Assignment: Attribution Errors

Contents lists available at ScienceDirect

Psychology of Sport & Exercise

journal homepage: www.elsevier.com/locate/psychsport

Confirmation bias and the stereotype of the black athlete

Gordon B. Moskowitz∗, Devon Carter Lehigh University, USA

A R T I C L E I N F O

Keywords: Implicit bias Stereotyping Subtyping Confirmation bias Person perception Social judgment Black athlete

A B S T R A C T

Objectives: Examine how the subtype “black athletes” differs from that of black men. More centrally, to illustrate that the category black athletes can be triggered and lead to implicit bias. This bias is implicit and impacts a perceiver's evaluation outside conscious awareness. Design and method: Participants were white residents of the United States. Experiment one assessed familiarity with three qualities comprising the cultural stereotype of black athletes. Experiment two explored implicit bias from the stereotype using a within subjects design. Participants read a series of quotes and rated how strongly they implied these stereotypical qualities. Each participant rated such quotes from a black athlete, white athlete, black salesman, and a white salesman. Results: Implicit bias was revealed in the type of ratings provided of black athletes. The ratings for the black athletes along these two stereotypical dimensions significantly differed from the ratings made to the same quotes when uttered by other men. Conclusions: The stereotype guided how participants viewed the person's actions/words. The stereotype of black athletes is distinct from black men generally. Participants did not rate men negatively along the stereotypic dimensions when they were black salesmen, only when they were black athletes. Additionally, although the quality “athletic,” which is part of the global stereotype of black men, has a positive connotation, the stereotype that is built around this quality is not positive. It is a stereotype that this group has natural athletic ability that leads to an arrogance and lack of work ethic (since one is naturally gifted).

Implicit bias occurs when stereotyping and prejudice impacts people outside of awareness and without conscious intent. It arises when either a stereotype or an attitude is triggered by thoughts of a social group. This “triggering,” (known as priming, e.g., Higgins, 1996) then shapes how we respond. A wide set of responses are affected by implicit bias – from how we categorize, where attention is allocated, the types of judgments and inferences we form, the expectations and standards we set, what we consider valid and veridical, how we feel, our approach and avoidance tendencies, and how we more generally act (e.g., Moskowitz & Olcaysoy Okten, 2018; Nelson, 2018). Our current con- cern is with how stereotypes of black athletes implicitly bias one's judgment. In a first experiment we examine the stereotype of the black athlete, identifying if there is consensus around a set of attributes posited to be part of the cultural beliefs for this category. In Experiment two we illustrate how the implicit triggering/priming of the stereotype “black athlete” alters the types of judgments one makes, and does so in a manner distinct from the stereotype of black men more generally. The influence exerted is consistent with a theory of aversive racism in which the bias occurs outside of awareness and without conscious intent (e.g., Dovidio & Fiske, 2012).

The stereotype of the black athlete has a long history of study, gaining prominence as media coverage of sporting events became more regular. Rainville and McCormick (1977) raised the argument that a persistent bias exists in how the media talks about athletes, with an analysis showing that announcers place more praise on white players and more criticism on comparable black players. Staples and Jones (1985) argued that these portrayals reinforce the image of white su- periority and black inferiority, thus stabilizing and justifying the ex- isting disparities. The prevalence of the negative stereotype was sum- marized by Eastman and Billings (2001): “The traditional prejudices about Black players and concomitant flattering of White players persist, despite changing times and an increased number of minority and women announcers in college basketball. Stereotypes seem to be the language of sport, at least in college basketball, and few sportscasters make an effort to break out of the patterns of speech used by their predecessors” (p. 198). This has been supported by studies using a biased coverage index (BCD) to test for announcer bias in television coverage. These analyses reveal that announcers emphasize athleticism when discussing African American players but emphasize cognitive abilities (sports intelligence) when discussing White players (Rada,

https://doi.org/10.1016/j.psychsport.2018.02.010 Received 25 May 2017; Received in revised form 16 February 2018; Accepted 18 February 2018

∗ Corresponding author. E-mail address: [email protected] (G.B. Moskowitz).

Psychology of Sport & Exercise 36 (2018) 139–146

Available online 20 February 2018 1469-0292/ © 2018 Elsevier Ltd. All rights reserved.

T

1996). The stereotype of the black athlete stands at the intersection of two

well-known (and more global) stereotypes – those of athletes and of African Americans. Established literatures exist that delineate the social stereotypes of both African Americans (e.g., Allport, 1954; Devine & Elliot, 1995; Katz & Braly, 1933) and athletes (e.g., Edwards, 1973; 1984; Stone, Harrison, & Mottley, 2012). Complicating the matter is the fact that the quality of “athleticism” is part of the global stereotype of African Americans; it is a positive quality nested among a set of mostly negative traits comprising the global stereotype (Barden, Maddux, Petty, & Brewer, 2004). This might imply that the sub-stereotype of the black athlete is positive, like the attribute “athletic” (see Devine & Baker, 1991). In contrast, it remains possible that the stereotype of the black athlete is not seen positively, but that the global stereotype of African Americans overwhelms perceptual experience and guides how the black athlete is viewed (in essence, there is no difference in how the black man versus the black athlete is seen). A third possibility is that the category “black athlete” is negative like the global stereotype, but with a set of distinct negative qualities that define this group in a manner different to the global stereotype – that it is a distinct subtype of the group more generally (e.g., Kunda & Oleson, 1995).1

We explore here this third possibility – that the stereotype of the black athlete is not positive, but is also quite separate from the (nega- tive) global stereotype of African Americans. As such, the impact of the black athlete stereotype on judgment and behavior will be distinct from the impact of the stereotype for black men. The subtype of the “black athlete” is defined by an instance of an otherwise positive characteristic (athletic) being isolated from the global stereotype, and rather than building a positive stereotype around this quality, a subtype is created in which the positive quality is morphed into a set of negative char- acteristics describing the subgroup. The single, positive quality of ath- leticism from the global stereotype has become the central, unifying quality around a new stereotype (e.g., Asch, 1946) that identifies ath- leticism as a natural ability one is lucky to have. Given that the attribute is seen as innate and not earned, it is not seen as indicative of anything positive about the person's character, or as stemming from a psycho- logical quality internal to the person (goals, traits, effort). While ath- leticism is a quality that is unarguably about the person, the person endorsing this stereotype attributes its origins not to the qualities of the individual athlete, but to external forces such as luck and breeding (Harrison, 1998).

