In the 1968 Olympics, two African-American athletes, Tommie Smith and John Carlos, stood proudly on top of the podium to claim their medals. As the national anthem of USA began to play, both athletes raised a black-gloved clenched fist in a momentous stand for equality and human rights that became known as the "Black Power" salute. The importance of this historical act was enhanced by the notable venue and the ability to reach such a large audience. In their quest to advance social justice on arguably the world's biggest stage, Smith and Carlos embodied two distinct identities: the activist and the athlete.

Activist identity is defined as an individual's developed, relatively stable, yet changeable orientation to engage in various collective, social-political, problem-solving behaviors spanning a range from low-risk, passive, and institutionalized acts to high-risk, active, and unconventional behaviors (Corning & Myers, 2002). Activist identity can develop differently across individuals. "Lifelong activists" who grow up being involved in social-political or community activism, can be distinguished from "circumstantial activists," who engage in activism due to a specific life event or experience (Ollis, 2011). There is evidence that individuals with higher cognitive abilities, who come from more cohesive families, and who have been previously involved in activism are more likely to engage in social or political activism (Pancer, Pratt, Hunsberger, & Alisat, 2007; Rosenthal, Feiring, & Lewis, 1998).

Athletic identity is defined as the degree to which an individual identifies with the athlete role (Brewer, Van Raalte, & Linder, 1993). There is evidence that the strength of athletic identity is associated with personal and social factors. Personal factors include rhythm of daily life (Stephan & Brewer, 2007), bodily dimension (Stephan & Brewer, 2007), and showing talent at a young age (Webb, Nasco, Riley, & Headrick, 1998). Social factors include socioprofessional responsibility, sports staff, peers and teammates, and social recognition (Stephan & Brewer, 2007). Research has shown that although a strong athletic identity may provide certain benefits, such as greater enjoyment of sport, stronger commitment to sport, and an expanded social network (Horton & Mack, 2000), overadherence to the athlete role can be detrimental. Individuals with a strong athletic identity have been shown to be more likely to have depressive reactions to injury (Brewer, 1993), premature return to sport after injury (Podlog et al., 2013), extreme distress, loss, and confusion of identity upon retirement from sport (Warriner & Lavallee, 2008; Webb et al., 1998), and lower career maturity or readiness for post-sport careers (Danish, Petitpas, & Hale, 1993; Murphy, Petitpas, & Brewer, 1996).

Due to the societal idolization of athletes and their high visibility, athletes possess the ability to aifect many people by shining a spotlight on areas of inequality and injustice. Both scientific and popular media literature have made the case that sport and politics are undeniably linked (Carrington, 2010; Houlihan, 1997, 2000; Jackson & Haigh, 2008), and encouraged those involved in the development of athletes to promote the "athlete-as-citizen" model throughout sport (Fisher, Butryn, & Roper, 2003; Sage, 1993). Recently, there has been an influx of athletes advocating for causes such as anti-racism and LGBT rights (Pierce, 2015). Pierce suggested that a possible reason for this increase is due to our increasingly technological world.

We are living with a generation of athletes who have grown up as the children of social media, of Facebook and Twitter, Instagram and Snapchat and whatever-the-hell-comes-after. They communicate directly with the world and directly with each other. Informal networks have sprouted up within the leagues and within teams. Events are discussed. Opinions are exchanged. Movements are formed. (Pierce, 2015).

Athletes who have used their sport to engage in activism have indicated that four values embedded in sport (i.e., social consciousness, meritocracy, responsible citizenship, and interdependency) are in alignment with an activist orientation (Kaufman & Wolff, 2010). Given these shared dimensions, researchers have argued that the sporting realm is a logical and needed setting for the advocacy of social justice (Agyemang, Singer, & DeLorme, 2010; Kaufman & Wolff, 2010; Mehlsak, Tannenwald, & Guillory, 2009; Travers, 2013).

Societal influences can also discourage athletes from taking public stances on political or social issues. Henderson (2009) suggested that there is a common existing schema that athletes should just play and not pontificate; that advocacy is radical and undesirable. Athletes who engage in activism are sometimes treated as deviants and face booing crowds, fines, suspensions, bans, and loss of endorsements or sponsors (Kaufman, 2008). The type of activism an athlete may engage in also matters. Athletes involved in controversial anti-war activism have been perceived as less trustworthy than those involved in noncontroversial anti-obesity activism (Cunningham & Regan, Jr., 2011). Anti-activism backlashes send a clear message to athletes about what is acceptable behavior and appropriate to do as an athlete and what is off-limits. Due to the possible detrimental consequences experienced by athletes who engage in activist behaviors, athletes may be discouraged from assuming an activist role (Candaele & Dreier, 2004; Smith, Bundon, & Best, 2016).

