Revisiting the Relationship Between Sports Fandom and the Black Criminal Stereotype: A Replication and Extension Study

Abstract

In news media, one stereotype that is continually over-represented is Black men as criminals, leading to what many refer to as “the Black criminal stereotype.” Although research demonstrates that distorted news portrayals of criminals can provoke stereotypical responses in viewers, limited evidence connects these effects to sport media. Anderson and Raney (2018) explored this in an experimental study (n = 234) and found evidence of the Black criminal stereotype among sports fans. However, more research was needed to further explore this phenomenon. The current study employed a similar experimental design (n = 603) in an attempt to (1) replicate their 2018 study to examine whether sports fans’ perceptions of criminal athletes have changed over the past several years, especially in light of the Black Lives Matter movement, and (2) determine whether the sport played by an alleged criminal athlete might be linked to judgments regarding race and crime in sports. Our findings indicate that sports fandom still predicts stereotypical judgments. However, individual difference variables—particularly social conservatism, African American stereotype endorsement, and gender—were more strongly related. Our findings point toward the potential power of social movements in influencing attitudes and beliefs regarding race and crime.

Keywords

Black criminal stereotype Black Lives Matter #metoo sports fandom media

Decades of research demonstrates that media portrayals play a significant role in the development of perceptions about racial and ethnic groups (for a recent summary, see Dixon, 2020). Studies repeatedly find that portrayals of crime distort perceptions of Black criminality in the United States, as news media overrepresent Black people as perpetrators of crime and overrepresent violent Black criminality in general (e.g., Dixon, 2006; Dixon & Azocar, 2007; Dixon & Maddox, 2005; Leshner, 2006). Such exposure can activate the so-called “Black criminal stereotype” (Dixon, 2008a), which can subsequently influence attitudes and behavioral tendencies toward African Americans. Despite numerous studies examining these issues in general crime news coverage, comparatively little attention has been paid to the activation of the Black criminal stereotype in a sports news context (for notable exceptions, see Anderson & Raney, 2018; Brown et al., 2021). In the wake of league, team, and athlete responses to the Black Lives Matter (BLM) movement, an examination—or in this case, a reexamination—of sports news coverage of athlete criminality was warranted. To that end, we replicated our previous study (Anderson & Raney, 2018) to explore racial perceptions and attitudes toward Black and White American football players accused of violent and nonviolent crimes. The current study also extends this line of research to a different sporting context by exploring audience responses toward Black and White soccer players accused of the same crimes.

Study Context

On September 1, 2016, San Francisco 49ers quarterback Colin Kaepernick kneeled during the national anthem of the team’s final National Football League (NFL) preseason game to call attention to police brutality against and systemic oppression of African Americans in the United States. Despite increasing media attention to the burgeoning BLM movement at the time, Kaepernick was widely criticized across the NFL and even called a “traitor” by NFL executives, who vowed he would never play for their organization (Niven, 2021). The following season, when Kaepernick’s contract expired, no NFL team signed him, and he has not played in the league since.

Following the killing of George Floyd by police officer Derek Chauvin on May 25, 2020, the NFL could no longer ignore the reasons behind Kaepernick’s protests. As TePoel and Nauright (2021) observed, “suddenly and seemingly overnight, individual voices, groups, organizations and businesses from all sectors of society cried out that ‘Black Lives Matter’ … in an about-face that turned swiftly from prior stances of silence or opposition” (p. 693). In June 2020, NFL Commissioner Roger Goodell released a video stating that the league was wrong for previously silencing players’ peaceful protests against police brutality, and on August 23, 2020, Goodell publicly apologized to Kaepernick himself. In an interview on former NFL linebacker Emmanuel Acho’s YouTube show, Uncomfortable Conversations with a Black Man, Goodell said, “I wish we had listened earlier, Kaep, to what you were kneeling about and what you were trying to bring attention to” (Boren, 2020, para. 2). The NFL subsequently pledged to donate $250 million over a 10-year period to combat systemic racism in the United States, including leveraging its media properties (e.g., NFL Films, NFL Network) to increase awareness of social justice issues (Reid, 2020) and using their facilities and resources to fight racism and promote social justice reform (Assimakopoulos, 2020). As part of this effort, the league also vowed to imprint “End Racism” and “It Takes All of Us” in the end zones at each stadium, while also allowing similar phrases to be worn on players’ helmets; many of these messages are readily visible during televised NFL games.

Although the NFL—and most other professional sports leagues and team owners— generally resisted or ignored the BLM movement until 2020 (TePoel & Nauright, 2021), the National Basketball Association (NBA) and its players embraced the BLM movement early on and have arguably been the most vocal in offering support. For instance, less than one month after a hooded sweatshirt-wearing Trayvon Martin was fatally shot by George Zimmerman in February 2012, NBA star LeBron James posted a photo on Twitter of him and numerous Miami Heat teammates also wearing hoodies, with heads bowed. Afterwards, Heat coach Eric Spoelstra expressed support for his players: “For them to come together, to draw more light on the subject, I think, is a powerful move, and we all stand behind them” (Mayo, 2012, para. 3). Many other instances of support for the BLM Movement in the NBA have appeared, including top players such as Kobe Bryant and Derrick Rose wearing “I Can’t Breathe” t-shirts during pregame warmups to honor Eric Garner’s last words before dying in a chokehold by an arresting police officer (Rohrbach, 2020).

Unquestionably, the murder of George Floyd in 2020, along with several other public killings of people of color by police officers, resulted in riots and protests across the United States and led to people of all races and backgrounds protesting police brutality and racial injustice. During this “awakening,” professional sports leagues in the United States could no longer remain silent. As Nicolaides and Lim (2020) suggested, “the global pandemic, accompanied by stay-at-home curfews, forced many people to slow down and reflect on the civil unrest in America in the wake of the murder of George Floyd, Breonna Taylor, and Ahmaud Arbery” (p. 844). Sports organizations and league executives were among those demanding change, while also admitting fault for their previous ignorance towards racial discrimination and the need for criminal justice reform. Unsurprisingly, the NBA remained at the forefront of this movement, adding BLM signage to their courts. Major League Baseball (MLB), the National Hockey league (NHL), and Major League Soccer (MLS) also offered support through a mix of fundraising efforts, social media campaigns, and other messages of unity. That same year, after the murder of Jacob Blake, Jr by a White police officer, the sports world’s response was visible. Players on the Milwaukee Bucks NBA team, located just 40 miles from the shooting, boycotted their playoff game against the Orlando Magic. Teams in MLB, MLS, and the WBNA also canceled or postponed games in protest, and tennis star Naomi Osaka sat out of her semifinal match at the Western Southern Open, explaining, “Before I am an athlete, I am a Black woman … and as a Black woman I feel as though there are much more important matters at hand” (Li & Stelloh, 2020, para. 3).

Because of these and numerous other public actions by sport leagues, teams, and players—as well as extensive media coverage accompanying them—issues of race and racial inequality have played an increasingly prominent role in the reporting and coverage of U.S. sports over the past several years. In light of these developments, it seems reasonable to question whether this increased presence of race in sports conversations might have impacted more broadly held race-related attitudes and beliefs within the sports context, particularly among sports fans. To examine one piece of this much-larger puzzle, we replicated a study conducted prior to the increased salience of race in sports coverage, exploring possible shifts in race-related perceptions of athlete criminality.