With such attributions, the seemingly positive quality of “athletic” takes on a new meaning built around a largely negative narrative (e.g., Asch, 1946). In Experiment one we explore this issue, examining whether people categorized by this subtype of the black athlete are believed to be (are stereotyped as): 1) having been born with natural athletic ability (e.g., Stone et al., 2012), 2) taking their ability for granted and possessing a poor work ethic as a result of this reliance on natural talents (shunning practice, persistence, and effort), and 3) de- veloping a personality marked by arrogance (and showiness/ostenta- tiousness) due to this sense of having innate, superior ability (Devine & Baker, 1991). Harrison (1998) describes the public perception of the overpaid and arrogant black athlete, who is supposed to be impervious to injury and pain.

1. Experiment one

In Experiment one we explored the question of whether participants in our sample were aware of the components comprising the stereotype (subtype) of the black athlete. The degree to which they were familiar

with this subtype was assessed through a short survey. A perceiver cannot be influenced in how they judge another person if they are unaware (do not have available in memory) the stereotypic content. Experiment one sought to establish the stereotype's content, which then allowed us to explore its influence on judgment in Experiment two. Participants in Experiment one who were aware of the stereotype were then eligible to participate in our second experiment exploring the in- fluence of the subtype on social judgment. Experiment one also ex- plored if individual differences in motivation to be an egalitarian person might shape knowledge of the stereotype. It is possible that people who are particularly afraid of appearing biased to others (ex- ternally motivated) might be more reluctant to report the stereotype.

1.1. Participants

Amazon Mechanical Turk was used to recruit participants. Participants in the experiment were 241 white, English speaking re- sidents of the United States. Participants completed the survey in ex- change for fifty cents in compensation.

1.2. Materials

Knowledge of the Stereotype. Participants were asked three ques- tions, each assessing their familiarity with one element of the stereo- type of the black athlete. Each question was on a seven point scale, anchored by a score of “1” defined as “not at all familiar” and a score of “7” defined as “extremely familiar”. Scores of “3” and higher indicated familiarity with the stereotype to varying degrees (familiar, somewhat familiar, moderately familiar, strongly familiar, extremely familiar). One question asked about lack of work ethic (rate to what extent you have heard the following stereotype about black athletes in US culture: Black athletes are less hard working than white athletes). One question asked about arrogance (rate to what extent you have heard the fol- lowing stereotype about black athletes in US culture: Black athletes are more arrogant than white athletes). A final question asked about nat- ural athletic ability (rate to what extent you have heard the following stereotype about black athletes in US culture: Black athletes have more natural athletic ability than white athletes).

Demographic Questions. Participants answered a series of demo- graphic questions that assessed their age, race, gender, and experience with athletics (professional, recreational, etc.).

Motivation to be Egalitarian. Participants answered four questions that assessed their strength and type of motivations for being egali- tarian to black men taken from a scale developed by Plant and Devine (1998). There were internal motivations assessed (“I attempt to act in nonprejudice ways toward other ethnic groups because it is personally important to me”; Being nonprejudiced toward other ethnic groups is important to my self-concept) and external motivations assessed (“I try to hide any negative thoughts about other ethic groups in order to avoid negative reactions from others”; “I try to act nonprejudiced toward other ethnic groups because of pressure from others”).

2. Results

A set of frequency counts was produced to indicate what percentage of participants were familiar with each element of the stereotype. Scores on the stereotype assessment scale less than three (indicating no familiarity with the specified element of the stereotype) were compared against scores three and higher (that indicated familiarity with the stereotype). For the question that asked about lack of hard work, we found that 52% of participants were familiar with this stereotype (25% of the total using scores of 3 and 4, and 27% using scores of 5, 6, and 7). For the question that asked about arrogance, we found that 64% of participants were familiar with this element of the stereotype (22% of the total using scores of 3 and 4, and 42% using scores of 5, 6, and 7). Finally, for the question that asked about natural athletic ability, we

1 It is possible that another subtype has developed surrounding athleticism that is ir- relevant to race – that of the superstar athlete. Here, athletes such as Michael Jordan, Tom Brady, Babe Ruth, Wayne Gretzky, Muhammed Ali, Tiger Woods, Serena Williams, etc. transcend notions of race and instead become iconic figures of the culture. This “super- star” subtype is not our focus.

G.B. Moskowitz, D. Carter Psychology of Sport & Exercise 36 (2018) 139–146

140

found that 89% of participants were familiar with this stereotype (15% of the total using scores of 3 and 4, and 74% using scores of 5, 6, and 7). Our focus for further analysis turned to the first two items since they will be the items assessed in experiment two.

We next explored if there were any demographic differences in ratings of the stereotypic qualities of arrogance and lack of hard work. Ratings of familiarity with the arrogant stereotype do not correlate with external motivations to respond without prejudice (r = −.04) or to internal motivations to respond without prejudice (r = −.04). Ratings of the lack of hard work component of the stereotype also do not cor- relate with either external (r = −.1) or internal (r = −.1) motivation to respond without prejudice. Ratings of familiarity with the stereotype do not correlate with age (r = −.04 and r = −.06 for the arrogant and for the hard working component of the stereotype respectively). Looking at political ideology, 64% of people who claim to be politically neutral, 65% of liberals, and 65% of conservatives report being familiar with the arrogant stereotype. For the lack of hard work stereotype, 48% of people who claim to be politically neutral, 52% of liberals, and 52% of conservatives report being familiar with this stereotype. Thus, no differences in knowledge of the stereotype emerge for ideology either. Turning to gender, 59% of women were familiar with the arrogant stereotype, 75% of men report being familiar with the arrogant ste- reotype. For the lack of hard work stereotype, only 43% of women were familiar with this stereotype, while 64% of men report being familiar with this stereotypic quality. Overall, the stereotype of the black athlete seems to be shared equally by people in a wide variety of demographic categories, with the one exception being that women report being less familiar with the stereotype than men (yet still knowledgeable of the stereotype to sufficient degrees, as discussed below).

Are the rates with which participants indicated familiarity with the hard work and arrogance aspects of the stereotype sufficiently high to indicate this is a generally held stereotype in the culture? By compar- ison, when Devine (1989) assessed stereotypes of blacks in the United States, white participants had consensus on a set of traits comprising the stereotype. Out of the eight top traits that made up the stereotype in Devine's work, six had percentages that ranged between 50% and 60% agreement that the trait was an element of the stereotype. This is consistent with the 52% and 64% ratings we found for the lack of hard work and arrogant components of the stereotype, respectively (and the 43% of women familiar with the lack of hard work component of the stereotype is in the same range as the ninth and tenth traits Devine identified). In Devine's experiment the top two traits of the global ste- reotype were poor (78%) and criminal (70%). Natural athletic ability was seen here as recognized by 89% of people.