Few studies have been conducted to examine how athletes and non-athletes differ in activist activity or engagement. Gayles, Rockenbach, and Davis (2012) examined national survey data from college students and found that student-athletes did not differ from their nonathlete peers in their goals for social activism. When it came to actually acting on those goals, however, differences existed. Nonathletes reported engaging in more charitable behaviors than low- and high-profile athletes, with high-profile athletes reporting the lowest level of engagement in charitable behaviors. The authors concluded that possible reasons for the discrepancy between the student-athletes' goals and their actions include a constraint on time or lack of opportunities for athletes to carry out their social activism goals when they are involved in high-profile sport. More recently, Hoffman, Kihl, and Browning (2015) found that college student-athletes reported engaging in service activities (e.g., volunteering at a soup kitchen, reading aloud to school children) more frequently and political activities (e.g., signing a petition, taking part in a demonstration) less frequently than their nonathlete peers. Hoffman et al. suggested that the discrepant findings for service and political activities may be attributed to student-athletes' participation in the NCAA CHAMPS/Life Skills program, an initiative intended to facilitate and support student-athlete development that has a strong emphasis on volunteerism, but not broader political engagement.

The Gayles et al. (2012) and Hoffman et al. (2015) studies provided important information, but the surveys used in the studies were broad and did not focus specifically on athletic and activist identities. For example, Gayles et al. categorized participants as nonathletes, low-profile athletes, high-profile athletes, or both low- and high-profile athletes. They did not take into account the participants' level of athletic identity, a factor that is associated with a wide variety of athlete behaviors (Brewer & Cornelius, 2001). Considering identity is important, as it offers a means of understanding why athletes do or do not engage in activism. It is possible, for example, that individuals strongly and exclusively identified with the athlete role are less likely than those less identified with the athlete role to view themselves as activists and engage in activism.

The current study aimed to address the limitations of previous research by Gayles et al. (2012) and Hoffman et al. (2015) by using a multi-trait, multi-method analysis and specifically assessing activist and athletic identities. Self-report measures and a Facebook content analysis were used to examine the associations between activist and athletic identities in college students and to explore the differences between athletes and nonathletes regarding these identities. Facebook content was used to augment self-report measures of activist and athletic identities because social media offer a means of displaying one's identity to the world (Bundon, 2016) and Facebook profiles have been shown to be quite accurate and representative of offline identities (Gosling, Gaddis, & Vazire, 2007; Waggoner, Smith, & Collins, 2009; Weisbuch, Ivevic, & Ambady, 2009). Further, college age youth have been found to follow political campaigns, post comments to foster political dialogue and civic engagement (Fernandes, Giurcanu, Bowers, & Neely, 2010), and express their opinions (Hoffman et al., 2015) on Facebook. Thus, research provides evidence that college students' Facebook profiles contain identity- and activism-related information and are likely to accurately reflect who they are offline.

College students were considered an appropriate population for the current study for several reasons. First, examining activism in association with the sport involvement of college students facilitates comparison with previous research on the topic (Gayles et al., 2012; Hoffman et al., 2015). Second, the college years overlap with the developmental periods of late adolescence and "emerging adulthood," during which individuals wrestle with questions of who they are and what they will become (Arnett, 2004, 2007; Chickering, 1969; Erikson, 1963). Patterns of self-identification during the formative college years have important implications for activism in adulthood. Third, from a methodological standpoint, athlete and nonathlete college students share a common living and learning environment, representing a degree of control not typically found among adult populations.

In light of research suggesting that strong or exclusive investment in one aspect of identity may result in other aspects of identity being less developed (Brewer et al., 1993; Danish et al., 1993; Murphy et al., 1996, Pearson & Petitpas, 1990), it was hypothesized that the self-report and Facebook indices of athletic identity would be negatively correlated with the comparable measures of activist identity. Conversely, it was expected that the cross-method correlations (i.e., self-report and Facebook) would be positive within each domain of identity. Based on research showing that athletes rate themselves as having a stronger athletic identity than their nonathlete peers (Brewer & Cornelius, 2001), it was hypothesized that the athlete participants of this study would have higher scores on the self-report and Facebook measures of athletic identity than the nonathlete participants. Because athletes report engaging in fewer political activities (Hoffman et al., 2015) and charitable behaviors (Gayles et al., 2012) than nonathletes, it was expected that nonathletes would have higher self-reported activist identity scores than athletes. Athlete-nonathlete differences were not predicted for the Facebook measure of activist content, however, given that Gayles et al. found that student-athletes did not differ from their nonathlete peers in their goals for social activism, and social media has been suggested to be an avenue where an individual can present their "hoped-for possible selves" (Zhao, Grasmuck, & Martin, 2008).

Method

Participants

The participants were 234 students (145 women and 89 men) enrolled at colleges and universities across the United States. The mean age of the participants was 19.72 (SD = 2.13) years. The majority of the sample identified themselves as Caucasian/White (87%). The remaining participants reported their race/ethnicity as American Indian (1%), Asian (2%), Black or African American (5%), Native Hawaiian or Pacific Islander (1%), or multiracial (4%). Both nonathletes (n = 95) and athletes (n = 139) were included. Instruments

