2017-2018 Doctoral Dissertation Awardee: Joseph Vitriol

“The (In)Egalitarian Self: On the Motivated Rejection of Alleged Implicit Racial Bias”


Most Americans reject hostile forms of prejudice and are motivated to eliminate racial bias in their own judgment, preferring to view themselves as egalitarian and society as fair (Henry & Kosterman, 2000). Yet significant racial disparities persist across a broad range of contexts (Blank, 2001).

Considerable evidence has established the existence of racially biased implicit attitudes, or unconsciously held and automatically activated evaluations, which have been offered as an explanation for consequential racial disparities in many important life domains (e.g., Borgida & Fisher, 2012). Individuals, organizations, and institutions committed to antidiscrimination principles would be remiss to ignore the empirical evidence on implicit racial bias and its relation to discriminatory behaviors (Jost et al., 2009). Several promising empirical investigations have identified strategies that can reduce implicit biases and their consequences (Lai, Hoffman, & Nosek, 2013). These interventions, and theoretical models of prejudiceregulation in general, share the assumption that increasing awareness of implicit racial bias is an effective strategy for reducing its impact on judgment and behavior (e.g., Monteith & Mark, 2009). Providing individuals with accurate information about their implicit racial biases can be valuable for prejudice-reduction interventions and prejudice-regulation.

While policymakers, industry leaders, and organizations invest substantial resources on prejudice-interventions (Department of Justice, 2016) many of these proposed remedies lack empirical evidence (Paluck & Green, 2009) or fail to induce stable, long-term change in attitudes and behavior (Forscher et al., 2016). One potential explanation for the inefficacy of many antiprejudice interventions concerns the relative difficulty of increasing awareness of the existence of unconscious forms of racial bias (Perry et al., 2015) and the challenge of helping individuals to implement strategies to regulate implicit biases over time (Lai et al., 2015). Thus, for anti-bias interventions to be effective they must succeed in persuading potentially skeptical individuals that they are in possession of socially undesirable and stigmatizing attitudes that can influence their actions, despite their commitment to egalitarian values (Hillard, Ryan & Gervais, 2013).

Herein lies the problem. Few social labels are more aversive than being categorized as a racist (Crandall, Eshelman, & O’Brien, 2002). Many White Americans, including self-identified egalitarians, are fundamentally threatened by and avoidant of situations that may lead to attributions of racist intent or challenge their egalitarian self-image (Dovidio & Gaertner, 2000; Frantz et al., 2004). Consequently, defensiveness may be a common reaction to anti-bias interventions—especially when individuals encounter feedback suggesting that their unconscious biases can influence their behavior, without their knowledge, and in ways that are inconsistent with their explicit beliefs (Kaiser & Miller, 2001).

However, prior work had neglected to study the psychological obstacles to the kind of awareness necessary for the effective implementation of anti-bias interventions or how to overcome it. Given the central importance of bias awareness for prejudice-regulation (Perry, Murphy, & Dovidio, 2015), strategies to reduce defensive responding is of clear importance. Across four experiments, my dissertation systematically examined the psychological responses to implicit bias feedback and tested theory-based strategies that were designed to decrease defensive responding and increase bias awareness and prejudice-regulation.


Study 1 and 2 addressed the psychological responses induced by implicit bias feedback. I focused on two mediating processes—guilt and defensiveness–leading from feedback to bias awareness. Prominent theoretical models of prejudice-regulation hold that when individuals feel guilty about prejudiced-responding they are able to control its expression through conscious commitment to egalitarianism (e.g., Moskowitz et al., 1999). However, recent research suggests that many White Americans are motivated to reject implicit bias feedback (Howell et al., 2013). Consistent with these perspectives, guilt was expected to increase, whereas defensive responding was expected to decrease, bias awareness.

To test these hypotheses, I experimentally manipulated Amazon MTurk participants’ exposure to implicit bias feedback using a false-feedback paradigm following completion of the Implicit Association Test (Greenwald, McGhee, & Schwartz 1998). Study 1 (n=368) employed a single independent variable design (IAT Feedback: Bias, No Bias, No Feedback). A similar design was adopted for Study 2 (n=444; IAT Feedback: Bias, Accurate, No feedback), however, some participants received accurate feedback about their implicit attitudes. Before completing the IAT, participants completed measures of racial prejudice and egalitarian motivations. After completing the IAT and receiving feedback, participants completed measures of affect, defensive responding, bias awareness, prejudice-regulation, and public policy attitudes.

Study 3 and 4 were designed to test the effectiveness of pre-feedback interventions in attenuating defensive responding and, as a result, increasing bias awareness and prejudice-regulation. Two interventions, administered prior to feedback, were tested on two samples. Both studies employed a 2 (IAT Feedback: Bias, Control) x 3 (Intervention: self-affirmation, collective bias, control) design, and adopted the same methodology as Study 1 and 2, with intervention conditions administered prior to feedback.

The first intervention (“self-affirmation”) was intended to mitigate the threat of the feedback by bolstering the integrity of the self before implicit bias feedback. Affirming the global adequacy of the self, prior to encountering self-threatening information, can reduce defensiveness and biased responding (e.g., Sherman & Cohen, 2002; Steele, 1988). In intergroup contexts, for example, self-affirmation has been found to reduce the perceived threat of being perceived as racist (Frantz et al., 2004). Thus, in the “self-affirmation” condition, participants were instructed to affirm a valued self-identity by writing about a personally important value (e.g., Cohen et al., 2000).