We can conclude based on our findings that the stereotypic qualities of the black athlete as not working as hard as white athletes, and being more arrogant than white athletes, are familiar to most participants and can be described as part of a cultural stereotype. To examine if knowledge of this stereotype can subtly influence how people judge a black athlete, we chose to focus Experiment two only on participants who had stated they were familiar with the two components of the stereotype of interest. Our concern is whether among those who know the stereotype, is there an unknown influence of the stereotype on how they see the world?

Given that all the participants selected for Experiment two know the two stereotypic aspects (arrogance and lack of hard work), would they be equally influenced in rating a black athlete by the arrogant stereo- type as the lack of hard work stereotype? We predicted that when the rating being made seems obvious to participants as designed to assess their stereotypes, then these participants would avoid being influenced by the stereotype as a way to manage their appearance and not seem to be biased. However, as predicted by the theory of aversive racism (e.g., Pearson, Dovidio, & Gaertner, 2009), when the rating could be justified as not being based on a stereotype, the influence of the stereotype would emerge; Black athletes would be judged negatively relative to black men who were not athletes (and relative to white men).

3. Experiment 2

3.1. Implicit bias in social judgment

Much of the scientific inquiry relating to the subtype of the black athlete pertains to stereotype threat – or how identity cues that make one's membership in this group salient can then impact one's perfor- mance in a stereotype-relevant domain (e.g., Beilock & McConnell, 2004; Czopp, 2010; Stone et al., 2012; Stone, Lynch, Sjomeling, & Darley, 1999). Our concern is not with issues of identity threat, or with the behavior of black athletes who are the targets of bias, but with issues relating to confirmation bias and implicit stereotyping. These issues concern how a perceiver judges a black athlete differently than white athletes (and differently than other categories of black men), seeing what one expects, with expectations determined by a social stereotype that one may not even realize one holds (e.g., Allport, 1954; Darley & Fazio, 1980; Duncan, 1976; Hamilton & Trolier, 1986; Kunda, 1987).

The literature on implicit bias informs that a perceiver whose ste- reotype of black men is triggered will make judgments of black men that are more negative, with the dominant qualities guiding judgment being those specified by the stereotype – aggression, hostility, crimin- ality, and unintelligence (e.g., Blair, 2002; Correll et al., 2007; Devine, 1989; Duncan, 1976; Payne, 2006). For example, when a perceiver observes someone shoving another person in an ambiguous way, if the person doing the shoving is black the behavior is seen as an act of ag- gression and hostility, but if the person doing the shoving is white it is seen as playfulness and “joking around.”2

Outside of the lab these subtle biases manifest in pernicious ways: the sentencing decisions of judges are biased by how stereotypically black a defendant accused of a violent crime is (Eberhardt, Davies, Purdie-Vaughns, & Johnson, 2006). Doctors make different diagnoses given identical symptoms and patient profiles if the patient is black (Schulman et al., 1999). Science faculty rate male candidates for a position as better qualified than women with identical portfolios (Moss- Racusin, Dovidio, Brescoll, Graham, & Handelsman, 2012). And in the domain of sport, referees call more fouls on black athletes than white athletes (Price & Wolfers, 2007).3

3.2. Implicit bias and the subtype of the black athlete

What are the judgmental effects of triggering a subtype, such as black athlete? Presumably, the subtype would be an altogether different set of content from the global stereotype (e.g., Kunda & Oleson, 1995). If the subtype of black athletes (described above) is what becomes triggered in the mind of a perceiver, then it will have a distinctive impact on how perceivers make judgments. It would not result in one seeing the qualities of hostility and criminality that mark the impact of the global stereotype (and may even inhibit such qualities, e.g., Macrae,

2 While the influence of the stereotype on judgment is typically to make one see a member of the stereotyped group in a fashion that matches the global evaluation of the stereotype, this is not always the case. It is possible for one to be biased in the direction of a contrast effect, where one interprets a person's behavior in the opposite way that is implied by the stereotype (e.g., Moskowitz & Skurnik, 1999; Wegener & Petty, 1995).

3 Implicit bias is not limited to stereotypes affecting our judgment, but attitudes (evaluations of valence) are subtly impacted as well (e.g. Nosek et al., 2007). Negative associations to the group (prejudices) can be triggered without one realizing, as is re- vealed by a large literature on the implicit association test (IAT; Nosek, et al., 2007). Bias in how one feels can emerge both from the prejudice one has to the group triggering negativity, or from an individual element or characteristic of the group triggering an affective reaction. For example, Barden et al. (2004) show that if a perceiver is thinking about the quality of athleticism as it relates to black men, then the perceiver is more positive in how one evaluates a black man. It is important to note that this work does not involve the triggering of the subtype of the black athlete, but a focus on one aspect or characteristic of the entire group (black men) with a strong (positive) evaluative com- ponent. In sum, both positive and negative evaluation can be triggered, and the triggering can come from a global feeling of prejudice to a group, or from a single quality.

G.B. Moskowitz, D. Carter Psychology of Sport & Exercise 36 (2018) 139–146

141

Bodenhausen, & Milne, 1995). Nor would it make one evaluate the person positively, as when the single quality of “athleticism” is used as a standard of judgment (e.g., Barden et al., 2004). Instead, guided by the content of the subtype, one should be ready to perceive natural ability, and an arrogance and lack of hard work that comes with re- liance on that ability; perceivers believe the athlete arrogantly feels that he/she does not need to nurture innate ability through effort and hard work.

Stone, Perry, and Darley (1997) performed one of the few experi- ments examining this issue of how black athletes are perceived. Their participants were asked to rate the attributes and performance of a basketball player after listening to a radio broadcast of a collegiate basketball game. For half the participants, the player was said to be a white man; for the other half he was depicted as black. Participants then gave estimates of the player's natural ability and evaluations of his hustle, court-smarts, and position-play. Despite judging the exact same behavior, those who rated a black athlete saw greater natural ability, less hustle, and less intelligence.

Of course, laziness (lack of hustle), athleticism, and lack of in- telligence are all part of the global stereotype of African Americans (e.g., Devine, 1989). Thus, it remains unclear if black athletes in the Stone et al. (1997) experiment were triggering a specific subtype that shapes judgment, or if a global stereotype of black men was being ap- plied to athletes, just the same as it would be applied to all other black men. Their experiment did not ask participants to judge black athletes versus non-athletes, so it is unclear if the effect is due to an athlete subtype or to the global stereotype simply being assessed in an athletic context. Our experiment sought to illustrate the unique effects of the stereotype of black athletes, with black athletes being judged differently than white athletes and black non-athletes (salesman) when the exact same behavior is observed. It asks whether the triggering of the ste- reotype of black athletes leads to a specific subtype being activated or does it instead lead to the activation of the global stereotype of black men.