Demographic information was gathered with a brief questionnaire. In one of the items, participants were asked to self-identify as a nonathlete or an athlete. Activist identity was assessed with the Activism Orientation Scale (AOS; Corning & Myers, 2002). The AOS is a 35-item scale that focuses on the extent to which individuals would engage in activist behaviors. Responses to AOS items are given on a scale from 0 (extremely unlikely) to 3 (extremely likely). The AOS has been used in multiple studies and has been shown to be a reliable measure with satisfactory internal consistency, construct validity, criterion-related validity, and discriminant validity (Corning & Meyers, 2002). In the current study, the internal consistency of the AOS was high ([alpha] = .98). The 7-item version of the Athletic Identity Measurement Scale (AIMS; Brewer & Cornelius, 2001) was used to measure athletic identity. Responses to AIMS items are given on a scale from 1 (strongly disagree) to 7 (strongly agree). The AIMS is a 7-item scale that has been shown to be psychometrically sound, with strong internal consistency (Brewer & Cornelius, 2001) and cross-cultural factorial validity (Li & Andersen, 2008; Visek, Hurst, Maxwell, & Watson, 2008). In the current study, the internal consistency of the AIMS was high ([alpha] = .94). In accordance with the recommendations of Downey and King (1998) and Bono, Ried, Kimberlin, and Vogel (2007), missing data on the AOS (for up to 7 items, n = 12 participants) and the AIMS (for up to one item, n = 8 participants) were prorated by replacement with the mean of the completed items on the respective scale.

Procedure

After Institutional Review Board approval was obtained, professional listservs, an alumni network, and social media were used to recruit participants for the study. Students interested in participating followed a link to the online survey on Survey Monkey, viewed and expressed agreement with an informed consent statement, and completed the AOS and the AIMS. Following the method used by Ridout, Campbell, and Ellis (2012) when they examined "alcohol identity" on Facebook, participants could "friend" the study's Facebook page to enable the researchers to view their Facebook content. The survey and the Facebook page for the study were kept live until the content analysis was complete and were then deactivated.

Data Analysis

Pearson correlations were computed to determine if there were significant associations among scores on the AOS, the AIMS, and the Facebook indices of activist and athletic identities. T-tests were used to determine if there were significant differences between the athlete and nonathlete participants on the AOS, AIMS, Facebook activist content, and Facebook athletic content. Effect sizes were interpreted in accordance with the guidelines of Cohen (1992). Content analysis was performed on the qualitative data collected from participants' Facebook pages. Based on coding schemes that have been developed to measure concepts such as narcissism (Mehdizadeh, 2010), depression (Moreno et al., 2011), alcohol identity (Ridout et al., 2012), ethno-racial identity (Grasmuck, Martin, & Zhao, 2009), and overall identity construction (Zhao et al, 2008) using Facebook content, a coding form was devised specifically for the current study. The form lists each of the sections of the Facebook page that were included in the coding process: profile pictures and cover photos, timeline posts, groups, pages liked, quotes, and about me. For each of the categories, the number of items that reflected an activist identity and the number of items that reflected an athletic identity were counted and divided by the total number of items in the category to create activist and athletic identity indices. Items were considered to reflect an activist identity or an athletic identity if they met the criteria listed on the coding form as outlined below.

All profile pictures and cover photos were included in the analysis. Inspired by the autophotographic coding scheme of Dollinger (1996), pictures considered to reflect an athletic identity included those in which: (a) the participant was wearing sport-related clothing, (b) the participant was engaging in sport-related activity, (c) a sport environment or a sporting event was depicted, or (d) teams, an individual athlete, or a sport-related logo was shown. Pictures considered to reflect activist identity included those in which: (a) the participant was wearing clothing promoting a political or social cause, (b) a political or activist figure was featured, (c) a political or social service event was depicted, or (d) a political or social cause insignia was shown. The number of pictures considered athletic and activist in nature were divided by the total number of pictures to create separate athletic and activist identity indices, respectively. For example, if a participant had 50 total pictures, 5 pictures that met the criteria for reflecting athletic identity and 2 pictures that met the criteria for reflecting activist identity, the athletic identity index would be 0.10 and the activist identity index would be 0.04.

When coding timeline posts, all posts on the participants' timeline for the six months prior to the date that they completed the demographic questionnaire, the AIMS, and the AOS were considered. Coding ceased on the date of participants' completion of the self-report measures to ensure that participation in the study had not prompted them to think about their athletic identity, level of political involvement or social advocacy and alter their behavior on Facebook. Posts considered to reflect athletic identity included: (a) mentions of sports figures, (b) mentions of current events in the sport world, (c) mentions of games or competitions, (d) mentions of being an athlete (going to practice, the gym, etc.), and (e) mentions of being involved in sporting events. Posts considered to reflect activist identity included: (a) mentions of a political or activist figure; (b) mentions of current world events; (c) advocating or expressing a position on any political, social, or service cause; (d) attempts to spread awareness of any political, social, or service concern; and (e) mentions of being involved in any political, social, or service events.

All quotes listed were included in the analysis. Quotes considered to reflect athletic identity included those by a sport figure or about being an athlete. Quotes considered to reflect activist identity included those by political/ activist figures or pertaining to any political or social issue.

All Facebook groups that participants were a member of were included in the analysis. All sport groups and groups dedicated to a sport team or athlete were considered to reflect athletic identity. All political groups, groups advocating for any social justice causes, or religious groups were considered to reflect activist identity.

All of the pages that the participants liked were included in the analysis. The pages were clicked on and examined. The criteria for a page to reflect athletic identity included: (a) a page dedicated to a sport team, (b) a page dedicated to an individual athlete, or (c) a page with a sport focus. The criteria for a page to reflect activist identity included: (a) a page dedicated to a political, social justice, or service group; (b) a page for a political, activist, or service figure; or (c) a page dedicated to a religion if the "About Me" section calls for action (i.e. reaching out, spreading the word).

