Ferrand, A. (2009). Sponsor and sponsees interactions: Effects
882 on consumers’ perceptions of brand image, brand attachment, and purchase intention.
883 Journal of Sport Management, 23(5), 644-670.
884 Chaplin, L.N. & John, D. R. (2005). The development of self-brand connections in children
885 and adolescents. Journal of Consumer Research, 32(1), 119-129.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 36
886 Chadwick, S., & Burton, N. (2008). From Beckham to Ronaldo: assessing the nature of football
887 player brands. Journal of Sponsorship, 1(4), 307-317.
888 Cheung, G. W. & Lau, R. S. (2008). Testing mediation and suppression effects of latent
889 variables bootstrapping with structural equation models. Organizational Research Methods,
890 11(2), 296-325.
891 Cobbs, Groza, & Rich (2015). Brand spillover effects within a sponsor portfolio: The
892 interaction of image congruence and portfolio size. Marketing Management Journal, 25(2),
893 107-122.
894 Cortsen, K. (2013). Annika Sörenstam – a hybrid personal sports brand. Sport, Business and
895 Management: An International Journal, Vol. 3(1), 37–62.
896 Daniels, J., Kunkel, T., & Karg, A. (2019). New brands: Contextual differences and
897 development of brand associations over time. Journal of Sport Management, 33(2), 133-
898 147.
899 Dwivedi, A, Johnson, L. W., & McDonald, R. E. (2015). Celebrity endorsement, self-brand
900 connection and consumer-based brand equity. Journal of Product & Brand Management,
901 24 (5), 449-461.
902 Escalas, J. E. (2004). Narrative processing: Building consumer connections to brands.
903 Journal of Consumer Psychology, 14(1-2), 168-179.
904 Escalas, J. E., & Bettman, J. R. (2003). You are what they eat: The influence of reference
905 groups on consumers' connections to brands. Journal of Consumer Psychology, 13(3), 339-
906 348.
907 Escalas, J. E., Bettman, J. R. (2009). Connecting with celebrities: Celebrity endorsement,
908 brand meaning, and self-brand connections. Journal of Marketing Research, 13(3), 339-
909 348.
910 Escalas, J. E., Bettman, J. R. (2015). Managing brand meaning through celebrity
911 endorsement. Brand Meaning Management, 12(1), 29-52.
912 Escalas, J. E., & Bettman, J. R. (2017). Connecting with celebrities: How consumers
913 appropriate celebrity meanings for a sense of belonging. Journal of Advertising, 46(2),
914 297-308.
915 Farrelly, F. J., & Quester, P. G. (2005). Examining important relationship quality constructs
916 of the focal sponsorship exchange. Industrial Marketing Management, 34(3), 211-219.
917 Fink, J. S., Parker, H. M., Cunningham, G. B., Cuneen, J. (2012). Female athlete endorsers:
918 Determinants of effectiveness. Sport Management Review, 15(1), 13-22.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 37
919 Finn, A., & Wang, L. (2014). Formative vs. reflective measures: Facets of variation. Journal
920 of Business Research, 67, 2821-2826.
921 Forbes (2017). Full list: The world's 50 most valuable sports teams 2017. Accessed 09
922 January at: https://www.forbes.com/sites/kurtbadenhausen/2017/07/12/full-list-the-worlds-
923 50-most-valuable-sports-teams-2017/#7300ee014a05.
924 Forbes (2018). The NBA’s highest paid players for 2018. Accessed 7 February at:
925 https://www.forbes.com/sites/kurtbadenhausen/2018/02/07/the-nbas-highest-paid-players-
926 on-and-off-the-court-for-2018/#2a9824a32853
927 Fournier, S. (1998). Consumers and their brands: Developing relationship theory in consumer
928 research. Journal of Consumer Research, 24(4), 343-373.
929 Funk, D. C., James, J. (2004). The fan attitude network (FAN) model: Exploring attitude
930 formation and change among sport consumers. Sport Management Review, 7(1), 1-26.
931 Funk, D. C., Beaton, A., & Alexandris, K. (2012). Sport consumer motivation: Autonomy and
932 control orientations that regulate fan behaviours. Sport Management Review, 15(3), 355-
933 367.