Media Stereotyping

Research in communication and psychology points to mass media as a significant source for the production and reinforcement of racial prejudice, primarily through stereotyped content (Dixon, 2020). Media stereotyping research generally focuses on two themes: how media play a role in stereotype formation and how media may trigger existing stereotypes (Oliver et al., 2014); the current study examined the latter. Examinations of media-triggered stereotypes generally rely upon priming theory and network models of memory (e.g., Ewoldsen & Rhodes, 2020) as explanatory mechanisms for the observed effects; the current study did as well. Priming “refers to the process by which recently activated information about a group (e.g., stereotypes) is used in making subsequent judgments of group-related stimuli” (Dixon & Maddox, 2005, p. 1556). These judgments may occur regardless of one’s personal beliefs, as stereotypes and beliefs are conceptually distinct cognitive structures (e.g., Devine, 1989)

As distinct cognitive structures (or schema), stereotypes affect the encoding and processing of information by directing attention to certain stimuli over others and leading individuals to associate (often negative) personal attributes to all members of a specific group, regardless of variation on those attributes within the group. When activated through media exposure, stereotypes can influence judgments of and behavior towards outgroup members, especially when individuals lack firsthand experience with that outgroup. As Ash and Cranmer (2020) explain, “given that the vast amount of knowledge individuals acquire about the social world is mediated, much of their schema about social groups is informed by portrayals in media, particularly when interpersonal contact with those groups is limited” (p. 393).

One particular stereotype continually perpetuated through news media is that of Black individuals as criminals. For instance, Blacks are overrepresented as perpetrators of crime in news stories, proportionally when compared to Whites and in comparison to real-world crime statistics (e.g., Campbell, 1995; Dixon & Linz, 2000). Further, news media routinely overrepresent violent Black criminality (Dixon, 2008a; Leshner, 2006). Black suspects are consistently portrayed as violent, aggressive, and dangerous, and exposure to such messaging can activate what scholars have termed “the Black criminal stereotype” (e.g., Dixon, 2008a), further strengthening cognitive associations between Blacks and violent criminality in audience members and reinforcing prejudice and hostility toward Blacks (Bjornstrom et al., 2010).

Although the Black criminal stereotype is often activated through exposure to media coverage of Black criminality, for the purposes of this study, it is important to note that it may also be activated when an audience member is exposed to any crime story. Research shows that, when presented with a story of a race-unidentified suspect, individuals are more likely to infer that the suspect is Black than any other race (e.g., Dixon, 2007). Because viewers often associate crime with Blackness, they tend to react similarly to both Black criminal suspects and race-unidentified criminal suspects regarding attitudes and judgments of the perpetrator (Dixon, 2008b). Thus, stereotypes not only influence categorization of information, but they also lead us to “fill-in” missing information and influence memory.

Sport Media Stereotyping

Racial stereotyping in sports media—both in general and in relation to crime—has been extensively examined in the communication literature. Numerous critical content analyses have examined broader media portrayals of race in sport, finding that White and Black athletes are portrayed differently across the media landscape (e.g., Billings, 2003, 2004; Billings & Eastman, 2002; Denham et al., 2002). Such portrayals contribute to negative stereotypical portrayals of Blackness, where White athletes are often described by their intelligence, strong work ethic, and leadership skills, whereas Black athletes are often described by their natural physical ability and strength (e.g., Ferrucci et al., 2013). As Grainger et al. (2006) explained, stereotyping of natural Black athleticism “suggests that African Americans possess innate physiological advantages while conversely lacking the necessary skills and intelligence to succeed in other occupational areas” (p. 452). Stereotypical portrayals of race in sport are more pervasive among “intellectually challenging” positions, such as quarterback in American football (e.g., Ferrucci & Tandoc, 2017, 2018). Whereas the success of a White quarterback is commonly attributed to effort rather than physical ability, Black athletes are often stereotyped as lacking the intelligence needed to succeed at the position. Arth and Billings (2019) described this phenomenon as positional “stacking,” where certain positions are segregated based on race. Such portrayals have become so pervasive that they facilitate both general public stereotypical attitudes, as well as within-group stereotype endorsement (e.g., Ferrucci & Tandoc, 2017).

Stereotypical portrayals of criminality in sports media coverage have also garnered attention from scholars. Similar to patterns in general news, Black athletes are routinely overrepresented as criminals in sports news compared to their White counterparts, despite the lack of data supporting the notion that Black athletes commit more crimes (Berry & Smith, 2000; Mastro et al., 2011). Moreover, media representations of Black criminal athletes commonly reinforce general portrayals of African American males as violent, deviant, and threatening (Grainger et al., 2006), further reinforcing the stereotype of violent Black criminality. Because media frames play an important role in audience perceptions of issues (for a complete summary, see Tewksbury & Scheufele, 2020), such “racist rhetoric” can shape individuals’ perceptions of both people of color and broader social issues related to race and crime. Therefore, stereotypical media representations of Black athletes and their off-field transgressions may contribute to negative portrayals of Blackness and wider racialized discourses of criminalizing Black men in society (Leonard, 2006).

Despite this rich tradition of research, few experimental studies have directly examined the effects of Black criminal stereotyping among consumers of sports news. We earlier explored this phenomenon (Anderson & Raney, 2018), finding that the Black criminal stereotype can indeed be activated with sports news. Sports fans were more likely than nonfans to infer that a racially unidentified criminal athlete was Black as opposed to White. However, the more a participant claimed to be a sports fan—and, moreover, the more they reported being a fan of combative sports like American football—the more favorable they viewed an athlete charged with a violent crime. Brown et al. (2021) further explored this phenomenon, examining how athlete race and image-restoration strategies affected perceptions of athletes accused of criminal activity among sports news consumers. Similar to our 2018 findings, they found that high sports news consumers perceived Black criminal athletes more positively than White athletes, providing evidence of a “racial contradiction” in sports. That is, “because sports media provide an unrealistic and superficial portrayal of Black criminal athletes, media consumers could possibly either be surprised and disappointed that a White athlete engaged in criminal activity” (p. 6).

The increased attention to race-related issues and information in U.S. sports coverage in the wake of the BLM movement led us to question whether more general, race-related attitudes among sports fans may have shifted and, if so, in what direction. Specifically, we were curious if the ways that sports fans perceive athletes arrested for a crime have changed. It seems possible that the widespread support for BLM from athletes, teams, and leagues may have created greater awareness of and sensitivity to racial inequalities among sports fans, which could encourage less stereotypical judgments regarding race (and crime) in sport. It is also possible that this may be the case for only certain fans (e.g., those who are also less socially conservative). Still yet, it is possible that (some) fans dismiss public statements about racial inequality by leagues and athletes as political theater, or “wokeness,” leading to a strengthening of racial-stereotypical thinking. To explore these possibilities, we replicated our previous study (Anderson & Raney, 2018) to see if attitudes related to Black athlete criminality have shifted, especially among sports fans. In doing so, we interrogated the same research questions explored in the prior study:

RQ1

Is the Black criminal stereotype (still) more prevalent among sports fans versus nonfans?

RQ2

Do sports fans (still) differ from nonfans in basic attitudes and beliefs about crime and violence in sport?

In addition to a strict replication of the previous study, the current project included other factors that might influence sports fans’ attitudes and judgments related to race and crime in sport. First, in the initial experiment, we systematically manipulated the crime committed by the athlete (i.e., marijuana distribution vs. domestic abuse) but held the sport played constant (i.e., football). However, because African Americans are traditionally “over-represented” (relative to their percentage in the overall population) as participants in some sports (e.g., football, basketball) and “underrepresented” in other sports (e.g., soccer, hockey; see Rasmussen et al., 2005), certain sports themselves—at least anecdotally—are stereotyped as (more) “Black” or “White” (at least within the United States). Because of this, social reality judgments regarding race and crime may differ across athletes depending upon the sport they play. Thus, in the current study, both the crime committed and the sport played by the athlete (football vs. soccer, with the latter presumed to represent a more “White sport”) were systematically manipulated.

Secondly, previous research on racial stereotyping outside the sporting context (e.g., Dixon, 2006, 2008a, 2008b) has suggested that individual difference variables may be linked to social judgments regarding race and crime. It stands to reason, then, that such variables may also be influential in the sports context. Of particular interest were potential differences related to stereotype endorsement, racial prejudice, political ideology, and social conservatism, all of which have been explored in similar research in the past (see Dixon, 2020). We chose to investigate these relationships through additional research questions:

RQ3

To what extent does the sport played by an athlete affect judgments of an athlete-suspect of violent and nonviolent crimes?