When coding the About Me section, all words were considered their own individual items. It is notable that many participants did not have anything listed in this section. Words considered to reflect athletic identity were those pertaining in some way to sport, whereas those considered to reflect activist identity pertained to a political insignia or to a political, social, or service issue.

The first author coded all of the Facebook pages. A single coder was used because of the dynamic nature of Facebook and the fact that content can be changed at any time. To assess the reliability of the coding scheme for profile pictures and cover photos, two independent coders were trained by the first author and then coded 20 of the same Facebook pages, at the same time, using the coding form. The athletic identity (ICC = 0.94) and activist identity (ICC = 0.73) indices demonstrated strong agreement of consistency in ratings (LeBreton & Senter, 2007). Pilot work was also undertaken to ensure that the Facebook athletic identity index tapped participants' identification as athletes as opposed to their identification as sports fans. A sample of 88 college students (69 women and 19 men) gave access to their Facebook pages for coding purposes after completing the AIMS and the Sport Spectator Identification Scale (SSIS), a 7-item scale for which internal consistency (Cronbach's [alpha] = .91), test-retest reliability (r = .60), and criterion-related validity have been documented (Wann & Branscombe, 1993). Results indicated that the Facebook athletic identity index was correlated more strongly with the AIMS (r = .56) than the SSIS (r = .32) and that when statistically controlling for the association between the SSIS and the Facebook athletic identity index, AIMS scores accounted for 27% of the variance in Facebook athletic identity index scores over and above the 10% accounted for by SSIS scores. Thus, it appears that the Facebook athletic identity index assesses athletic identity to a greater extent than fan identity.

Results

Before reporting the results, it is important to discuss how the Facebook content scores were handled in the analysis. The content analysis of the Facebook pages revealed a lack of consistency in the amount of activist content across the various categories. For example, even though individuals might have belonged to many activist-oriented groups, that group membership was not necessarily accompanied by posting activist content on their timeline or liking activist-oriented pages. The category in which activist content was displayed most consistently was the profile and cover photo category. Consequently, the values for this category (i.e., number of activist-oriented pictures divided by total number of pictures and number of athletic-oriented pictures divided by total number of pictures) were used to represent the Facebook activist and athletic content scores, respectively.

Of the 234 participants who completed the AIMS and the AOS, 114 "friended" the Facebook page for the study. There were no significant differences in age, AIMS, or AOS scores between participants who friend-ed and those who did not (ps > .05). There were also no differences in the distributions of gender, race, ethnicity, or athlete status (ps > .05). Means and standard deviations of AOS, AIMS, Facebook activist content, and Facebook athletic content scores are presented in Table 1. Mean AOS scores for both athletes and nonathletes were significantly lower than the mean value for the scale (M = 37.81, SD = 19.05) obtained in the initial validation study (Corning & Myers, 2002), both ts > 3.20 and ps < .002. Compared to the norms provided for the AIMS (Brewer & Cornelius, 2001), the mean AIMS scores were in the 40th to 45th percentile range for both athletes and nonathletes. In response to feedback obtained during the review process, a power analysis was conducted to determine the effect size detectable with the Facebook sample size (n = 114). Using G*Power 3.1.9.2 (Faul, Erdfelder, Buchner, & Lang, 2009) and an alpha of .05, the current sample would provide an 80% chance to detect a medium effect size for the two-tailed t-test analysis (d = .53) and the correlation analysis (r = +/- .26)

For the sample that completed both the AIMS and AOS (w = 234), the correlation between AIMS and AOS was significant (r = -.18, p = .006) and in the predicted direction. The differences between athletes and non-athletes on AIMS and AOS were examined using independent sample t-tests not assuming equality of variances. Athletes and nonathletes differed significantly on the AIMS, t(152.81) = 12.59, p < .001, but not on the AOS, t(214.34) = 1.24, p = .21.

Analyses that examined relationships among AIMS, AOS, and Facebook content were conducted on the subsample that friended the Facebook page (n =114). AIMS scores were not significantly correlated with Facebook activist content scores, r = .06, p = .51, but were significantly correlated with Facebook athletic content, r = .45, p < .001. AOS scores were not correlated with either Facebook athletic content scores, r = -.17, p = .07, or Facebook activist content scores, r = .06, p = .55. The correlation between Facebook athletic score and Facebook activist scores was not significant, r = 10, p = .31. Results of independent samples t-tests not assuming equality of variances indicated that although athletes had significantly higher Facebook athletic content scores than nonathletes, t(111.94) = 4.43,p < .001, the Facebook activist content scores of athletes and nonathletes were not significantly different, t(87.06) = 0.61, p = .54.

These correlations suggest that when individuals have a high score on a self-report measure of athletic identity, there will be a high level of athletic content on their Facebook page and they will tend to score lower on a self-report measure of the likelihood to engage in activism. Identifying as an athlete may not be wholly compatible with identifying as an activist and vice versa.