3.3. Aversive racism

Implicit bias as defined above can appear even in a person with egalitarian goals (e.g., Pearson, et al., 2009). Such a person may control stereotyping when the context signals to the person that bias may exist or that a category such as race or gender may be an influential force in one's reaction (or how others interpret one's reactions). Because they truly desire to be fair, concern that one is not being fair leads such people to compensate for this shortcoming if they are aware of their bias (e.g., Monteith, 2001; Moskowitz, Gollwitzer, Wasel, & Schaal, 1999). However, when the circumstances are less clearly about race/ gender, or when one has what appears to be substantive justifications for a response that do not lean on race, bias may appear. Implicit ste- reotypes that co-exist with one's egalitarian goals are not dampened because one feels the response is not racialized, but justified (e.g., Dovidio & Fiske, 2012; Pearson et al., 2009). This is known as aversive racism.

Aversive racism makes assessing bias difficult – one cannot ask di- rectly if a person is biased. First, because this form of bias is invisible to the person who is biased and they could not accurately tell you. A second reason researchers should not directly ask if a person has racial (or any type of) bias is because it would signal that the evaluation being asked for is racially charged. Under such circumstances an aversive racist will fear being labelled as biased, their egalitarian goals will be triggered to compensate for those fears, and they will report being fair, not being biased. Indeed, research on aversive racism shows that such concerns about being seen as biased not only occur if you directly ask people about their biases, but if they are in a situation where they feel their bias could be revealed – when something in the situation other than a direct question makes this concern salient. For example, if white voters are asked if they will vote for a white candidate or a black

candidate they are more likely to publicly say the black candidate than their actual private voting behavior would justify. But if the choice is less clearly about race, such as asking if their votes are in support of a policy issue that a black candidate opposes, they now will freely state their opposition to the black candidate and attribute their response to the policy issue, not race (e.g., Pearson et al., 2009). Some types of judgment of a stereotyped group similarly trigger concerns that what is really being assessed is bias. When Darley and Gross (1983) asked participants to judge the intelligence of a rich child versus a poor child they showed no difference – they controlled their bias to see poor children as less intelligent because all they were told about the child was her socioeconomic status, and they knew one should not judge others on such factors alone. However, a bias to judge poor children as less intelligent did emerge as soon as participants were given a shred of additional information about the child (that revealed nothing about her intelligence) that they could then use to justify their response. When aversive racists see an assessment of their beliefs/attitudes as an op- portunity to look biased, they control that bias and are careful not to respond negatively (e.g., Monteith & Voils, 2001; Moskowitz et al., 1999).

In the current experiment we explore this issue – participants will make judgments about people both when the judgment is seemingly unable to reveal a bias, and when the judgment has the potential to reveal them to be biased. Stereotypes should impact their judgments in the former case, but they should control their bias in the latter case, as the theory of aversive racism predicts. In our task participants read quotes and judged each author of the quote on a 7-point scale that ranged from a negative characteristic as the explanation for the quote, to a positive characteristic. We then manipulated the scales so that the characteristics that anchored the high and low end were either qualities that signaled to the perceiver that racial bias could be revealed by their answer, or failed to signal such concerns. This “signaling” was done in two ways. First, the negative items that signaled bias were more ex- treme in their valence. Second, these negative items were harder to justify because their negativity was really all that separated them from the item it was contrasted with at the other end of the scale. Much like participants who would not call a poor girl “stupid” or voters who would not claim to lack support for a black candidate simply because he was black, participants should control bias to a black athlete when the option they are asked to select is extremely negative and has no other justification for being selected. However, if the negative quality offered is both less extreme, and seems justifiable as a choice (as voting based on “policy” not race is justifiable), then bias will emerge.

Let us provide a concrete example from our materials: Judging whether a behavior is confident versus arrogant is essentially asking about the same quality – a person's belief in his/her superiority over others in a given domain. The decision to label that person as confident versus arrogant is largely a decision based on valence, or negativity. And arrogance is quite an extreme negative quality compared to the extremity of calling someone confident, and compared to the other negative qualities asked about with other quotes presented in filler trials. An aversive racist might be reluctant to judge a black athlete as arrogant when such cues signal race is relevant. In contrast, judging whether the behavior of a black athlete who repeatedly attempts the same task is “persistent” or “stubborn” does not offer participants an extreme negative quality that might invoke control attempts. Additionally, it avoids invoking race by pointing to a justification that labels the negative option not merely negative, but a wholly different form of explanation. Labelling the person as stubborn and not persis- tent, allows one to say the person is not working hard without having to call them lazy and lacking a work ethic. It provides a semantically se- parate alternative one can use as a non-race explanation for calling the person not hard working. And a somewhat moderate negative ex- planation at that. The choice can be rationalized by the different se- mantic content. Thus, a bias to express the stereotype is free to emerge, when, as the tenets of aversive racism theory suggest, there is a

G.B. Moskowitz, D. Carter Psychology of Sport & Exercise 36 (2018) 139–146

142

justifiable reason for making that choice that does not seem to invoke race. A choice between persistent versus stubborn as an explanation for behavior invokes a difference not only in valence, but in the semantic meaning of the contrasting items – one means closed-minded, the other hard working. If the choices were hard working versus lazy, then, as with confidence and arrogance, the choice might raise participants' awareness of the dangers of seeming racially biased. But persistent and stubborn afford one an opportunity to respond in ways that do not appear to be racially motivated.

Thus, we expect that when judging black athletes as arrogant versus confident, participants will attempt to control the expression of bias since the question seems on its face to be racialized. Yet when judging the black athletes as persistent versus stubborn, bias will emerge. These effects will not emerge when judging white athletes, since these ste- reotypes do not exist for this group and will not be triggered. But we further predict that these effects will only emerge when judging black athletes, not black men more generally. Once again, these stereotypes are specific to the subtype of the black athlete, not black men more generally, and thus black men of other occupations (e.g., salesmen, professors, architects) would not be rated differently from white men on these particular rating scales. Only black athletes, with the stereotype that their natural ability makes them less hard working and more ar- rogant (Devine & Baker, 1991; Sailes, 2000; Sheldon, Jayaratne, & Petty, 2007), would be judged differently.