RQ4

To what extent do demographics, sports fanship, and ideological dispositions influence judgments of an athlete-suspect of violent and nonviolent crimes?

Methodology

To explore these questions, we utilized a 3 (prime: Black vs. White vs. racially unidentified, non-athlete criminal) x 2 (target athlete crime: marijuana distribution/nonviolent crime vs. domestic violence/violent crime) x 2 (target athlete sport: football vs. soccer) online experimental design. A total of 516 undergraduate students enrolled in communication courses at a state-supported, research university in the southern United States and 87 undergraduate students enrolled in communication courses at a private university in the northeastern United States participated in the study between November 2020 and April 2021, in exchange for course or extra credit. Based on previous research, no reason could be determined to examine the subsamples from each site separately. Nevertheless, an ANOVA procedure was conducted to explore potential differences between the sites on the four primary outcome variables (i.e., suspect perception, emotional reactions, guilt and recidivism, proposed sentencing; see details below) across all conditions; as expected, no differences (p > .1) were observed. As a result, data from both sites were combined for a total of n = 603 participants. A majority of the participants self-identified as women (68.7%) and Caucasian (78.8%), with an average age of 20.50 (range = 18–40) years.

Procedures

The procedures and all study materials for the current study were identical to those from our 2018 study, with one exception. Due to COVID-19 restrictions, data in the replication study were collected online (through Qualtrics), whereas the initial study was lab-based. After providing IRB-approved consent, participants completed several independent measures (see below) before reading a distractor article (word count = 207) about the U.S. economy presented as a screen capture from CBSNews.com. Dates in the article were changed from the original version to reflect 2020 information. Participants completed memory and attitudinal items immediately afterwards, though the data were not analyzed. The distractor article was intended to familiarize participants with the study protocols and to reduce hypothesis guessing.

Participants then read a second article (i.e., “priming article,” word count = 183) describing the arrest of a male college student in conjunction with the murder of a male student at a university in the Midwestern U.S. The content of the article was manipulated from an actual crime report, though names were altered; the priming article was also presented as a screen capture from CBSNews.com. The intended purpose of the second article was to systematically activate and make equally salient crime-related thoughts among the participants.¹ Memory and attitudinal items about the story content were again collected but not analyzed.

Finally, participants were randomly assigned to a third news article (i.e., “target article”), again manipulated from an actual crime report. Presented as a screen capture from CBSSports.com, the article chronicled the arrest of a college (1) football player or (2) soccer player on charges associated with (1) marijuana distribution (hereafter, nonviolent crime) or (2) domestic violence (hereafter, violent crime). Thus, participants were randomly exposed to one of four articles: football player arrested for a nonviolent crime (word count = 321, n = 152); football player arrested for a violent crime (word count = 320, n = 153); soccer player arrested for a nonviolent crime (word count = 323, n = 134); and soccer player arrested for a violent crime (word count = 322, n = 164). In all four versions of the story, minor content differences (e.g., MLS vs. NFL prospect; hit his girlfriend vs. was selling marijuana) were necessary. One additional, minor change from the 2018 article was made to promote greater attention to the article: The suspect’s university affiliation was changed from “Western Oregon University” to “University of Michigan.” Otherwise, the articles were identical, including the player’s name (Brandon Johnson) and the person filing charges (Johnson’s unnamed girlfriend). Importantly, no version mentioned (or pictured) the athlete’s race. As with the previous two articles, participants responded to memory and attitudinal items about the article’s contents. Finally, participants completed a posttest questionnaire on the extent to which they endorsed stereotypes about African Americans. Upon completing the posttest questionnaire, participants read an IRB-approved debriefing script and were given the option to delete their responses (though none did).

Independent Measures

As in the initial study, gender and race were self-reported. General sports fanship was measured with 11 items employing a 7-point scale (Raney & Depalma, 2006; Raney & Kinnally, 2009); our initial study used the same measure with one fewer item (i.e., “soccer”). For the replication portion of the analysis, we utilized the 10-item measure (i.e., without soccer fanship; α = .94), whereas for the extension, we used the 11-item version (α = .94) because soccer was the sport played by the athlete in the two added conditions. Combative sport fanship was measured as the average of the “football” and “hockey” items from the general fanship measure (r = .44, p < .001).

Dixon (2006) suggested that political ideology and racial bias might also be linked to judgments regarding race and crime. Neither were assessed in the original study; however, given the racial and political aspects of recent league and player responses to the BLM movement, both were measured in this study. Political ideology was operationalized and measured as social conservatism using a 7-item subscale from the Social and Economic Conservatism Scale (SECS; Everett, 2013); responses were indicated on an 11-point scale. One item (i.e., feelings toward abortion) was dropped for the sake of internal consistency; the resulting 6-item measure showed strong reliability (α = .89; M = 5.98, SD = 2.34). To assess racial bias, Dixon’s (2006) 20-item stereotype endorsement measure was used, which asks participants the extent to which they agree with a number of common stereotypes associated with African Americans on a 7-point scale. After dropping four items for the sake of internal consistency, the resulting 16-item measure was reliable (α = .81; M = 3.50, SD = .69).

Dependent Measures

All dependent measures were identical to those in the earlier study. First, to measure Black criminal stereotype activation, participants were asked to identify the race of the athlete alleged to have committed the crime in the target article, though the article made no mention of race. A 3-item response set was offered: “White,” “Black,” and “I don’t know.”

Perceptions of the suspect were measured on eight, 5-point Likert-type scale items, adopted from Dixon (2008a, 2008b). Five items were negatively valenced (threatening, dangerous, violent, criminal, aggressive); three items were positively valenced (happy, peaceful, nice) and reverse coded, yielding a reliable (α = .93) suspect perception factor, with higher scores associated with more negative perceptions of the suspect. Two items, adopted from Peffley et al. (1996), measured emotional reactions (anger, fear) to the suspect, with higher scores associated with greater negative reactions. Responses were indicated on a 5-point, Likert-type scale; the two items were significantly correlated (r = .76, p < .001).

A guilt and recidivism factor was calculated as the average of responses to three items: perceived guilt for this crime, likelihood to commit the crime again, and likelihood to act aggressively in the future (Peffley et al., 1996). Responses were indicated on a 5-point, Likert-type scale, with higher scores reflecting greater perceptions of guilt and likely recidivism; the scale showed sufficient reliability (α = .68). Finally, a single-item assessed the proposed sentence (in years) if the alleged suspect was convicted of the crime (Peffley et al., 1996). Responses were indicated on a 5-point, Likert-type scale, with higher scores indicating a more severe sentence recommendation for the suspect.

Results

Replication of Anderson and Raney (2018)

As is the case with most replication studies, the current one relied upon data from a different sample of the population. However, we are in the favorable position of having access to the full 2018 dataset. The following information is offered for the sake of comparison and as a lens for the interpretation of the results. As a reminder, two groups in the current study served as a strict replication of the 2018 study: the football player/nonviolent crime and football/violent crime conditions. The analyses presented in this section compared the 2018 data with the new data from those two conditions only.

Comparison of the Samples and Measures

The 2018 study was based on a sample of n = 238, a majority of which self-identified as women (75.6%) and Caucasian (73.8%), with an average age of 19.53 years. In the replication study, of the n = 305 participants across the two relevant conditions, a majority self-identified as women (69.5%) and Caucasian (76.4%), with an average age of 20.64 years. A set of χ² tests revealed no significant differences between the samples with regard to gender or racial representation; a t test also revealed no difference between the two samples with regard to age. In terms of general sports fanship, the 2018 sample (M₂₀₁₈ = 3.15, SD = 1.50) was statistically similar (p > .05) to the current one (M = 3.13, SD = 1.58). The same was the case for combative sports fanship: M₂₀₁₈ = 3.26, SD = 1.53; M = 3.00, SD = 1.64.