Understanding of the activist identity of the participants was enriched by their answers to questions about whether or not they had engaged in political or social activism, and if so, what they had done. Out of the 234 participants, 82 individuals (34% of the athletes and 37% of the nonathletes) reported that they had engaged in activism. Some of the 82 individuals listed multiple actions, resulting in 93 activities. These activities were categorized by type of action and included: raising/donating/collecting money, food, or other resources (37), signing petitions (19), going on service trips (18), initiating or being involved in local community campaigns/volunteer work (10), speaking to or supporting a political candidate (3), and being involved in protests (3).

Discussion

In addressing the issue of athletes advocating for social justice (Agyemang, Singer, & DeLorme, 2010; Kaufman & Wolff, 2010; Mehlsak, Tannenwald, & Guillory, 2009; Travers, 2013) and heeding the call for greater involvement of sport psychology in enabling social missions (Schinke, Stambulova, Lidor, Papaioannu, & Ryba, 2016), the current study extends previous research comparing the activist identities of athletes and nonathletes (Gayles et al., 2012; Hoffman et al., 2015) through use of a multi-trait, multi-method design. Incorporation of measures of both activist and athletic identities facilitated a more fine-grained examination of the association between activism and sport involvement than those of previous investigations. From a methodological standpoint, the current study builds upon the autophotographic identity research of Dollinger (1996) and converges with the work of Smith et al. (2016) in using social media as a means of gaining insight into the activist and athletic identities of sport participants.

The weak negative correlation between the AIMS and the AOS indicates that when college students strongly adhere to either an athletic or activist identity they tend to be less likely identify with the other. This association, which accounts for 3% of the variance and corresponds to a small effect size, supports previous research suggesting that a high athletic identity may restrict development of other aspects of a multifaceted identity (Brewer et al., 1993; Danish et al., 1993; Murphy et al., 1996, Pearson & Petitpas, 1990), that political engagement may not be as encouraged as community service for college athletes (Hoffman et al., 2015), and that the time commitment of being involved in a collegiate sport may interfere with athletes' civic engagement rates (Gayles et al., 2012). The lack of a significant difference in activist content on Facebook between athletes and nonathletes also supports previous research (and stands in contrast with the negative association between AIMS and AOS scores). Social media has been suggested to be an avenue where an individual can present their "hoped-for possible selves" (Zhao, Grasmuck, & Martin, 2008), and Gayles et al. (2012) found that student-athletes did not differ from their nonathlete peers in their goals for social activism. Student-athletes also did not differ significantly from their nonathlete peers on the AOS. The lack of difference (and corresponding small effect size), coupled with the inverse relationship between AIMS and AOS scores, suggests that the likelihood of an individual endorsing an activist orientation is not negatively related to being involved in sport per se, but how strongly athletes identify with the athlete role.

In support of the claim that Facebook profiles are quite accurate and representative of individuals' offline identities (Gosling, Gaddis, & Vazire, 2007; Waggoner, Smith, & Collins, 2009; Weisbuch, Ivevic, & Ambady, 2009), the self-report and Facebook measures of athletic identity were positively correlated and indicative of a medium effect size. The lack of a significant correlation between the self-report and Facebook measures of activism stands in contrast to the correspondence between the two athletic identity measures and may be attributable to the element of risk involved with publicly espousing activist beliefs. Publicly displaying involvement in sport may be far less risky than portraying investment in a controversial topic that can generate a backlash from other Facebook "friends" who do not share the same views. As discussed earlier, previous research findings suggest that it may be more acceptable for athletes to be engaged in service activities (Hoffman et al., 2015) and noncontroversial activism (Cunningham & Regan, Jr., 2011) than engaging in or endorsing more radical activities. Although the majority of individuals with a strong athletic identity may not also have an activist identity, individuals who claim that they would participate in activist behaviors do not consistently represent that on their public depiction of who they are on their Facebook page. Thus, even when individuals do have a self-reported interest in taking a stand regarding social or political issues, they may not display their interest publicly.

Although Fernandes et al. (2010) found that youth use Facebook to engage in political dialogue and civic engagement, the lack of consistency of activist content across Facebook categories in the current study suggests that Facebook may not be an accurate means of assessing activist identity. In contrast to the current study, however, Fernandes et al. examined only Facebook groups with specific goals of supporting presidential candidates rather than entire Facebook profiles. Therefore, the sample of the Fernandes et al. study was already made up of individuals that had some interest, investment, or willingness to engage in political dialogue. In addition, if people posted comments in the group format studied by Fernandes et al., the comments were portrayed only on the group page rather than on the people's personal pages and were likely to be perceived favorably given that the whole purpose of the group was to engage in that discussion.

The findings of this study have implications for the role and expectations of athletes in addressing social justice issues. Given that self-reported adherence to an activist identity was inversely related to that of an athletic identity, college athlete activists likely represent only a small subset of the broader college athlete population. Further, because of the amount of risk and backlash that exists for athletes who choose to engage in activism (Cunningham & Regan, Jr., 2011; Dreier & Candaele, 2004; Henderson, 2009; Kaufman, 2008; Smith et al., 2016), the subset of athletes who do align with an activist identity may choose not to publicly engage in activism out of fear of the consequences or negative perception from the public.