To explore these questions we borrow from two old lines of work in social judgment and person perception. Lorge (1936) gave participants quotes such as these: “I hold it that a little rebellion, now and then, is a good thing, and as necessary in the political world as storms are in the physical.” Participants were told either that the quotes came from someone in a group they liked and belonged to (Jefferson) or to a group they disliked and did not belong to (Lenin). Agreement with the quotes was shown to be greater if you believed it was said by a liked group- member. Asch (1946) questioned whether these effects reveal an effect of valence (liking the quote better when said by someone you like) versus a change in meaning (the quote means something different when said by someone you like – implying bloody revolution from Lenin, but interpreted as an honest questioning of corruption when attributed to Jefferson). In our experiment we similarly provided quotes and ma- nipulated to whom they were attributed. Our concern was with how people judge the men when the quotes are attributed to a black athlete. But we also heed Asch's concern with (and that raised by aversive ra- cism) distinguishing among 1) interpretation of the quote being driven by valence and a global stereotype, versus 2) interpretation of the quote being driven by semantic meaning and differing interpretations. Thus, rather than assessing agreement with the quotes, we had participants judge the qualities implied by the quotes, pitting a negative inter- pretation against a positive (arrogant versus confident), but also pitting one possible meaning of the quote against another (stubborn versus persistent).

To summarize, our prediction is that the subtype of the black ath- lete, not the global stereotype, will guide judgment. This impact will be seen in two ways: 1) the trait of lacking persistence/hard work will be judged to be more prevalent in black athletes because the semantically distinct opposing anchor of stubbornness will mask the bias, 2) the trait of being arrogant, being more unjustifiable as a choice when pitted against confidence, will trigger control of the subtype (avoiding ap- pearing negative) so that people actually rate black athletes as less arrogant than white athletes. This is especially likely to occur among people with chronic concerns about not being seen by others as having prejudice.

4. Method

4.1. Overview

Participants read 120 quotes, one at a time, attributed to white and

black men of ten different occupations. On the 24 critical trials of the experiment the quotes were pertinent to the trait dimensions of arro- gant-confident and persistent-stubborn, and the occupations of the men were either athlete or salesman. The experiment had a Race (white, black) by Occupation (athlete, salesman) by Trait Type (arrogant-con- fident, stubborn-persistent) within subjects design. Our hypothesis is that race and occupation will interact to shape perception such that a black athlete will be seen as less hard working (more stubborn) than a black salesman and a white athlete or white salesman. Despite a similar stereotype of the black athlete as arrogant, we expect that an interac- tion will not emerge that reveals black athletes will be rated as more arrogant. Participants will exert control over expressing such an ex- tremely negative and less justifiable (due to its semantic similarity to confidence) view of black athletes.

4.2. Participants

Amazon Mechanical Turk was used to recruit participants to a survey said to last approximately 15 min. Participants were recruited from the pool of participants who had performed Experiment one who had provided responses in that experiment that indicated they were familiar with the aspects of the stereotype of black athletes being ex- amined Experiment 2 – black athletes as arrogant and black athletes as lacking work ethic. They were not aware that their ratings on these items in Experiment 1 were the basis for their inclusion in participation in Experiment two. The participants in Experiment two were 152 English speaking residents of the United States who completed the survey in exchange for fifty cents in compensation.

4.3. Materials

Quotes. There were 120 quotes constructed that were each attrib- uted to a single individual whose occupation was specified prior to the quote. Ninety six of these quotes were filler items, which consisted of 12 quotes per occupation attributed to eight different occupations. Across 24 critical trials, each participant read 12 quotes attributed to two occupations.

Race. Each quote was preceded by a photograph of the man said to have been responsible for the quote that followed it. Across the critical trials race was manipulated so that 12 trials contained faces of white men and 12 trials contained faces of black men. On filler trials the faces were all white men, so that the ratio of white men to black men in our procedure was similar to the ratio in the United States population (in an attempt to not make race a salient factor of the experiment to the participants).

Occupation Manipulation. The occupations designated in the critical trials were “athlete” and “salesman”. These were selected because athlete is the specific subtype of African Americans we wished to ex- plore, and salesman was considered a control condition with which no specific subtype was associated.

Trait Type. The critical trials of the experiment asked participants about two traits associated with the subtype of African Americans – arrogance and lack of hard work. There were 12 quotes that were re- lated to each of these trait dimensions, each rated on a scale that cap- tured the ambiguity in interpreting the quotes; the scales offered a positive and a negative interpretation of each quote (the 96 filler trials offered similar opposing valence rating scales for the irrelevant traits implied in those quotes). One scale ranged from arrogant (a score of 1) to confident (a score of 7), while the other scale ranged from stubborn (a score of 1) to persistent (a score of 7). These two dimensions were se- lected not merely to provide two tests of the impact of the stereotype on judgment, but because the opposing interpretations built into each scale differ in a significant way relevant to aversive racism. Arrogant versus confident represents one trait dimension, differing only in the valence of the interpretation – each asks about the strong belief in one's own efficacy/ability but either possessing such a belief in a positive way

G.B. Moskowitz, D. Carter Psychology of Sport & Exercise 36 (2018) 139–146

143

(confident) or an extremely negative way (arrogance). Stubborn versus persistent represent not only opposing valence interpretations of a be- havior, but two different trait dimensions. Stubborn suggests one is refusing to yield because one is unwilling to listen, too narrow of focus, unreasonably obstinate, and difficult to move. Persistence similarly implies a refusal to yield, but because one is persevering, hard-working, and unbroken. Thus, arrogance-confidence is a rating simply about valence, whereas stubborn-persistent is a rating both about valence but also about interpretation/meaning of the behavior.

Experiment Version. Participants were randomly assigned to one of four different versions of the experiment. The versions different only in that they provided different pairings of faces with quotes. This was done to rule out the possibility that some subset of faces might (ran- domly) have features that strongly communicated arrogance or con- fidence, or strongly communicated persistence or stubbornness.

Demographic questions and Motivation to be Egalitarian. Participants had been selected to this experiment based on knowledge of the stereotype as assessed in Experiment 1. That earlier experiment also assessed a series of demographic questions concerning participant age, race, gender, and experience with athletics (professional, recrea- tional, etc.). They also answered four questions that assessed their strength and type of motivations for being egalitarian to black men (see Experiment 1).

Design. The study had three within subject factors (Race, Occupation, and Trait Type) and one between subject factor (Version). The Race (black, white) by Occupation (athlete, salesman) by Trait Type (arrogant/confident, stubborn/persistent) by Experiment version design presented to each participant: 3 quotes relevant to the trait di- mension of arrogant-confident attributed to a black athlete, 3 quotes relevant to the trait dimension of stubborn-persistent attributed to a black athlete, 3 quotes relevant to arrogant-confident attributed to a white athlete, 3 quotes relevant to stubborn-persistent attributed to a white athlete, 3 quotes relevant to arrogant-confident attributed to a black salesman, 3 quotes relevant to stubborn-persistent attributed to a black salesman, 3 quotes relevant to arrogant-confident attributed to a white salesman, 3 quotes relevant to stubborn-persistent attributed to a white salesman. Each version had different faces of white and black men assigned to each of the 3 quotes of each category.