934 Geurin, A. & Burch, L. (2017). User-generated branding via social media: An examination of
935 six running brands, Sport Management Review. 20, 273-284
936 Geurin-Eagleman, A. N., & Burch, L. M. (2016). Communicating via photographs: A gendered
937 analysis of Olympic athletes’ visual self-presentation on Instagram. Sport Management
938 Review, 19(2), 133-145.
939 Gladden, J. M., & Funk, D. C. (2001). Understanding brand loyalty in professional sport:
940 Examining the link between brand associations and brand loyalty. International Journal of
941 Sports Marketing & Sponsorship, 3(1), 67–91.
942 Hair, J. F., Anderson, R. E., Tatham, R. L. and Black, W.C. (2005). Multivariate Data Analysis
943 (5th ed). Upper Saddle River, NJ: Prentice-Hall.
944 Hair, J. F. Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a silver bullet. Journal of
945 Marketing Theory and Practice, 19, (2), 139-151.
946 Hasaan, A., Kerem, K., Biscaia, R., & Agyemang, K. (2018). A conceptual framework to
947 understand the creation of athlete brand and its consequences. International Journal of
948 Sports Marketing and Management, 18(3), 169-198.
949 Hinkin, T. R. (1995). A review of scale development practices in the study of organizations.
950 Journal of Management, 21(5), 967–988.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 38
951 Hulland, J. Baumgartner, H., & Smith, K. M. (2018). Marketing survey research best practices:
952 evidence and recommendations from a review of JAMS articles. Journal of the Academy of
953 Marketing Sciences, 46(1), 92-108.
954 Inoue, Y., Sato, M., Du, J., & Funk, D. C. (2017). Sport spectatorship and life satisfaction: A
955 multicountry investigation. Journal of Sport Management, 31(4), 419–432.
956 Jarvis, C. B., MacKenzie, S. B., & Podsakoff, P. M. (2003). A critical review of construct
957 indicators and measurement model misspecification in marketing and consumer
958 research. Journal of Consumer Research, 30, 199–218.
959 Keller, K.L. (1993). Conceptualizing, measuring, and managing customer-based brand equity.
960 Journal of Marketing, 57(1), 1-22.
961 Kerr, A. K., & Gladden, J. M. (2008). Extending the understanding of professional team brand
962 equity to the global marketplace. International Journal of Sport Management and
963 Marketing, 3(1–2), 58-77.
964 Kline, R. B. (2005). Principles and practices of structural equation modeling. New York, NY:
965 The Guilford Press.
966 Kunkel, T., Doyle, J. P., Funk, D. C., Du, J., & McDonald, H. (2016). The development and
967 change of brand associations and their influence on team loyalty over time. Journal of Sport
968 Management, 30(2), 117-134.
969 Kunkel, T., Funk, D. C., & King, C. (2014). Developing a conceptual understanding of
970 consumer-based league brand associations. Journal of Sport Management, 28(1), 49-67.
971 Kunkel, T., Funk, D. C., & Lock, D. (2017). The effect of league brand on the relationship
972 between the team brand and behavioral Intentions: A formative approach examining brand
973 associations and brand relationships. Journal of Sport Management, 31(4), 317-332.
974 Kunkel, T., Scott, O., & Beaton, A. (2016). Interview with Michael Lahoud, professional
975 Soccer player: Lessons of personal athlete branding via social media. International
976 Journal of Sport Communication, 9(4), 415-423.
977 Lee, J. S., & Kwak, D. H. (2016). Consumers’ responses to public figures’ transgression:
978 Moral reasoning strategies and implications for endorsed brands. Journal of Business
979 Ethics, 137, 101-131.
980 Lunardo, R., Gergaud, O., & Livat, F. (2015). Celebrities as human brands: an investigation of
981 the effects of personality and time on celebrities’ appeal. Journal of Marketing
982 Management, 31(5–6), 685–712.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 39
983 Magnusen, M. Kim, J. W., & Kim, Y. K. (2012) A relationship marketing catalyst: the salience
984 of reciprocity to sport organization–sport consumer relationships. European Sport
985 Management Quarterly, 12(5), 501-524.
986 McCracken, G. (1986). Culture of consumption: A theoretical account of the structure and
987 movement of the cultural meaning of consumer goods. Journal of Consumer Research,
988 13(1), 71-84.
989 McGuire, W. J. (1985). Attitude and attitude change, In G. Lindzey & E. Aronson (Eds.),
990 Handbook of social psychology (pp. 233-346). New York, NY: Random House.
991 Montoya, P. (2002). The Personal Branding Phenomenon. Nashville, TN: Personal Branding
992 Press.
993 Moore, D. J., & Homer, P. M. (2008). Self-brand connections: The role of attitude strength and
994 autobiographical memory primes. Journal of Business Research, 61(7), 707–714.