Replication of the Results

First, participants were asked to identify the race of the athlete-suspect in the target article, though the article made no mention of race. In the 2018 study, 163 (or 69.7%) of the respondents correctly responded “I don’t know”; in contrast, 71 (or 30.3%) inferred a race. In the replication study, 234 (or 76.7%) of the respondents correctly selected the “I don’t know” response; thus, 71 (or 23.3%) inferred a race. A χ² test indicated that these differences were not statistically significant, though they approached traditional levels of statistical significance: χ² (1, N = 539) = 3.40, p = .07. When these data were inspected by condition, (descriptively speaking) more participants in the nonviolent crime condition in 2018 inferred a race (n = 45) than in the violent crime condition (n = 26). However, in the current study, the pattern was reversed, with more participants in the violent crime condition inferring a race (n = 43) than in the nonviolent crime condition (n = 28).

Across the 2018 sample, participants who inferred the race of a racially unidentified suspect (n = 71) were significantly more likely to report that the suspect was Black (77.5%) than White: Wald (1) = 18.90, p < .001, Exp(B) = 3.44. Thus, participants who inferred a race when none was identified were nearly 3.5 more likely to report that the suspect was Black than White. Furthermore, in the 2018 study, general sport fanship (Wald (1) = 4.45, p < .05, Exp(B) = 1.70) and combative sport fanship (Wald (1) = 6.14, p < .05, Exp(B) = 1.79) were both significant predictors of this phenomenon. In the replication study, participants who inferred the race of the suspect (n = 71) were descriptively more likely to infer that the suspect was Black (54.9%) than White, though this difference was not statistically significant. Furthermore, no differences were observed when sports fanship or combative sports fanship was entered into the analyses.

Small cell sizes render the same logistic regression analyses difficult to interpret on a crime-condition level. However, descriptive analyses can be revealing. As noted above, in the 2018 study, more participants who inferred a race reported that the suspect was Black than White, regardless of condition. However, the discrepancy appeared greater in the violent crime condition (84.6% inferring Black) than in the nonviolent one (73.3%). In the replication study, of the participants inferring a race, more again reported that the suspect was Black than White, regardless of condition; however, the proportion doing so appeared to be lower than those reported in the 2018 study: violent crime condition (53.5% inferring Black) and nonviolent crime condition (57.1%). Results of χ² tests confirmed that the observed differences between the two samples were significant in the violent crime conditions (χ² [1, N = 68] = 6.44, p = .01), but not the nonviolent ones.

Thus, with regard to RQ1, the Black criminal stereotype was observed to be prevalent in general and especially among sports fans versus nonfans in our initial study, both in terms of general sports and combative sports fanship. In the replication study, this appears to be significantly less so the case. The proportion of participants inferring that the athlete-suspect of a violent crime was Black was significantly lower in the current study than in the previous one; across both conditions, the proportion inferring race in general also trended lower. Furthermore, differences in general and combative sports fanship did not play significant roles in the inference of race for a racially unidentified athlete-suspect.

To explore the extent to which sports fans may differ from nonfans in attitudes and beliefs about crime and violence in sport (RQ2), like the 2018 study, we conducted a series of correlation and partial correlation (with gender and race controlled) analyses with the four outcomes variables and the two fanship factors. The replication results of those analyses are reported in Tables 1 and 2.

Table 1.

Correlation Analysis With General Sport Fanship, With and Without Controlling for Participant Gender and Race.

	Anderson and Raney (2018)				Current Study
	Overall (df = 1,232)		Domestic Abuse Condition (df = 1,115)		Overall (df = 1,304)		Domestic Abuse Condition (df = 1,149)
	r	Partial r	r	Partial r	r	Partial r	R	Partial r
Suspect perceptions	−.033	−.045	−.182^*	−.183^*	−.023	.031	−.201^***	−.102
Emotional reactions	−.113^****	−.107	−.208^*	−.176^****	−.011	.014	−.164^*	−.057
Guilt and recidivism	−.039	−.027	−.123	−.104	.000	.057	−.188^*	−.098
Proposed sentencing	−.125^****	−.116^****	−.181^****	−.163^****	.059	.089	−.025	.021

Note. ^****p < .10, ^*p < .05, , ^***p < .001.

Table 2.

Correlation Analysis With Combative Sport Fanship, With and Without Controlling for Participant Gender and Race.

	Anderson and Raney (2018)				Current study
	Overall (df = 1, 232)		Domestic Abuse Condition (df = 1,115)		Overall (df = 1,304)		Domestic Abuse Condition (df = 1,149)
	r	Partial r	r	Partial r	r	Partial r	r	Partial r
Suspect perceptions	−.072	−.089	−.214^*	−.215^*	−.044	−.003	−.175^*	−.109
Emotional reactions	−.130^*	−.129^*	−.237^*	−.208^*	−.052	−.035	−.207^*	−.147^***
Guilt and recidivism	−.028	−.021	−.096	−.076	−.026	.022	−.185^*	−.116
Proposed sentencing	−.187^**	−.182^**	−.252^**	−.237^*	−.021	.042	−.030	−.004

Note. ^***p < .10, ^*p < .05, ^**p < .01.

In the replication study, none of the significant relationships reported (without regard to condition) in the 2018 study were observed. In fact, across both conditions in the current study, neither sports fanship nor combative sports fanship were significantly correlated with any of the outcome variables, including when gender and race were controlled. In the nonviolent crime condition specifically, similar to the initial study, no significant relationships were observed in the replication analyses.

Within the violent crime condition, the relationships between higher levels on both sports fanship measures and more positive/lenient ratings of the alleged suspect were particularly notable in the 2018 study. A similar pattern was observed in the violent crime condition in the replication study, with both sports fanship and combative sports fanship being significantly correlated with more positive suspect perceptions, more positive emotional reactions to the suspect, and more sympathetic ratings of the suspect’s guilt and recidivism likelihood. However, in contrast to the 2018 study, those relationships became insignificant when the effects of race and gender were controlled (save for the combative sports fanship-emotional reactions relationship, which remained significant at p < .1).

Thus, addressing RQ2, the 2018 study concluded that evaluations of a suspected criminal athlete differed between sports fans and nonfans, with the former generally being more sympathetic and lenient. In the replication study, some evidence for this pattern was observed, in particular when sports fans evaluated an athlete suspected of a violent crime. However, in several cases, the observed associations were not as strong as reported in the 2018 study. Furthermore, unlike the 2018 study, nearly all of the observed relationships became insignificant when the gender and race of the participants were controlled.

With access to the 2018 dataset, we were able to explore potential explanations for the inconsistencies observed between the two samples. To do so, we first conducted between-sample, ANOVA procedures on the four suspect evaluations by crime condition. For the nonviolent crime condition, no differences were observed on any of the four primary outcome variables between the two samples. However, within the violent crime condition, the current sample generally rated the alleged athlete-suspect more severely than their 2018 counterparts. No differences were observed on the guilt and recidivism measure across the two samples. However, perceptions of the suspect were more negative (F_1,268 = 14.25, p > .001, η² = .05; M₂₀₁₈ = 3.29, SD = .41; M = 3.52, SD = .56), emotional reactions were more negative (F_1,268 = 18.10, p > .001, η² = .06; M₂₀₁₈ = 3.00, SD = .85; M = 3.45, SD = .89), and the proposed sentence was significantly longer (F_1,268 = 8.26, p = .004, η² = .03; M₂₀₁₈ = 1.64, SD = .73; M = 1.90, SD = .75) among participants in the current violent crime condition. Thus, the replication study found significantly more negative evaluations of the suspect (in the violent crime condition) but sports fanship factors did not predict those evaluations (as they had in 2018).