5. Results

A 4 (Experiment Version: version 1, version 2, version 3, version 4) x 2 (Race: black, white) x 2 (Trait Type: arrogant/confident, stubborn/ persistent) x 2 (Occupation: athlete, salesman) mixed Analysis of Variance (ANOVA) was conducted (with repeated measures on the last three factors). Ratings of the quotes that were presented on the critical trials served as the dependent variable. As expected, there was no effect of the sole between-subjects variable, Experiment Version, so this variable is not mentioned further. There were main effects observed for Race, Occupation, and Trait Type that were qualified by the predicted three-way interaction, F (1,151) = 15.35, p < .001. To explore the nature of this three-way interaction, a Race by Occupation repeated measures ANOVA was conducted for each of the two trait dimensions.

Examining ratings of the person as being stubborn/persistent we see a significant interaction of Race and Occupation, F (1,151) = 13.54, p < .001.4 Exploring the nature of this interaction, paired comparisons reveal that black athletes are seen as less persistent than all three other groups, despite the fact the evidence on which those ratings are based

was identical across the conditions. Black athletes (M = 3.76) are seen as less persistent than white athletes (M = 4.26), t (1,151) = −4.1, p < .001. They are also seen as less persistent than black salesmen (M = 4.38), t (1,151) = −4.2, p < .001. When given the opportunity to judge the target person's description of his own unwavering and perseverant behavior, perceivers see a mild example of persistence and hard work when judging most types of targets. However, when the target person is a black athlete the ratings switch to below the scale midpoint, indicating that perceivers prefer to interpret this behavior in a different way – as a moderate form of stubbornness.

Examining ratings of the person as being arrogant/confident we see a similar significant interaction of Race and Occupation, F (1,151) = 4.13, p < .05. However, the nature of the interaction is re- versed. No longer do people perceive the black athlete in the more negative light. On this trait dimension, where valence is the only dif- ference between the two endpoints of the rating scale, participants are choosing the more positive interpretation of the black athlete. Exploring the nature of this interaction, paired comparisons reveal that black athletes are seen as less arrogant than all three other groups, despite the fact the evidence on which those ratings are based was identical across the conditions. Black athletes (M = 4.07) are rated as more confident than white athletes (M = 3.33), t (1,151) = 5.5, p < .001. They are also rated as more confident than black salesmen (M = 3.79), t (1,151) = 2.0, p < .05. When given the opportunity to judge the target person's description of his superior abilities, the per- ceiver sees a mild example of arrogance when judging most types of targets. However, when the target person is a black athlete the same behavior is seen as moderately confident. They prefer to interpret this behavior in a different way, and in a fashion that is opposite to the stereotype that they are known to possess.

If this is indeed a form of control, the pattern should only be ob- served among people who desire to be egalitarian. Individuals who have no desire to be egalitarian should have no issue rating the black athlete as arrogant (e.g., Devine, 1989; Monteith & Voils, 2001. To explore this question, people with scores lower than 4 when rating the question “I attempt to act in non-prejudice ways toward Black people because it is personally important to me” were examined in the same 2X2 ANOVA for ratings of the person as being arrogant/confident. Al- though there were only 15 such “overtly racist” participants in the sample, the pattern of results reverses. The black athlete is still seen as different from the black salesman, but for these participants they label the black athlete as more arrogant (M = 3.5) than black salesmen (M = 4.2), though the paired comparison examining this difference does not rise to statistical significance due to the low power, t (1,14) = −1.63, p = .13. The stereotype-consistent response of seeing the black athlete as arrogant (relative to other black men) is not being controlled by these individuals.

6. Discussion

Black athletes were judged in a fashion different from all other targets on two rating scales that assessed stereotype-relevant attributes of black athletes (but that were not stereotypical attributes of black men more generally). Black athletes were seen to be less hard working and more stubborn, and, interestingly, were also rated to be more confident than arrogant (interesting because confident is the evaluative opposite to arrogance, the quality suggested by the stereotype). It is important to note that ratings of black salesman did not differ from ratings of white men, indicating that the stereotype being triggered was not the global stereotype of black men, or general feelings of prejudice (negativity) toward black men that extended to any evaluation. The effects are specific to black athletes and to traits that are relevant to the stereotype of black athletes. The exact same quote is interpreted differently when uttered by a black athlete, and he is thus seen to be a different type of person than other black (and white) men.

The results also suggest that people do not desire to have such

4 The 152 participants included 10 African Americans, and if the analyses are con- ducted with these individuals removed (so that only members of another ethnic group are used as the participants who are judging the target people in the experiment) the pattern of findings is unchanged and is even slightly stronger. Examining ratings of the person as being stubborn/persistent we see a significant interaction of Race and Occupation, F (1,141) = 13.31, p < .001. This is also true for ratings of arrogant/confident where we see a similar significant interaction of Race and Occupation, F (1,141) = 4.68, p < .04.

G.B. Moskowitz, D. Carter Psychology of Sport & Exercise 36 (2018) 139–146

144

biases. We specifically selected people who were aware of the stereo- types being assessed, and thus should be capable of having those ste- reotypes triggered by the presence of a black athlete. The fact that participants know the stereotype of black athletes as arrogant, yet they rate the black athletes in a fashion that is less arrogant and more con- fident, suggests that participants are trying to assert their egalitarian goals. They perhaps recognize the quotes are an opportunity for one to be seen as racist, particularly given that the rating they are being asked to make is to identify how positive (confident) or negative (arrogant) the person is presumed to be along one trait dimension (how to describe his belief in his abilities). Once aware the judgment is about race, and provided an opportunity for bias to be revealed, their egalitarian goals are triggered and they engage in attempts at control of the bias (e.g., Monteith & Voils, 2001; Moskowitz et al., 1999; Pearson et al., 2009). One might call into question this interpretation along two lines, each of which we believe can be ruled out. First, one might wonder if the quotes simply are not arrogant enough to judge a black athlete as ar- rogant. However, white athletes and black salesmen making the same utterances are seen as arrogant. The behavior is diagnostic enough, participants just seem reluctant to use the label “arrogant” to describe the behavior when it is performed by black athletes. Second, one might wonder if the attribute of confidence is also part of the stereotype, perhaps even a more strongly held attribute to describe black athletes as arrogant. If so, the effect seen is not participants controlling their stereotype, but simply using the stereotype in the same manner they use their stereotype of lacking hard work. The stereotype is simply one of confidence with a hint of arrogance, not excessive arrogance. This in- terpretation also seems unlikely due to our examination of the parti- cipants who do not have high external motivation to control their ste- reotypes. These participants freely express the bias they are feeling without worry as to how it would look and without trying to control it. And these participants clearly rate the black athletes as the most ar- rogant from among all the target men who utter these quotes that imply arrogance. This suggests that the stereotype is not one of confidence more than arrogance, but that arrogance is indeed triggered and most participants then try to control expressing such sentiments when they are concerned about coming across as racist.