995 Morhart, F., Malär, L., Guèvremont, A., Girardin, F., & Grohmann, B. (2015). Brand
996 authenticity: An integrative framework and measurement scale. Journal of Consumer
997 Psychology, 25(2) 200-218.
998 Müller, O., Simons, A., & Weinmann, M. (2017). Beyond crowd judgments: Data-driven
999 estimation of market value in association football. European Journal of Operational
1000 Research, 263(2), 611-624.
1001 Muthen, L.K., & Muthen, B.O. (2002). How to use a Monte Carlo study to decide on sample
1002 size and determine power. Structural Equation Modeling, 9, 599–620.
1003 Nunnally, J. C., & Berstein, I. H. (1994). Psychometric theory (3rd ed). New York, NY:
1004 McGraw-Hill.
1005 Oliver, R. L. (1999). When consumer loyalty? Journal of Marketing, 63, 33-44.
1006 Pandian, A. (2017). Steve Kerr suggests Steph Curry might be more popular with kids than
1007 Michael Jordan. Retrieved on July 4th, 2018, from https://www.cbssports.com/nba/news
1008 /steve-kerr-suggests-steph-curry-might-be-more-popular-with-kids-than-michael-jordan/
1009 Park, C. W., MacInnis, D. J., Priester, P., Eisingerich, A. B., & Iacobucci, D. (2010). Brand
1010 attachment and brand attitude strength: Conceptual and empirical differentiation of two
1011 critical brand equity drivers. Journal of Marketing, 74(6), pp. 1-17.
1012 Parmentier, M. (2011). When David Met Victoria: Forging a strong family brand. Family
1013 Business Review, 24(3) 217-232.
1014 Parmentier, M., & Fischer, E. (2012). How athletes build their brands’, International Journal
1015 of Sport Management and Marketing, 11(1-2), 106-124.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 40
1016 Parmentier, M, Fischer, E., & Reuber, A. R. (2013). Positioning person brands in established
1017 organizational fields. Journal of the Academy of Marketing Sciences, 41(3), 373-387.
1018 Pegoraro, A. (2010). Look who’s talking – Athletes on Twitter: A case study. International
1019 Journal of Sport Communication, 3(4), 501-514
1020 Pegoraro, A., & Jinnah, N. (20120). Sponsorship: Research paper Tweet ’em and reap ’em:
1021 The impact of professional athletes’ use of Twitter on current and potential sponsorship
1022 opportunities. Journal of Brand Strategy, 1(1), 85-97.
1023 Ratten, V. (2015). Athletes as entrepreneurs: the role of social capital and leadership ability.
1024 International Journal of Entrepreneurship and Small Businesses, 25(4), 442-455.
1025 Ross, S., Russell, K. C., & Bang, H. (2008). An empirical assessment of spectator-based
1026 brand equity. Journal of Sport Management, 22(3), 322.
1027 Rovell, D. (2018). Inside Kevin Duran’s growing empire. Retrieved 4 April, 2019, from:
1028 http://www.espn.com/nba/story/_/id/23532594/how-kevin-durant-building-silicon-valley-
1029 empire
1030 Shapiro, S. L., DeShriver, T. D., & Rasher, D. A. (2017). The Beckham effect: examining the
1031 longitudinal impact of a star performer on league marketing, novelty, and scarcity.
1032 European Sport Management Quarterly, 17(5), 610-634.
1033 Spry, A., Pappu, R., & Cornwell, T. B. (2011). Celebrity endorsement, brand credibility and
1034 brand equity. European Journal of Marketing,45(6), 882-909.
1035 Stevens, S., & Rosenberger, P. J. (2012). The influence of fan involvement, following sport
1036 and fan identification on fan loyalty: An Australian perspective. International Journal of
1037 Sports Marketing & Sponsorship, 13(3), 220-234.
1038 Summers, J., & Johnson Morgan, M. (2008). More than just the media: Considering the role
1039 of public relations in the creation of sporting celebrity and the management of fan
1040 expectations. Public Relations Review, 34(2), 176-182.
1041 Sumino, M., & Harada, M. (2004). Affective experience of J. League fans: The relationship
1042 between affective experience, team loyalty and intention to attend. Managing Leisure,
1043 9(4), 181-192.
1044 Swaminathan, V., Page, K. L., & Gürhan-Canli, Z. (2007). “My” brand or “our” brand: The
1045 effects of brand relationship dimensions and self-construal on brand evaluations. Journal
1046 of Consumer Research, 34(2), 248-259.