Acknowledging the barriers to athletes adopting the activist role does not diminish the idea that athletes hold an inherent power to affect change or argue that athletes should not be taking on activist roles. Rather, understanding the barriers that exist is an important first step to help further the goal of developing athletes as engaged citizens, if they so choose. The findings suggest that there may be a need to create spaces for athletes who may be interested in engaging in activism to discuss activist issues in safe environments where there is minimal risk of losing tangible or intangible resources. If one agrees with the suggestion in the literature for athletes to use their sport as a vehicle for social justice (Agyemang et al., 2010; Kaufman & Wolff, 2010; Mehlsak et al., 2009; Travers, 2013), it will be important to create opportunities that help foster and encourage that social consciousness rather than place all of the risk and responsibility solely on the athletes to navigate the course of action alone.

It is important to address the strengths and weaknesses of the methodology and the participant sample of the current study. One strength was the use of multiple methods to measure the constructs. Assessing a construct through different means can improve the validity of that construct, particularly when one of the means of measurement is anchored in real world behavior. Additionally, the self-report measures chosen for this study to assess activist and athletic identity were shown to be internally consistent. A methodological weakness of the study, however, was the challenge that accompanied measuring activist on Facebook. Although activists can use social media to help bring awareness to their causes, there were not many individuals in the sample of the current study who used Facebook as a consistent means to engage in activism. This was evident in the results, as activist content on Facebook was not correlated with self-reported activist identity.

Although the broad sample of college students from across the United States was a methodological strength, a larger, more diverse sample would have afforded an analysis of how cross-campus differences in activism and athletic identity might have contributed to the current findings. Further, the fact that not all participants who completed the surveys added the research Facebook page leaves a gap of data that could have further informed the findings. A second limitation of the study is that there were more athletes than nonathletes in the sample. This limitation is likely explained by the purpose of the study, which was to understand if athletes are or are not identifying as activists. The sample would likely have been different if the research question was aimed at understanding the activist population. If the target population had been activists rather than athletes, the Facebook measure of activist content may have ended up being a better predictor of self-reported activist identity than it was in the current study, as there may have been more individuals who publicly portrayed activist identities on their Facebook page. A third limitation associated with the composition of the participant sample is the exclusive focus on college students in the United States. Further research is needed to determine the generalizability of the current findings to athlete populations outside the North American college environment. Cross-cultural inquiry would enable an exploration of the extent to which adoption of an activist identity by athletes is influenced by national cultural factors versus aspects of being an athlete that are common across cultures.

The results of this study provide implications for future research. In light of the low magnitude of the association between the self-report measures of activist and athletic identity, a systematic replication of the study is recommended. Moreover, although this studied analyzed the existing associations between athletic and activist identities, it would helpful to understand how athletes believe the two identities should intersect. In addition to the athletes themselves, how do others in the sport world (e.g., coaches, fans) perceive athlete activists? Given the backlash against athlete activists that has been shown to exist (Cunningham & Regan, Jr., 2011; Henderson, 2009, Kaufman, 2008), it would be interesting to explore what topics currently are "safer" and less controversial than others and if or how what is acceptable has changed over time. Further, as this study specifically examined college students, it would be of interest to assess how activist and athlete identities develop. For example, are there differences between college and professional athletes in adopting an activist identity? Are there different challenges or barriers for college athletes and professional athletes in using their sport as a platform to advocate for social change? To what extent do coaching practices and stereotypes of athletes (e.g., "dumb jocks") contribute to the development of activist identity among athletes? Lastly, for those who value athlete activism, it will be important to understand how best to support the subset of athletes who are interested in engaging in activist behaviors, but are not currently doing so. Qualitative data would be helpful to gain a better understanding of athletes' perceptions of using their platform in sport as a vehicle for social justice issues that they care about or are pertinent to them. Do they experience the barriers suggested by this study or are there different or additional obstacles that they encounter? What do they believe is needed or would be helpful for them to enhance their ability to voice their opinions and use their power for the greater good? The questions posed above are simply suggestions of potential avenues to explore when continuing this line of inquiry. Future research can be conducted to facilitate understanding of the phenomenon of athlete activism and to learn how to best support athletes who choose to engage in activist behaviors.

Conclusion

The key findings from this study are the nuances that are associated with the relationships between athletic and activist identities in the population sampled. The weak negative correlation between the AIMS and the AOS indicates that when college students strongly adhere to either an athletic or activist identity, they tend to be less likely to identify with the other. Athletes did not differ significantly from their nonathlete peers, however, in terms of their AOS scores and activist content on Facebook. The lack athlete-nonathlete differences combined with the inverse relationship between AIMS and AOS scores suggests that the reduced likelihood of an individual endorsing an activist identity may be less related to being involved in sport per se than to identifying with the role of athlete.

Erica G. Beachy and Britton W. Brewer, Springfield College

Judy L. Van Raalte

Springield College and Wuhan Sports Univeristy

and

Allen E. Cornelius

University of the Rockies

Address Correspondence to: Britton W. Brewer, Department of Psychology, Springfield College, Springfield, MA 01109. E-mail: [email protected]

References

Arnett, J. J. (2004). Emerging adulthood: The winding road from the late teens through the twenties. New York: Oxford University Press.

Arnett, J. J. (2007). Emerging adulthood: What is it, and what is it good for? Child Development Perspectives, 1, 68-73.