Such attempts at control, where one suspects one's judgment has been biased and then attempts to “de-bias” or correct/adjust that judgment, often lead to what is called a contrast effect (e.g., Moskowitz & Skurnik, 1999; Wegener & Petty, 1995). Contrast effects occur where one attempts to remove the influence of some biasing agent and the result is to produce a bias in the opposite direction. For example, having one's own child in a class may lead a teacher to suspect he/she is biased to show favoritism, so in an attempt to not play favorites, the teacher inadvertently treats the child more harshly than other students in the attempt to treat the child equally. In trying to be less favorable he/she goes too far. This same process happens with stereotypic judgments – if one suspects one has been unfairly harsh in judging a person as arrogant due to a negative stereotype that one fears is serving as a biasing agent, one attempts to remove that influence (e.g., Wegener & Petty, 1995). In doing so one may see the person as less arrogant and more confident as a result of the attempt to be fair.

However, despite this optimistic possibility that participants are constantly engaged in control of their stereotypes, the theory of aver- sive racism (e.g., for a review see Pearson et al., 2009) posits that such control only occurs when one is aware of the possibility of bias. When the task is not known to be about race (see also Devine, 1989), or when a justification for one's response can be made that is not attributable to race, then a bias to judge and treat people in a stereotypic way will emerge. For example, in political elections one can point to the policies of the disliked candidate, not her gender, as the cause for a negative opinion; when selecting employees to work on a project one can point to their objective skill sets and not their race as the cause for selection (when systematic evaluation reveals that, in these two examples, po- licies and skill sets are not the dominant factor – judgments are driven

by gender and race respectively). And when judging a black athlete's espoused refusal to change course, one can point to the obstinance and stubbornness one believes is evident in the behavior, not race, as the justification for the negative ratings. Obstinance versus hard work are not only different in valence but offer two interpretations for the be- havior, and this gives one the freedom to attribute the rating to the unique semantic meaning, not dislike related to race. The impact of the stereotype is observed when the judgment is not known to implicate race and can be justified. An aversive racist may not worry to choose the negative valence item (stubborn), even if concerned the judgment is about race, because they can point to the semantic content as the basis for their choice, not the valence. Their stereotype that black athletes are “less hard working” can be displayed freely because there is an ap- parent justification, one that is seemingly not racist, readily available on which they can lean.

Implicit bias was shown here in participants favoring a stereotypic explanation for behavior over a nonstereotypic explanation. However, bias can also manifest as the favoring of one stereotype over another stereotype. People belong to multiple groups, each with a unique ste- reotype (e.g., Macrae et al., 1995). Activation of one of these group stereotypes can inhibit the stereotypes associated with other groups to which one belongs. For example, if a perceiver judging a black woman has the stereotype of “woman” triggered, the “black” stereotype is in- hibited. A perceiver is more likely to judge her in a fashion consistent with the stereotype of women, but not of African Americans. We are biased not merely to use stereotypes, but to use only one stereotype at a time as a guide in judgment. One might ask whether triggering a sub- type has a similar inhibitory effect on the global stereotype, just as when one stereotype interferes with the stereotype of another group to which a person belongs. This would be an important question to address for future research – what is the relationship between the subtype and the global stereotype if they suggest different interpretations of a target person?

7. Conclusion

Subtypes of a category or stereotype often develop as a way to protect a stereotype from change; as a way of “fencing off” the excep- tions to the rule so that the category/stereotype can remain intact (Kunda & Oleson, 1995). Often this serves to keep a negative stereotype from being revised or updated by placing the positive members of the group into a separate group with a set of positive characteristics that distinguish them from other “more typical” members of the larger stereotypic category (Kunda & Oleson, 1995). In the case of the black athlete, the subtype appears to emerge in a way that takes an ostensibly positive quality, athleticism, and morphs it into a negative set of qua- lities, constructing a narrative around the downside of having natural abilities. Natural athletic ability is treated in the subtype as a force that can demotivate the nurturing of intelligence and hard work; one has the ability to compensate for such “tiresome efforts” with one's natural gifts. As such, a form of implicit bias is made possible whereby a black athlete is seen not positively for being athletic, but as arrogant, lacking work ethic, and lacking sport's intelligence. Such bias can be controlled if one is aware that one is being influenced (and is the type of person that desires to be fair and not be biased). But in the many cases where one believes one is not and cannot be biased, ironically, the implicit stereotype will influence judgment and cause one to believe one sees evidence that confirms the expectation put forward by the subtype.

References

Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley. Asch, S. E. (1946). Forming impressions of personality. Journal of Abnormal and Social

Psychology, 41, 258–290. Barden, J., Maddux, W. W., Petty, R. E., & Brewer, M. B. (2004). Contextual moderation of

racial bias: The impact of social roles on controlled and automatically activated at- titudes. Journal of Personality and Social Psychology, 87(1), 5.

G.B. Moskowitz, D. Carter Psychology of Sport & Exercise 36 (2018) 139–146

145

Beilock, S. L., & McConnell, A. R. (2004). Stereotype threat and sport: Can athletic per- formance be threatened? Journal of Sport & Exercise Psychology, 26(4), 597–609.

Blair, C. (2002). School readiness: Integrating cognition and emotion in a neurobiological conceptualization of children's functioning at school entry. American Psychologist, 57(2), 111.

Correll, J., Park, B., Judd, C. M., Wittenbrink, B., Sadler, M. S., & Keesee, T. (2007). Across the thin blue line: Police officers and racial bias in the decision to shoot. Journal of Personality and Social Psychology, 92(6), 1006.

Czopp, A. M. (2010). Studying is lame when he got game: Racial stereotypes and the discouragement of Black student-athletes from schoolwork. Social Psychology of Education, 13(4), 485–498.

Darley, J. M., & Gross, P. H. (1983). A hypothesis-confirming bias in labelling effects. Journal of Personality and Social Psychology, 44, 20–33.

Darley, J. M., & Fazio, R. H. (1980). Expectancy confirmation processes arising in the social interaction sequence. American Psychologist, 35, 867–881.

Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled com- ponents. Journal of Personality and Social Psychology, 56, 5–18.

Devine, P. G., & Baker, S. M. (1991). Measurement of racial stereotype subtyping. Personality and Social Psychology Bulletin, 17(1), 44–50.

Devine, P. G., & Elliot, A. J. (1995). Are racial stereotypes really fading? The Princeton trilogy revisited. Personality and Social Psychology Bulletin, 21(11), 1139–1150.