1047 The Guardian (2016). No room for half-measures in Cristiano Ronaldo’s gilded world of
1048 110%. Retrieved, 16th January, 2018, from: https://www.theguardian.com/football
ATHLETE BRANDS AND SELF-BRAND CONNECTION 41
1049 /blog/2016/jul/22/cristiano-ronaldo-hotel-watford-chesterfield
1050 The Guardian (2017). Ronaldo has done so much for Real Madrid – so why do some fans
1051 whistle him? Retrieved, 26th March, 2018, from: https://www.theguardian.com/football
1052 /blog/2017/apr/19/cristiano-ronaldo-real-madrid-fans-whistle-genius
1053 The Guardian (2018). Liverpool’s Mohamed Salah, an Egyptian hero for the 21st century.
1054 Retrieved, 17th June, 2018, from: https://www.theguardian.com/world
1055 /2018/may/06/liverpools-mohamed-salah-an-egyptian-hero-for-the-21st-century
1056 Thomson, M. (2006). Human brands: Investigating antecedents to consumers’ attachment to
1057 celebrities. Journal of Marketing, 70(3), 104-119.
1058 Till, B. D., & Busler. M. (2000) The match-up hypothesis: Physical attractiveness, expertise,
1059 and the role of fit on brand attitude, purchase intent and brand beliefs. Journal of
1060 Advertising, 29(3), 1-13.
1061 Tsiotsou, R. H. (2013). Sport team loyalty: integrating relationship marketing and a hierarchy
1062 of effects. Journal of Services Marketing, 27(6), 458–471.
1063 Uggla, H. (2006). The corporate brand association base: A conceptual model for the creation
1064 of inclusive brand architecture. European Journal of Marketing, 40(7/8), 785-802.
1065 Uhrich, S., & Benkenstein, M. (2012). Physical and social atmospheric effects in hedonic
1066 service consumption: customers' roles at sporting events. The Service Industries Journal,
1067 32(11), 1741-1757.
1068 Väätäinen, M., & Dickenson, P. (2018). (Re)examining the effects of athlete brand image
1069 (ABI) on psychological commitment: an empirical investigation using structural equation
1070 modelling (SEM) and fuzzy set qualitative comparative analysis (fsQCA). European Sport
1071 Management Quarterly, doi: 10.1080/16184742.2018.1508242.
1072 Walsh, P., & Williams, A. (2017). To extend or not extend a human brand: An analysis of
1073 perceived fit and attitudes toward athlete brand extensions. Journal of Sport Management,
1074 31(1), 44-60.
1075 Wang, R. T., Zhang, J. J., & Tsuji, Y. (2011). Examining fan motives and loyalty for the
1076 Chinese Professional Baseball League of Taiwan. Sport Management Review, 14(4), 347-
1077 360.
1078 Watkins, B. A. (2014). Revisiting the social identity-brand equity model: An application to
1079 professional sports. Journal of Sport Management, 28(4), 471–480.
1080 Westland, J.C. (2010). Lower bounds on sample size in structural equation modeling.
1081 Electronic Commerce Research and Applications, 9(6), 476-487.
ATHLETE BRANDS AND SELF-BRAND CONNECTION 42
1082 Williams, A., Kim, D. Y., Agyemang, K., & Martin, T. G. (2015). All brands are not created
1083 equal: Understanding the role of athletes in sport-brand architecture. Journal of
1084 Multidisciplinary Research, 7(3), 75-86.
1085 Wolf, E.J., Harrington, K.M., Clark, S.L., & Miller, M.W. (2013). Sample size requirements
1086 for structural equation models: An evaluation of power, bias, and solution propriety.
1087 Educational and Psychological Measurement, 73, 913–934.
1088 Wu, S., Tsai, C.D., & Hung, C. (2012). Toward team or player? How trust, vicarious
1089 achievement motive, and identification affect fan loyalty. Journal of Sport Management,
1090 26(2), 177-191.
1091 Yoo, B., Donthu, N., & Lee, S. (2000). An examination of selected marketing mix elements
1092 and brand equity. Journal of the Academy of Marketing Science, 28(2), 195-211.
1093 Yu, C. (2005). Athlete endorsement in the international sport industry: a case study of David
1094 Beckham. International Journal of Sports Marketing and Sponsorship, 6(3), 189–99.
1095
1096
ATHLETE BRANDS AND SELF-BRAND CONNECTION 43
1100
1101 Figure 1. Hypothesized model.
1102
1103
1104 Figure 2. Summary results of the structural model. n.s. = not significant; * = significant at the .05 level
ATHLETE BRANDS AND SELF-BRAND CONNECTION 44
Skill An athlete’s ability, level of talent Arai et al. (2014) Messi’s talent is often
and competence. Hasaan et al. (2018) highlighted by professional
colleagues and media.