To examine why this may have been the case, the roles of gender and race were further explored, given the partial correlation findings reported above. Specifically, we regressed each of the four outcome variables onto whether the participant inferred a race (yes = 0, no = 1), crime condition (violent = 0, non-violent = 1), gender (dichotomized: woman = 0, man = 1), race (dichotomized: White = 0, non-White = 1), and general sports fanship.² With the 2018 study data, all four of the regression models were statistically significant: suspect perceptions (F_5,226 = 21.76, p < .001, R² = .32), emotional reactions (F_5,226 = 22.21, p < .001, R² = .33), guilt and recidivism (F_5,226 = 4.46, p < .001, R² = .09), and sentencing (F_5,226 = 5.36, p < .001, R² = .11). In each case, crime condition (i.e., being in the violent condition) emerged as the most significant predictor: suspect perceptions, β = −.56, p < .001; emotional reactions, β = −.54, p < .001; guilt and recidivism, β = −.28, p < .001; sentencing β = −.29, p < .001. In fact, condition was the only significant predictor of suspect perceptions and guilt/recidivism. Sports fanship significantly predicted more negative emotional reactions (β = −.12, p = .041) and approached traditional levels of significance as a predictor of harsher sentencing (β = −.12, p = .063).

In the replication study, all four regression models were also significant. Having inferred a race for the suspect (β = −.11, p = .013), being in the violent condition (β = −.61, p < .001), and identifying as a woman (β = −.16, p < .001) all predicted more negative suspect perceptions (F_5,295 = 40.60, p < .001, R² = .41). With regard to emotional reactions (F_5,295 = 46.44, p < .001, R² = .41), having inferred a race for the suspect (β = −.10, p = .022) and being in the violent condition, (β = −.65, p < .001) significantly predicted more negative reactions, whereas identifying as a woman approached traditional levels of significance (β = −.09, p = .067). Having inferred a race for the suspect (β = −.17, p = .002), being in the violent condition (β = −.34, p < .001), and identifying as a woman (β = −.16, p = .007) also predicted less sympathetic perceptions of guilt and recidivism (F_5,295 = 12.31, p < .001, R² = .17). Finally, being in the violent condition (β = −.45, p < .001) and identifying as a woman (β = − .12, p = .037) predicted more harsh sentencing (F_5,295 = 17.30, p < .001, R² = .23).

Extending Anderson and Raney (2018)

Another goal of the current project was to extend this line of research into a different sport context, one that is routinely stereotyped as (more) “White” (at least in the U.S.). To do so, we replaced “football” with “soccer” as the sport played by the alleged suspect in the target article, leading to two additional conditions: a racially unidentified soccer player arrested for marijuana distribution (i.e., nonviolent crime; n = 134) and a racially unidentified soccer player arrested for domestic violence (i.e., violent crime; n = 164). The same analyses reported above were conducted on the resulting data.

In terms of inferring race (when no race was identified), within the two soccer conditions, without regard to the crime committed, 223 (or 74.8%) of the respondents correctly selected the “I don’t know” response; thus, 75 (or 25.2%) inferred a race. This proportion is statistically similar to the football conditions in the current study. Of those inferring a race, participants were overwhelmingly more likely to infer that the soccer athlete-suspect was White (84.0%) than Black; as a reminder, participants who inferred the race of the football athlete-suspect were descriptively more likely to infer that the suspect was Black (54.9%) than White. A χ² test indicated that these differences were statistically significant: χ² (1, N = 146) = 24.32, p < .001. Without regard to the sport played, descriptively speaking, more participants reading the violent crime story (n = 84) inferred a race than those reading the nonviolent crime story (n = 62), though this difference was not statistically significant.

Logistic regression analyses were used to explore potential explanations for these findings. First, the inference of a race (or not) was regressed on gender (dichotomized: woman, man), race (dichotomized: White, non-White), general sports fanship (using the 11-item measure which includes soccer fanship), combative sports fanship, social conservatism, and racial stereotype endorsement, without regard to sport or crime condition. The model was significant (χ² (6, N = 596) = 30.22, p < .001), with social conservatism (Wald (1) = 5.62, p = .02, Exp(B) = .89) and stereotype endorsement (Wald (1) = 14.23, p < .001, Exp(B) = .57) emerging as significant predictors. The same models tested at the sport-condition level were also both significant, with stereotype endorsement (Wald (1) = 13.11, p < .001, Exp(B) = .45) predicting a race inference for those in the football conditions (χ² (6, N = 301) = 26.88, p < .001) and social conservatism (Wald (1) = 4.29, p = .038, Exp(B) = .87) predicting a race inference for those in the soccer conditions (χ² (6, N = 295) = 13.38, p = .038). At the crime-condition level, the same models were again significant, with social conservatism (Wald (1) = 4.17, p = .041, Exp(B) = .87) and stereotype endorsement (Wald (1) = 6.56, p = .01, Exp(B) = .61) predicting a race inference for those who read a story about violent crime (χ² (6, N = 314) = 16.85, p = .01), and stereotype endorsement (Wald (1) = 7.85, p = .005, Exp(B) = .54) predicting a race inference for those who read a story about nonviolent crime (χ² (6, N = 282) = 19.23, p = .004).

Next, we isolated cases of participants who inferred a race (without regard to condition) and regressed the inference of White or Black onto the same factors. The regression model revealed no significant relationships. We ran the same model by sport condition and by crime condition respectively, and again the models were not significant.

To summarize: Overall, participants at higher levels of social conservatism and African American stereotype endorsement were more likely to infer (vs. not to infer) a race for a criminal when no race was identified. Some combination of those factors predicted race inferences made at the sport- and crime-condition level. Participant gender, race, or sports fanship level (neither in general nor for combative sports specifically) did not predict whether they made race inferences or not. Among those who did make a race inference, none of the measured factors predicted the specific race that was inferred (i.e., Black vs. White), regardless of sport or crime condition.

An ANOVA procedure (with a Tukey’s post hoc analysis) explored differences between conditions with regard to each of the four primary outcome variables. The same pattern emerged for each outcome: Participants in the two violent crime conditions rated the suspect more negatively/harshly than participants in the two nonviolent crime conditions. The sport played was inconsequential. See Table 3 for details.

Table 3.

Athlete-Suspect Evaluations, by Condition.

	Football, Violent		Football, Nonviolent		Soccer, Violent		Soccer, Nonviolent		Significance Tests
	M	SD	M	SD	M	SD	M	SD	F	η ²
Suspect perceptions	3.52^b	.56	2.59^a	.63	3.67^b	.57	2.61^a	.55	149.42	.428
Emotional reactions	3.45^b	.89	1.85^a	.97	3.65^b	.92	1.91^a	.99	161.30	.447
Guilt and recidivism	3.54^b	.61	3.05^a	.65	3.63^b	.66	3.10^a	.50	36.49	.155
Proposed sentencing	1.90^b	.75	1.21^a	.53	2.10^b	.96	1.23^a	.54	59.96	.231

Note. Superscripts in the same row indicate significant differences. All tests were significant at p < .001.

In the final set of analyses, linear regression models were used to examine the four outcome variables. Each outcome was regressed onto whether the participant inferred a race (yes = 0, no = 1), crime condition (violent = 0, non-violent = 1), sport condition (football = 0, soccer = 1), gender (0 = woman, 1 = man), race (0 = White, 1 = non-White), general sports fanship, social conservatism, and stereotype endorsement. Across the sample, all of the models were significant: suspect perception (F_8,587 = 62.51, p < .001, R² = .46), emotional reactions (F_8,587 = 68.51, p < .001, R² = .48), guilt and recidivism (F_8,587 = 20.21, p < .001, R² = .22), and sentencing (F_8,587 = 25.66, p < .001, R² = .26). Having inferred a race for the suspect (β = −.06, p = .041), being in the violent condition (β = −.64, p < .001), being in the soccer condition (β = .06, p = .044) and identifying as a woman (β = −.17, p < .001) predicted more negative suspect perceptions. With regard to emotional reactions, having inferred a race for the suspect (β = −.10, p = .001), being in the violent condition, (β = −.66, p < .001), and identifying as a woman (β = −.17, p < .001) all significantly predicted more negative responses, whereas being in the soccer condition also approached traditional levels of significance (β = .06, p = .065). Less sympathetic perceptions of guilt and recidivism were predicted by having inferred a race for the suspect (β = −.12, p = .001), being in the violent condition, (β = −.37, p < .001), identifying as a woman (β = −.21, p < .001), and stereotype endorsement level (β = .10, p = .012). Finally, having inferred a race for the suspect (β = −.11, p = .002), being in the violent condition (β = −.46, p < .001) and identifying as a woman (β = −.15, p < .001) also predicted more harsh sentencing.