Dovidio, J. F., & Fiske, S. T. (2012). Under the radar: How unexamined biases in decision- making processes in clinical interactions can contribute to health care disparities. American Journal of Public Health, 102(5), 945–952.

Duncan, B. L. (1976). Differential perception and attribution of intergroup violence: Testing the lower limits of stereotyping Blacks. Journal of Personality and Social Psychology, 34, 590–598.

Eastman, S. T., & Billings, A. C. (2001). Biased voices of Sports: Racial and gender ste- reotyping in college basketball announcing. Howard Journal of Communications, 12, 183–201.

Eberhardt, J. L., Davies, P. G., Purdie-Vaughns, V. J., & Johnson, S. L. (2006). Looking deathworthy: Perceived stereotypicality of Black defendants predicts capital-senten- cing outcomes. Psychological Science, 17(5), 383–386.

Edwards, H. (1973). The black athletes: 20th century gladiators for white America. Psychology Today, 7(6), 43–52.

Edwards, H. (1984). The Black ‘dumb jock’: An American sports tragedy. The College Board Review, 131, 8–13.

Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview of the cognitive approach. In J. F. Dovidio, & S. L. Gaertner (Eds.). Prejudice, discrimination, and racism (pp. 127–163). Orlando, FL: Academic Press.

Harrison, K. (1998). The assassination of the Black male image in sport. Journal of African American Men, 3, 45–56.

Katz, D., & Braly, K. W. (1933). Racial stereotypes of 100 college students. Journal of Abnormal and Social Psychology, 28, 280–290.

Higgins, E. T. (1996). Knowledge activation: Accessibility, applicability, and salience. In E. T. Higgins, & A. W. Kruglanski (Eds.). Social psychology: Handbook of basic principles (pp. 133–168). New York: Guilford.

Kunda, Z. (1987). Motivated inference: Self-serving generation and evaluation of causal theories. Journal of Personality and Social Psychology, 53(4), 636–647.

Kunda, Z., & Oleson, K. C. (1995). Maintaining stereotypes in the face of disconfirmation: Constructing grounds for subtyping deviants. Journal of Personality and Social Psychology, 68, 565–579.

Lorge, I., & Curtiss, C. C. (1936). Prestige, suggestion, and attitudes. The Journal of Social Psychology, 7.4, 386–402.

Macrae, C. N., Bodenhausen, G. V., & Milne, A. B. (1995). The dissection of selection in person perception: Inhibitory processes in social stereotyping. Journal of Personality and Social Psychology, 69(3), 397.

Monteith, M. J., & Voils, C. I. (2001). Exerting control over prejudiced responses. In G. B. Moskowitz (Ed.). Cognitive social psychology: The Princeton symposium on the legacy and future of social cognition. NJ. Lawrence Erlbaum Associates, Inc.

Moskowitz, G. B., Gollwitzer, P. M., Wasel, W., & Schaal, B. (1999). Preconscious control of stereotype activation through chronic egalitarian goals. Journal of Personality and Social Psychology, 77(1), 167–184.

Moskowitz, G. B., & Olcaysoy Okten, I. (2018). Social cognition. In T. Nelson (Ed.). Getting grounded in social psychology (pp. 37–78). New York: Psychology Press/Taylor and Francis.

Moskowitz, G. B., & Skurnik, I. W. (1999). Contrast effects as determined by the type of prime: Trait versus exemplar primes initiate processing strategies that differ in how accessible constructs are used. Journal of Personality and Social Psychology, 76, 911–927.

Moss-Racusin, C. A., Dovidio, J. F., Brescoll, V. L., Graham, M. J., & Handelsman, J. (2012). Science faculty's subtle gender biases favor male students. Proceedings of the National Academy of Sciences, 109(41), 16474–16479.

Nelson, T. D. (2018). Prejudice. In T. Nelson (Ed.). Getting grounded in social psychology. New York: Psychology Press/Taylor and Francis.

Nosek, B. A., Smyth, F. L., Hansen, J. J., Devos, T., Lindner, N. M., Ranganath, K. A., et al. (2007). Pervasiveness and correlates of implicit attitudes and stereotypes. European Review of Social Psychology, 18(1), 36–88.

Payne, B. K. (2006). Weapon bias: Split-second decisions and unintended stereotyping. Current Directions in Psychological Science, 15(6), 287–291.

Pearson, A. R., Dovidio, J. F., & Gaertner, S. L. (2009). The nature of contemporary prejudice: Insights from aversive racism. Social and Personality Psychology Compass, 3(3), 314–338.

Plant, E. A., & Devine, P. G. (1998). Internal and external motivation to respond without prejudice. Journal of Personality and Social Psychology, 75(3), 811–832.

Price, J., & Wolfers, J. (2007). Racial discrimination among NBA referees (No. w13206). National Bureau of Economic Research.

Rada, J. (1996). Color blind-sided: Racial bias in network television's coverage of pro- fessional football games. The Howard Journal of Communications, 7, 231–240.

Rainville, R. E., & McCormick, E. (1977). Extent of covert racial prejudice in pro football announcers' speech. Journalism Quarterly, 54, 20–26.

Sailes, G. A. (2000). Betting against the Odds: Professional sports aspiration among African American males. Vol. 10. Amer Alliance for Health Physical.

Schulman, K. A., Berlin, J. A., Harless, W., et al. (1999). The effect of race and sex on physicians' recommendations for cardiac catheterisation. New England Journal of Medicine, 340, 618–626.

Sheldon, J. P., Jayaratne, T. E., & Petty, E. M. (2007). White Americans' genetic ex- planations for a perceived race difference in athleticism: The relation to prejudice toward and stereotyping of Blacks. Athletic Insight: Online Journal of Sport Psychology, 9(3), 31–56.

Staples, R., & Jones, T. (1985). Culture, ideology and black television images. The Black Scholar, 16(3), 10–20.

Stone, J., Harrison, C. K., & Mottley, J. (2012). “Don't call me a student-athlete”: The effect of identity priming on stereotype threat for academically engaged African American college athletes. Basic and Applied Social Psychology, 34(2), 99–106.

Stone, J., Lynch, C. I., Sjomeling, M., & Darley, J. M. (1999). Stereotype threat effects on Black and White athletic performance. Journal of Personality and Social Psychology, 77, 1213–1227.

Stone, J., Perry, Z. W., & Darley, J. M. (1997). “White men can't jump”: Evidence for the perceptual confirmation of racial stereotypes following a basketball game. Basic and Applied Social Psychology, 19, 291–306.

Wegener, D. T., & Petty, R. E. (1995). Flexible correction processes in social judgment: The role of naive theories in corrections for bias. Journal of Personality and Social Psychology, 68, 36–51.

G.B. Moskowitz, D. Carter Psychology of Sport & Exercise 36 (2018) 139–146

146