Agyemang, K., Singer, J. N., & DeLorme, J. (2010). An exploratory study of black male collegiate athletes' perceptions of race and athlete activism. International Review for the Sociology of Sport, 45, 419-435.

Bono, C., Ried, L. D., Kimberlin, C., & Vogel, B. (2007). Missing data on the Center for

Epidemiologic Studies Depression Scale: A comparison of 4 imputation techniques. Research in Social and Administrative Pharmacy, 3, 1-27.

Brewer, B. W. (1993). Self-identity and specific vulnerability to depressed mood. Journal of Personality, 61, 343-364.

Brewer, B. W., & Cornelius, A. E. (2001). Norms and factorial invariance of the Athletic Identity Measurement Scale. Academic Athletic Journal, 15, 103-113.

Brewer, B. W., Van Raalte, J. L., Linder, D. E. (1993). Athletic identity: Hercules' muscles or Achilles heel 1 International Journal of Sport Psychology, 24, 237-254.

Bundon, A. (2016). The web and digital qualitative methods: Researching online and researching the online in sport and exercise studies. In B. Smith & A. C. Sparkes (Eds.), Routledge handbook of qualitative research in sport and exercise (pp. 355-367). London: Routledge.

Candaele, K., & Dreier, P. (2004, June 28). Where are the jocks for justice? The Nation, 278(25), 21-24.

Carrington, B. (2010). Race, sport, and politics: The sporting black diaspora. Thousand Oaks, CA: Sage.

Chickering, A. W. (1969). Education and identity. San Francisco: Jossey-Bass.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.

Corning, A. F., & Myers, D. J. (2002). Individual orientation toward engagement in social action. Political Psychology, 23, 703-729.

Cunningham, G. B., & Regan, Jr., M. R. (2011). Political activism, racial identity, and the commercial endorsement of athletes. International Review for the Sociology of Sport, 47, 657-669.

Danish, S. J., Petitpas, A. J., & Hale, B. D. (1993). Life developmental interventions for athletes: Life skills through sports. The Counseling Psychologist, 21, 352-285.

Dollinger, S. J. (1996). Autophotographic identities of young adults: With special reference to alcohol, athletics, achievement, religion, and work. Journal of Personality Assessment, 67, 384-398.

Downey, R. G., & King, C. (1998). Missing data in Likert ratings: A comparison of replacement methods. Journal of General Psychology, 125, 175-91.

Evans, N. J., Forney, D. S., Guido, F. M., Patton, L. D., & Renn, K. A. (2010). Student development in college; Theory, research, and practice (2nd ed.). San Francisco: Jossey-Bass.

Faul, F., Erdfelder, E., Buchner, A., & Lang, A. (2009). Statistical power analyses using G*Power 31: tests for correlation and regression analyses. Behavior Research Methods, 41, 1149-60.

Fernandes, J., Giurcanu, M., Bowers, K. W., & Neely, J. C. (2010). The writing on the wall: A content analysis of college students' Facebook groups for the 2008 presidential election. Mass Communication and Society, 13, 653-675.

Fisher, L. A., Butryn, T. M., & Roper, E. A. (2003). Diversifying (and politicizing) sport psychology through cultural studies: A promising perspective. The Sport Psychologist, 17, 391-405.

Gayles, J. G., Rockenbach, A. B., & Davis, H. A. (2012). Civic responsibility and the student athlete: Validating a new conceptual model. The Journal of Higher Education, 83, 535-557.

Gosling, S. D., Gaddis, S., & Vazire, S. (2007, March). Personality impressions based on Facebook profiles. Paper presented at the International Conference on Weblogs and Social Media, Boulder, CO.

Grasmuck, S., Martin, J., & Zhao, S. (2009). Ethno-racial identity displays on Facebook. Journal of Computer-Mediated Communication, 12, 158-188.

Henderson, S. (2009). Crossing the line: Sport and the limits of civil rights protest. International Journal of the History of Sport, 26, 101-121.

Hoffman, J., Kihl, L., & Browning, A. (2015). Civic engagement and activism: Do college students and college athletes differ? Journal of College and Character, 16, 75-89.

Horton, R. S., & Mack, D. E. (2000). Athletic identity in marathon runners: Functional focus or dysfunctional commitment? Journal of Sport Behavior, 23, 101-119.

Houlihan, B. (2000). Politics and sport. In J. Coakley & E. Dunning (Eds.), Handbook of sports studies (pp. 213-227). London: Sage.

Jackson, S. J., & Haigh, S. (2008). Between and beyond politics: Sport and foreign policy in a globalizing world. Sport in Society, 11, 349-358.

Kaufman, P. (2008). Boos, bans, and other backlash: The consequences of being an activist athlete. Humanity & Society, 32, 215-237.

Kaufman, P., & Wolff, E. (2010). Playing and protesting: Sport as a vehicle for social change. Journal of Sport & Social Issues, 34, 154-175.

LeBreton, J. M., & Senter, J. L. (2007). Answers to 20 questions about interrater reliability and interrater agreement. Organizational Research Methods, 11, 815-852.

Li, H. Y., & Andersen, M. B. (2008). Athletic identity in China: Examining the AIMS in a Hong Kong sample. International Journal of Sport and Exercise Psychology, 6, 176-188.