To summarize: With regard to RQ3, the results indicate that the nature of the crime, not the sport played by the criminal, was of greater import in predicting responses to an athlete-suspect. Concerning RQ4, participants at higher levels of social conservativism and racial stereotype endorsement were more likely to infer a race for a racially unidentified athlete-suspect. Those factors, though, did not predict what race was inferred of the athlete-suspect. In fact, the only factors found to be predictive of the race inferred was sport condition, with participants in the soccer condition particularly more likely to infer that the suspect was White. Finally, having inferred a race for the suspect, being in the violent crime condition, and gender—specifically, identifying as a woman—consistently played the most influential role in predicting negative evaluations of an athlete-suspect.

Discussion

The goals of the current project were (1) to replicate our 2018 study, examining whether sports fans’ perceptions of criminal athletes have changed over the past several years, and (2) to determine whether the sport played by an athlete, as well as individual-difference variables, might be linked to judgments regarding race and crime in sport. In terms of replication, two main findings warrant our attention: one in relation to the Black criminal stereotype, another in relation to sports fanship and attitudes toward crime in sport.

The findings of our 2018 study evidenced the activation of a Black criminal stereotype, especially among sports fans. This was less prominent in the replication. Across the entire sample, a lower percentage of participants inferred that a race-unidentified suspect was Black, with sports fanship and combative sports fanship no longer predictive of those results. One possible explanation for these findings is the increased attention paid to race-related issues in sports (and general current affairs) coverage over the past several years, which has (potentially) created greater awareness of racial inequality and implicit bias among fans, with downstream effects on stereotypical judgments. Of course, such a shift is not solely bound to the sports context, as Sawyer and Gampa (2018) found that overall implicit attitudes were less pro-White during the BLM movement and became increasingly less pro-White throughout the movement. Nevertheless, the current findings related to decreased Black criminal stereotype activation among sports fans is encouraging. Additional research is needed to identify the specific cause(s) for this observation.

Regarding attitudes toward crime in sport, the findings of our initial study showed that general sports fanship was predictive of less negative attitudes about an alleged criminal athlete. Some evidence for this pattern was observed in the replication study. In particular, when evaluating an athlete suspected of a violent crime, both sports fanship and combative sports fanship were significantly correlated with more positive suspect perceptions, more positive emotional reactions to the suspect, and more sympathetic ratings of the suspect’s guilt and recidivism likelihood. Several explanations for this phenomenon—also offered in our initial study—may apply to the current one. First, given the violent nature of sports, it is possible that fans have become desensitized to violence and therefore do not see violent transgressions committed by athletes as a big deal. Second, it is also possible that fans may only care about off-field transgressions when they directly affect the games played. Whereas drug use by an athlete could have tangible on-field effects, an athlete abusing his girlfriend does not (unless, of course, it results in a suspension). Third, given sports leagues’ mishandling of cases involving violence against women (arguably the most notable being the NFL’s handling of the 2014 Ray Rice assault case), it is plausible that sports fans are just as likely as the leagues to sweep violent crimes under the rug. Finally, because hegemonic masculinity is inherent in sports culture (Connell, 1995), misogynistic attitudes in sport may influence the ways in which fans view violence against women. Regardless of the explanation, the replication study provides further evidence that sports fans differ from nonfans in basic attitudes and beliefs about crime and violence in sports, viewing a criminal athlete in a more favorable light than nonfans.

Whereas the replication results are important, perhaps the more significant findings come from the extension of the research to another sport context and the testing of (more) individual-differences variables. First, if we compare inference of race (for a race-unidentified athlete suspect) between participants in football and soccer conditions, the sport played clearly matters. In short, a “sports stereotype” appears to trump the Black criminal stereotype in this situation. Of those who inferred race, participants were significantly more likely to infer the soccer player to be White and the football player to be Black, suggesting that the ways we colloquially racialize sports are important. As noted earlier, an abundance of research demonstrates that White and Black athletes are portrayed differently across the media landscape at the individual level (e.g., Ferrucci et al., 2013; Ferrucci & Tandoc, 2017, 2018). However, research on racial stereotyping of sports themselves is limited, and much more research in this area is warranted.

Another notable difference between the soccer and football conditions were perceptions of the criminal athlete who committed a violent crime. Participants in the soccer condition reported more negative perceptions of the athlete arrested for domestic abuse and proposed a harsher sentence for the criminal athlete compared to the football condition. This further suggests that perceptions of criminal athletes vary depending on the sport played. We offer several explanations for this phenomenon. First, American football is often regarded as a violent sport, which may impact perceptions of off-field violence. As a combative sport, American football is defined by strength, power, domination, physical contact between players, and “assaultive actions directed at competitors” (Sargent et al., 1998, p. 183). Because football glorifies on-field violence, it is possible that off-field violent offenses committed by football players are (somewhat) expected and, thus, deemed more socially acceptable (or expected) in the public eye. Another possible explanation is the NFL’s lack of consistency in punishing players for violence against women, compared, for instance, to the near zero-tolerance policy for such offenses in European football. For example, when Manchester City player Benjamin Mendy was charged with four counts of rape and one count of sexual assault in 2021, he was immediately suspended by the club (Baynes, 2021). In contrast, Houston Texans quarterback Deshaun Watson has been the subject of 22 civil suits accusing him of coercive and lewd sexual behavior, two of which also alleged sexual assault (Shpigel, 2021). However, Watson’s “punishment” by the team was a trade to the Cleveland Browns, which came with a new 5-year, USD$230 million, fully guaranteed contract; five months later, he was (finally) suspended by the NFL for 11 games.

A final potential explanation for finding more negative perceptions of the violent criminal soccer player ties together sports stereotyping and perceptions of criminality in sport. Recently, Brown et al. (2021) found evidence of a “racial contradiction” in media messages of criminal athletes, where exposure to messages in sports news could lead to White athletes being viewed more negatively than Black ones. Since Black athletes are overrespresented in news media as criminals and described more negatively compared to their White counterparts (e.g., Dixon, 2008a; Leshner, 2006), White athletes who are faced with criminal transgressions could be viewed more negatively by the public because they are not expected to face such accusations. Regarding our study, of those who inferred a race in the soccer conditions, a majority inferred the player to be White. The “racial contradiction” may explain why participants had more negative perceptions of the soccer player. In the absence of naming a race, the assumption apparently was that the soccer player was White, and the participants seemed to judge him more harshly in this context. Whereas Brown et al. (2021) found evidence for “racial contradiction” only among high sports news consumers, suggesting cultivation theory as a possible explanation for this effect, in our current study, no significant patterns were observed regarding sports media consumption; both high and low sports media consumers perceived the soccer player more negatively than the football player. With this in mind, we encourage others in the sports media research community to explore perceptions of criminal athletes across sporting contexts, especially in relation to general media, as well as sports media, consumption.

The current study also provided evidence for three individual-difference variables—social conservatism, African American stereotype endorsement, and gender—predicting attitudes and judgments of race and crime in sport. Specifically, participants at higher levels of social conservatism and African American stereotype endorsement were more likely to infer a race for a suspected criminal athlete when no race was identified. This is in line with previous research outside the sporting context, which posits that political ideology and racial bias both play a large role in judgments regarding race and crime (Dixon, 2006, 2008a, 2008b). Regarding perceptions of the criminal athlete, social conservatism was also more strongly associated with negative perceptions, emotions, and guilt of the athlete than either of the sports fandom factors, which is not surprising, given conservative views on criminality. Not only do these findings extend previous research on the Black criminal stereotype to the context of sport, but they also signify the importance of separating social and economic conservatism when measuring how political ideology may facilitate racial stereotypes (see Everett, 2013).