The second intervention (“collective bias”) provided participants information that was intended to promote egalitarian goal pursuits and minimize the perceived threat of feedback by (a) highlighting the potential for prejudiced-responding (e.g., Moskowitz & Li, 2011) and (b) increasing perceived efficacy in prejudice-regulation by emphasizing the malleability and controllability of implicit bias. This latter feature was included because a sense of efficacy in the pursuit of a goal can bolster goal achievement (Moskowitz, 2014). Furthermore, implicit bias feedback may be experienced as moral condemnation (e.g., O’Brien et al., 2010) or as inconsistent with assumptions about one’s objectivity (Pronin, 2007). For this reason, this intervention also characterized implicit bias as a basic feature of human cognition to diffuse the belief that one is unique, irrational, or immoral for harboring unconscious bias.


Results across both Study 1 and 2 indicate that implicit bias feedback (whether accurate or not) simultaneously produced both guilt and defensive responding with independent and countervailing effects, such that the former increased and the latter decreased bias awareness and prejudice-regulation. Importantly, higher levels of bias awareness predicted increased intentions for prejudice-regulation, reduced endorsement of negative racial stereotypes, and more favorable attitudes towards anti-bias interventions and public policies designed to help racial minorities.

Interestingly, individuals high (vs. low) in prejudice were less likely to experience guilt, perhaps because they did not regard their bias as problematic, inaccurate, or in need of change, or because the feedback was consistent with their sense of self and verified self-knowledge (e.g., Swann, 1986). These findings are in line with existing work indicating that affective responses to feedback about a failure to have responded in an egalitarian manner varies as a function of individual differences in one’s level of prejudice toward the targeted group (e.g., Devine, 1989; Monteith et al., 1991). The results of Study 1 and 2 demonstrate that individuals low (vs. high) in prejudice are more inclined to experience guilt upon learning that they have responded with bias.

Furthermore, egalitarian motivations moderated the effect of implicit bias feedback on defensive responding. Egalitarian values can manifest in a variety of ways, with the manner of expression impacting how one responds to awareness of bias. In particular, two ways of valuing egalitarianism have been extensively examined – internal and external motivations to control prejudice (e.g., Plant & Devine, 2003). Internal motives to be egalitarian are an idealized state that one is motivated to approach (Monteith, 1993). External motives to be egalitarian are an “ought” state in which people hold standards about how they should (ought to) behave based on perceived social norms or the expectations of others. While internal (vs. external) egalitarians are more motivated to minimize the influence of bias in their judgments, confrontation can nonetheless lead them to respond defensive (Czopp et al., 2016). The results of studies 1 and 2 support this prediction.

Finally, Study 3 and 4 both provide robust evidence for the success of the “collective bias” intervention, which reliably reduced defensive responding (but not guilt) and thereby increased awareness of bias, motivations for prejudice-regulation, and support for anti-bias interventions. Furthermore, the effectiveness of the feedback intervention did not vary across levels of implicit or explicit racial prejudice, egalitarian motivations, or sociopolitical orientations. I did not observe any evidence to suggest that self-affirmation reduced defensive responding. Together, these findings indicate that implicit racial bias feedback can backfire, but, when framed in ways consistent with the “collective bias” strategy, can nonetheless be an effective approach for increasing bias awareness, prejudice-regulation, and the success of anti-bias interventions.


My dissertation research helps elucidates the psychological responses that underlie reactions to implicit bias feedback and its implications for bias awareness and prejudice-regulation. The results of Study 1 and 2 demonstrate that prejudice-regulation is facilitated by bias awareness, which is promoted by guilt. This is consistent with existing models of prejudice-regulation (e.g., Monteith, 1993; Moskowitz et al., 1999). However, prejudice-regulation is interfered with, and replaced by backlash, when people do not experience guilt and heightened awareness, but instead respond defensively. That both of these processes emerge simultaneously, and as a function of theoretically predicted individual difference variables (i.e., levels of prejudice; egalitarian motivations), is both a novel contribution to existing research and offers a parsimonious and robust framework for understanding prejudice-regulation and anti-bias interventions.

However, these effects were observed using an exaggerated feedback paradigm that characterized the results of a test of unconscious racial bias in a particularly harsh and deterministic manner. With additional research support from the Society for Personality and Social Psychology, I have replicated the findings in Study 1 and 2 on an independent sample and across a broader range of feedback conditions (i.e., extreme vs. moderate vs. slight feedback), suggesting that this model will generalize well to different kinds of anti-bias interventions and means of providing feedback.

Given the central importance of bias awareness for prejudice-regulation, it is imperative to identify approaches for anti-bias interventions that motivates egalitarian goal pursuits but does not engender defensive responding. The failure to do so is a major theoretical oversight. Studies 3 and 4 remain, to my knowledge, the only existing attempt to evaluate the effectiveness of prefeedback interventions designed to reduce defensive responding and increase bias awareness. The results provide clear support for the success of the “collective bias” intervention, and, in doing so, also advance perspectives on the motivational factors that underpin contemporary racial attitudes. That is, when individuals feel threatened by the prospect of appearing racist, morally condemned for harboring implicit bias, and that they lack the ability to change their own behavior and avoid the social disapproval of others, they may be motivated to deny evidence of prejudice and discrimination.

However, the methodology in my dissertation did not experimentally isolate different features of the intervention, so it remained unclear which factors were driving these effects or whether they must be implemented in conjunction to be effective. With research support from the Society for the Psychological Study of Social Issues, I conducted a series of follow-up studies to manipulate each feature of the intervention, and to evaluate its long-term implications for bias awareness. The effectiveness of the “collective bias” intervention was clearly demonstrated across additional samples. Importantly, reducing perceived moral blameworthiness and increasing the perceived malleability and controllability of bias are both able to reduce defensive responding, although these features appear most effective when used together. By decreasing defensiveness, this intervention not only increased bias awareness in the short term but also led to intra-individual increases in bias awareness approximately 6-months after the interventions.