Off-field associations
Physical An athlete’s physical qualities and *Arai et al. (2014) Maria Sharapova and Anna
attraction characteristics that fans find Kournikova were famous for her
aesthetically pleasing.* aesthetic appeal.
Body An athlete’s body fitness in his/her *Arai et al. (2014) Cristiano Ronaldo is known as a
conditioning sport.* machine – he takes care of his
body to always be in the best
conditions to perform.
Lifestyle The athlete’s private life including Hasaan et al. (2018) David Beckham and the public
interests, behaviours, family, or Parmentier (2011) interest in his personal life.
friends.
Personality Visible human qualities of an Carlson & Donovan Stephen Curry being regarded as
athlete that permit fans to identify (2013) relatable to the common person
with and develop a relationship (e.g., a family man)
with the athlete.
Cultural Captures aspects such as race, Hasaan et al. (2018) Pakistani and Indian
background origins and other cultural patterns followership of Cricket players;
of an athlete. Muhamad Ali gauges strong
interest from black community.
Social The athlete’s engagement with Agyemang & Singer Recent athlete protests in NFL,
responsibility social concerns in a given (2013) WNBA, NBA; Philanthropy,
community. such as Serena Williams’
support of women and young
girls.
Role model An athlete’s behaviour off the field Arai et al. (2014) Larry Fitzgerald who is heralded
that society has determined is for his integrity and work ethic
worth emulating. on and off the field.
* = Original definition adopted.
1105
ATHLETE BRANDS AND SELF-BRAND CONNECTION 45
1106
Table 3. Correlation Matrix, Average Variance Extracted (AVE) and squared correlations among constructs.
AVE 1 2 3 4 5 6
1. On-Field image --- 1.00 .219 .076 .033 .011 .097
2. Off-Field image --- .468 1.00 .339 .349 .143 .032
3. Self-Brand connection .655 .275 .582 1.00 .412 .099 .026
4. Athlete commitment .817 .181 .591 .642 1.00 .238 .012
5. Team commitment .833 .106 .378 .315 .488 1.00 .004
6. Athlete sponsor preference .899 .312 .178 .161 .109 .065 1.00
Notes. Values below the diagonal are correlation estimates. Values above the diagonal are squared correlation
1107 estimates.
1108
1109
ATHLETE BRANDS AND SELF-BRAND CONNECTION 46
Table 4. Path Coefficients, Indicator Weights, and Explained Variance of Mediated Structural Model.
H Path Supported? β T LLCI ULCI
H1a On-field image → Athlete commitment No -.125 1.428 -.042 .467
H1b Off-field image → Athlete commitment Yes .313* 3.446 .145 .478
H2a On-field image → Team commitment No -.075 0.756 -.278 .180
H2b Off-field image → Team commitment Yes .326* 3.327 .127 .527
H3a On-field image → Athlete sponsor preference Yes .251* 2.100 .006 .467
H3b Off-field image → Athlete sponsor preference No -.008 0.075 -.178 .217
H4a On-field image → SBC No .006 0.042 -.125 .185
H4b Off-field image → SBC Yes .602* 9.178 .436 .707
H5a SBC → Athlete commitment Yes .490* 6.519 .421 .672
H5b SBC → Team commitment No .097 1.052 -.046 .292
H5c SBC → Athlete sponsor preference No .194 1.877 -.113 .267
H6a On-field image → SBC →Athlete commitment No .003 0.042 -.153 .065
H6b On-field image → SBC →Team commitment No .006 0.037 -.031 .020
H6c On-field image → SBC → Athlete sponsor preference No .012 0.026 -.025 .021
H7a Off-field image → SBC →Athlete commitment Yes .295* 5.200 .204 .425
H7b Off-field image → SBC →Team commitment No .058 1.038 -.046 .168
H7c Off-field image → SBC → Athlete sponsor preference No .117 1.783 .007 .258
Explained Variance
Self-brand connection R2 = .392
Athlete commitment R2 = .462
Team commitment R2 = .119
Athlete sponsor preference R2 = .113
Notes. β = beta weight; t = t-value; LLCI = lower level confidence interval; ULCI = upper level confidence
interval; SBC = Self-brand connection; * = significant at .05 level
1110