Finally, the study extension identified gender as a particularly important predictor of attitudes toward the criminal athlete. In the current study, the single factor that most strongly predicted negative attitudes towards a violent athlete (regardless of sport played) was identifying as a woman. In fact, identifying as a woman was significantly correlated with more negative/harsh scores on all four outcomes: stronger negative perceptions of the athlete (threatening, dangerous, etc.), stronger emotional reaction to the athlete (anger and fear), stronger perceptions of guilt of the athlete, and harsher suggested sentencing. These results mark a drastic shift from our 2018 study, which reported no gender-based differences. Although sports fanship still influenced perceptions of violent criminality by an athlete (i.e., sports fans reported more lenient attitudes), gender played a much more powerful role in those evaluations.

Given the findings from our 2018 study, these results were unexpected. But, in hindsight, we think that league and player responses to the BLM movement may not have been the only social conversation in sports coverage that affected attitudes about athlete criminality. Because the violent crime depicted in the target news article was domestic abuse, we think that sports (and general news) coverage of the #MeToo movement (see https://metoomvmt.org) may also be reflected in the current findings. Before the widespread coverage of the #MeToo movement, media accounts of violence against women—including violence perpetrated by athletes—often trivialized the problem, reinforcing victim blaming (Meyers, 1997). Such language supports and reproduces male supremacy and patriarchal views, which can influence how both men and women evaluate public cases of domestic abuse. Before the #MeToo movement, both men and women—including those in our 2018 sample—may have accepted domestic abuse committed in general, but specifically by a hypermasculine athlete, as “just the way things are.” But it seems reasonable to think that the millions of survivors sharing their stories (and the accompanying media coverage thereof) may have shifted societal beliefs surrounding violence against women. We think that it is possible that the social reverberations of the movement may have contributed to a shift in female attitudes and beliefs surrounding domestic abuse (see also Leopold et al., 2021; MacKinnon, 2018; Restrepo Sanín, 2019), which are reflected in the current data. Although athlete violence against women is not a new phenomenon, the #MeToo movement raised the visibility of these crimes and shifted societal understandings of violence against women. In the context of sports specifically, #MeToo perhaps created a space for women to push back and resist hypermasculine culture, where athlete abuse against women has been ignored for far too long. Admittedly, these proposed connections are pure speculation on our part. More research is needed to see if the current findings are replicable and, if so, to explore the possible causes of such a shift in sport fans.

Like all studies, the current one has limitations. As we tried to emphasize above, the proposed connection between the findings and the BLM and #MeToo movements are admittedly speculative, and much more research is needed to explain the (surely) varied reasons behind our findings. We also acknowledge that our findings may not generalize to international sporting contexts, and we urge more scholars to conduct similar research on sports fandom and perceptions of criminal athletes outside of the U.S. Further research is also needed to explore the Black criminal stereotype in sporting contexts other than football and soccer. Nevertheless, as a second attempt at examining the relationship between sports fanship and the Black criminal stereotype, we think that this study offers important findings.

Ultimately, our findings show that sports (still) matter. But sports are always played within larger, ever-evolving social and cultural contexts. Without a doubt, the findings related to sports fanship as a predictor of racial stereotyping and differing attitudes about violent athlete criminality are important. However, we think the more significant findings—fewer participants inferring that a race-unidentified suspect was Black and shifting female attitudes relating to perpetrators of domestic violence—may ultimately be explained by larger social and cultural forces. We hope that this study encourages others to explore these and related issues, in an attempt to better understand the complicated role of sport in those broader contexts.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Lauren C. Anderson

Notes

References

Anderson

L. C.

Raney

A. A.

(2018). Exploring the relationship between sports fandom and the Black criminal stereotype. Communication & Sport, 6(3), 263–282. https://doi.org/10.1177/2167479517713152

Arth

Billings

A. C.

(2019). Touching racialized bases: Ethnicity in Major League Baseball broadcasts at the local and national levels. Howard Journal of Communications, 30(3), 230–248. https://doi.org/10.1080/10646175.2018.1466746

Ash

Cranmer

G. A.

(2020). Transcontextual effects of racialized sports media framing: Students’ off the field perceptions of student-athletes. Communication & Sport, 8(3), 389–408. https://doi.org/10.1177/2167479519830621

Assimakopoulos

(2020, September 1). NFL will display social justice statements in the end zone this season. NBC Sports. https://www.nbcsports.com/washington/football-team/nfl-will-display-social-justice-statements-end-zone-every-stadium-season

Baynes

(2021, August 27). Manchester City footballer Benjamin Mendy charged with four counts of rape. Sky News. https://news.sky.com/story/manchester-city-footballer-benjamin-mendy-charged-with-four-counts-of-rape-12391379

Berry

Smith

(2000). Race, sport, and crime: The misrepresentation of African Americans in team sports and crime. Sociology of Sport Journal, 17(2), 171–197. https://doi.org/10.1123/ssj.17.2.171

Billings

A. C.

(2003). Portraying tiger woods: Characterizations of a “Black” athlete in a “White” sport. Howard Journal of Communications, 14(1), 29–37. https://doi.org/10.1080/10646170304274

Billings

A. C.

(2004). Depicting the quarterback in black and white: A content analysis of college and professional football broadcast commentary. Howard Journal of Communications, 15(4), 201–210. https://doi.org/10.1080/10646170490521158

Billings

A. C.

Eastman

S. T.

(2002). Selective representation of gender, ethnicity, and nationality in American television coverage of the 2000 Summer Olympics. International Review for the Sociology of Sport, 37(3–4), 351–370. https://doi.org/10.1177/101269020203700302

10.

Bjornstrom

E. E.

Kaufman

R. L.

Peterson

R. D.

Slater

M. D.

(2010). Race and ethnic representations of lawbreakers and victims in crime news: A national study of television coverage. Social problems, 57(2), 269–293. https://doi.org/10.1525/sp.2010.57.2.269

11.

Boren

(2020, August 25). Roger Goodell wishes the NFL ‘had listened earlier’ to Colin Kaepernick, supports players who kneel. Washington Post . https://www.washingtonpost.com/sports/2020/08/24/roger-goodell-wishes-nfl-had-listened-earlier-colin-kaepernick/

12.

Brown

K. A.

Dickhaus

Harrison

Rush

(2021). Explaining the “racial contradiction”: An experimental examination of the impact of sports media use and response strategy on racial bias towards athlete transgressors. Communication & Sport, 9(5), 833–855. https://doi.org/10.1177/2167479519886118

13.

Campbell

C. P.

(1995). Race, myth and the news. Sage.

14.

Connell

R. W.

(1995). Masculinities. Polity Press.

15.

Denham

B. E.

Billings

A. C.

Halone

K. K.

(2002). Differential accounts of race in broadcast commentary of the 2000 NCAA men’s and women’s final four basketball tournaments. Sociology of Sport Journal, 19(3), 315–332. https://doi.org/10.1123/ssj.19.3.315

16.

Devine

P. G.

(1989). Stereotypes and prejudice: Their automatic and controlled components. Journal of Personality and Social Psychology, 56(1), 5–18. https://doi.org/10.1037/0022-3514.56.1.5

17.

Dixon

(2006). Psychological reactions to crime news portrayals of black criminals: Understanding the moderating roles of prior news viewing and stereotype endorsement. Communication Monographs, 73(2), 162–187. https://doi.org/10.1080/03637750600690643

18.

Dixon

T. L.

(2007). Black criminals and White officers: The effects of racially misrepresenting law breakers and law defenders on television news. Media Psychology, 10(2), 270–291. https://doi.org/10.1080/15213260701375660

19.

Dixon

T. L.

(2008a). Crime news and racialized beliefs: Understanding the relationship between local news viewing and perceptions of African Americans and crime. Journal of Communication, 58(1), 106–125. https://doi.org/10.1111/j.1460-2466.2007.00376.x

20.