Mehdizadeh, S. (2010). Self-presentation 2.0: Narcissism and self-esteem on Facebook. Cyberpsychology, Behavior & Social Networking, 13, 357-364.

Mehlsak, A., Tannenwald, D., & Guillory, B. (2009). Comment: Use sport for social change. NCAA News, 1. Retrieved from http://fs.ncaa.org/Docs/NCAANewsAr chive/2009/Associationwide/comment%2buse%2bsport%2bfo%2bsocial%2bchange%2b-%2bncaa%2bnews%2b03-20-09.html.

Moreno, M. A., Jelenchick, L. A., Egan, K. G., Cox, E., Young, H., Gannon, K. E., &

Becker, T. (2011). Feeling bad on Facebook: Depression disclosures by college students on a social networking site. Depression and Anxiety, 28, 447-455.

Murphy, G. M., Petitpas, A. J., & Brewer, B. W. (1996). Identity foreclosure, athletic identity, and career maturity in intercollegiate athletes. The Sport Psychologist, 10, 239-246.

Ollis, T. (2011). Learning in social action: The informal and social learning dimensions of circumstantial and lifelong activists. Australian Journal of Adult Learning, 51, 248-268.

Pancer, S., Pratt, M., Hunsberger, B., & Alisat, S. (2007). Community and political involvement in adolescence: What distinguishes the activists from the uninvolved? Journal of Community Psychology, 35, 741-759.

Pearson, R., & Petitpas, A. (1990). Transitions of athletes: Developmental and preventive perspectives. Journal of Counseling and Development, 69, 7-10.

Pierce, C. P. (2015, June 29). NASCAR's confederate flag ban and the explosion of athlete activism. Grantland. Retrieved from http://grantland.com/the-triangle/nascar-confederate-flag-ban-sonoma-lebron-james-racing/

Podlog, L. Gao, Z., Kenow, L., Kleinert, J., Granquist, M., Newton, M., & Hannon, J. (2013). Injury rehabilitation overadherence: Preliminary scale validation and relationships with athletic identity and self-presentation concerns. Journal of Athletic Training, 48, 372-381.

Ridout, B., Campbell, A., & Ellis, L. (2012). 'Off your Face(book)': Alcohol in online social identity construction and its relation to problem drinking in university students. Drug and Alcohol Review, 31, 20-26.

Rosenthal, S., Feiring, C., & Lewis, M. (1998). Political volunteering from late adolescence to young adulthood: Patterns and predictors. Journal of Social Issues, 54, 477-493.

Sage, G. H. (1993). Sport and physical education in the new world order: Dare we be agents of social change? Quest, 45, 151-164.

Schinke, R. J., N. Stambulova, N., R. Lidor, R., A.G. Papaioannou, A. G., & Ryba, T. V. (2016).

International Society of Sport Psychology position stand: Sport as social missions. International Journal of Sport and Exercise Psychology, 14, 4-22.

Smith, B., Bundon, A., & Best, M. (2016). Disability sport and activist identities: A qualitative study of narratives of activism among elite athletes' with impairment. Psychology of Sport and Exercise, 26, 139-148.

Stephan, Y., & Brewer, B. W. (2007). Perceived determinants of identification with the athlete role among elite competitors. Journal of Applied Sport Psychology, 19, 67-79.

Travers, A. (2013). Transformative sporting visions. Journal of Sport & Social Issues, 37, 3-7.

Visek, A. J., Hurst, J. R., Maxwell, J. P., & Watson, J. C. (2008). A cross-cultural psychometric evaluation of the Athletic Identity Measurement Scale. Journal of Applied Sport Psychology, 20, 473-480.

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

Waggoner, A. S., Smith, E. R., & Collins, E. C. (2009). Person perception by active verses passive perceivers. Journal of Experimental Social Psychology, 45, 1028-1031.

Warriner, K., & Lavallee, D. (2008). The retirement experiences of elite female gymnasts: Self identity and the physical self. Journal of Applied Sport Psychology, 20, 301-317.

Webb, W. M., Nasco, S. A., Riley, S., & Headrick, B. (1998). Athlete identity and reactions to retirement from sports. Journal of Sport Behavior, 21, 338-362.

Weisbuch, M., Ivevic, Z., & Ambady, N. (2009). On being liked on the web and in the 'real world': Consistency in first impressions across personal webpages and spontaneous behavior. Journal of Experimental Social Psychology, 45, 573-576.

Zhao, S., Grasmuck, S., & Martin, J. (2008). Identity construction on Facebook: Digital empowerment in anchored relationships. Computers in Human Behavior, 24, 1816-1836.

Table 1
Descriptive Statistics for Athletes and Non-athletes

                                Athletes

Variable                   M        SD       n

AIMS *                   37.18      6.80    139
AOS                      25.44     17.85    139
FB athletic content *    0.28       0.19     65
FB activist content      0.01       0.02     65

                              Nonathletes

Variable                   M        SD       n      Cohen's d

AIMS*                    22.08     9.79      95       1.79
AOS                      30.65    18.61      95       0.28
FB athletic content*      0.12     0.14      49       0.96
FB activist content       0.01     0.03      49       0.08

* Athlete mean significantly different from non-athlete mean, p < .05
AIMS = Athletic Identity Measurement Scale; AOS = Activism Orientation
Scale. FB = Facebook.