Dixon

T. L.

(2008b). Who is the victim here? The psychological effects of overrepresenting White victims and Black perpetrators on television news. Journalism, 9(5), 582–605. https://doi.org/10.1177/1464884908094160

21.

Dixon

T. L.

(2020). Media stereotypes: Content, effects, and theory. In Oliver

M. B.

Raney

A. A.

Bryant

(Eds.), Media effects: Advances in theory and research (4th ed., pp. 243–257). Routledge.

22.

Dixon

T. L.

Azocar

C. L.

(2007). Priming crime and activating Blackness: Understanding the psychological impact of the overrepresentation of Blacks as lawbreakers on television news. Journal of Communication, 57(2), 229–253. https://doi.org/10.1111/j.1460-2466.2007.00341.x

23.

Dixon

T. L.

Linz

(2000). Overrepresentation and underrepresentation of African Americans and Latinos as lawbreakers on television news. Journal of Communication, 50(2), 131–154. https://doi.org/10.1111/j.1460-2466.2000.tb02845.x

24.

Dixon

T. L.

Maddox

K. B.

(2005). Skin tone, crime news, and social reality judgments: Priming the stereotype of the dark and dangerous Black criminal. Journal of Applied Social Psychology, 35(8), 1555–1570. https://doi.org/10.1111/j.1559-1816.2005.tb02184.x

25.

Everett

J. A.

(2013). The 12 item social and economic conservatism scale (SECS). PloS one, 8(12), e82131. https://doi.org/10.1371/journal.pone.0082131.

26.

Ewoldsen

D. R.

Rhodes

(2020). Media priming and accessibility. In Oliver

M. B.

Raney

A. A.

Bryant

(Eds.), Media effects: Advances in theory and research (4th ed., pp. 83–99). Routledge.

27.

Ferrucci

Tandoc

E. C.

(2017). Race and the deep ball: Applying stereotypes to NFL quarterbacks. International Journal of Sport Communication, 10(1), 41–57. https://doi.org/10.1123/ijsc.2016-0074

28.

Ferrucci

Tandoc

E. C.

(2018). The spiral of stereotyping: Social identity theory and NFL quarterbacks. Howard Journal of Communications, 29(2), 107–125. https://doi.org/10.1080/10646175.2017.1315693

29.

Ferrucci

Tandoc

E. C.

Painter

C. E.

Leshner

(2013). A black and white game: Racial stereotypes in baseball. Howard Journal of Communications, 24(3), 309–325. https://doi.org/10.1080/10646175.2013.805971

30.

Grainger

Newman

J. I.

Andrews

D. L.

(2006). Sport, the media, and the construction of race. In Raney

A. A.

Bryant

(Eds.), Handbook of sports and media (pp. 482–504). Lawrence Erlbaum Associates.

31.

Leonard

(2006). A world of criminals or a media construction? Race, gender, celebrity, and the athlete/criminal discourse. In Raney

A. A.

Bryant

(Eds.), Handbook of sports and media (pp. 564–584). Lawrence Erlbaum Associates.

32.

Leopold

Lambert

J. R.

Ogunyomi

I. O.

Bell

M. P.

(2021). The hashtag heard round the world: How #MeToo did what laws did not. Equality, Diversity and Inclusion: An International Journal, 40(4), 461–476. https://doi.org/10.1108/EDI-04-2019-0129

33.

Leshner

(2006). The effects of dehumanizing depictions of race in TV news stories. In Reynolds

Barnett

(Eds.), Communication and law: Multidisciplinary approaches to research (pp. 233–251). Lawrence Erlbaum Associates.

34.

Stelloh

(2020, August 26). Pro sports teams delay games after Milwaukee Bucks refuse to play in protest of Jacob Blake shooting. NBC. Retrieved from https://www.nbcnews.com/news/us-news/milwaukee-bucks-%20don-t-take-floor-nba-playoff-game-apparently-n1238250

35.

MacKinnon

(2018, February 4). MeToo has done what the law could not. The New York Times. https://www.nytimes.com/2018/02/04/opinion/metoo-law-legal-system.html#comments

36.

Mastro

Blecha

Atwell Seate

(2011). Characterizations of criminal athletes: A systematic examination of sports news depictions of race and crime. Journal of Broadcasting and Electronic Media, 55(4), 526–542. https://doi.org/10.1080/08838151.2011.620664

37.

Mayo

(2012, March 23). Miami Heat pose in hoodies to protest handling of the Trayvon Martin killing in Florida. Michigan Live. https://www.mlive.com/pistons/2012/03/miami_heat_pose_in_hoodies_to.html

38.

Meyers

(1997). News coverage of violence against women: Engendering blame. Sage.

39.

Nicolaides

Lim

(2020). Undergoing the experiences of the COVID-19 pandemic as ruptures in American civil society create conditions for right action. Reflective Practice, 21(6), 844–860. https://doi.org/10.1080/14623943.2020.1821632

40.

Niven

(2021). Who says shut up and dribble? Race and the response to athletes’ political activism. Journal of African American Studies, 25(2), 298–311. https://doi.org/10.1007/s12111-021-09534-6

41.

Oliver

Hoewe

Ash

Kim

Chun

Shade

(2014). Media and social groups. In Oliver

M. B.

Raney

A. A.

(Eds.), Media and social life (pp. 81–97). Routledge.

42.

Peffley

Shields

Williams

(1996). The intersection of race and crime in television news stories: An experimental study. Political Communication, 13(3), 309–327. https://doi.org/10.1080/10584609.1996.9963120

43.

Raney

A. A.

Depalma

A. J.

(2006). The effect of viewing varying levels and contexts of violent sports programming on enjoyment, mood, and perceived violence. Mass Communication & Society, 9(3), 321–338. https://doi.org/10.1207/s15327825mcs0903_4

44.

Raney

A. A.

Kinnally

(2009). Examining perceived violence in and enjoyment of televised rivalry sports contests. Mass Communication and Society, 12(3), 311–331. https://doi.org/10.1080/15205430802468744

45.

Rasmussen

Esgate

Turner

(2005). On your marks, get stereotyped, go: Novice coaches and Black stereotypes in sprinting. Journal of Sport & Social Issues, 29(4), 426–436. https://doi.org/10.1177/0193723504273122

46.

Reid

(2020, June 11). NFL pledges to donate $250 million over 10-year period. The Undefeated. https://theundefeated.com/features/nfl-pledges-to-donate-250-million-over-a-10-year-period/

47.

Restrepo Sanín

(2019). #MeToo: What kind of politics? Panel notes. Journal of Women, Politics and Policy, 40(1), 122–128. https://doi.org/10.1080/1554477X.2019.1563418

48.

Rohrbach

(2020, August 27). The brief history of NBA player protests, a trail blazed by Elgin Baylor and Bill Russell. Yahoo! https://www.yahoo.com/now/the-brief-history-of-nba-player-protests-a-trail-blazed-by-elgin-baylor-and-bill-russell-205734051.html

49.

Sargent

S. L.

Zillmann

Weaver

J. B.

(1998). The gender gap in the enjoyment of televised sports. Journal of Sport & Social Issues, 22(1), 46–64. https://doi.org/10.1177/019372398022001005

50.

Sawyer

Gampa

(2018). Implicit and explicit racial attitudes changed during Black Lives Matter. Personality and Social Psychology Bulletin, 44(7), 1039–1059. https://doi.org/10.1177/0146167218757454

51.

Shpigel

(2021, November 8). What to know about the sexual misconduct lawsuits against Deshaun Watson. New York Times. https://www.nytimes.com/article/will-deshaun-watson-play.html

52.

TePoel

Nauright

(2021). Black lives matter in the sports world. Sport in Society, 24(5), 693–696. https://doi.org/10.1080/17430437.2021.1901392

53.

Tewksbury

Scheufele

(2020). News framing theory and research. In Oliver

M. B.

Raney

A. A.

Bryant

(Eds.), Media effects: Advances in theory and research (4th ed., pp. 51–68). Routledge.