The Neuroscience of Stigma and Stereotype Threat Belle Derks, Michael Inzlicht, Sonia Kang

To cite this version:

Belle Derks, Michael Inzlicht, Sonia Kang. The Neuroscience of Stigma and Stereotype Threat. Group Processes and Intergroup Relations, SAGE Publications, 2008, 11 (2), pp.163-181. ￿10.1177/1368430207088036￿. ￿hal-00571682￿

HAL Id: hal-00571682 https://hal.archives-ouvertes.fr/hal-00571682 Submitted on 1 Mar 2011

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Group Processes & Intergroup Relations 2008 Vol 11(2) 163–181

The Neuroscience of Stigma and Stereotype Threat

Belle Derks Leiden University Michael Inzlicht and Sonia Kang University of Toronto

This article reviews social neuroscience research on the experience of stigma from the target’s perspective. More specifi cally, we discuss several research programs that employ electro- encephalography, event-related potentials, or functional magnetic resonance imaging methods to examine neural correlates of stereotype and social identity threat. We present neuroimaging studies that show brain activation related to the experience of being stereotyped and ERP studies that shed light on the cognitive processes underlying social identity processes. Among these are two projects from our own lab. The fi rst project reveals the important role of the neurocognitive confl ict-detection system in stereotype threat effects, especially as it pertains to stereotype threat ‘spillover’. The second project examines the role of automatic ingroup evaluations as a neural mediator between social identity threats and compensatory ingroup bias. We conclude with a discussion of the benefi ts, limitations, and unique contributions of social neuroscience to our understanding of stigma and social identity threat. keywords EEG, ERP, social identity, social neuroscience, stereotype threat

Over the years, methodological and technolo- to important new insights into the cognitive gical advances have allowed social psychologists processes and neural mechanisms underlying to reduce their reliance on explicit self-report intergroup phenomena such as categorization, measures of people’s emotions, attitudes, and prejudice, and stereotyping. For example, the motivations. The movement within the fi eld mechanisms underlying automatic processes has been toward a focus on implicit measure- that were previously difficult or impossible ments, providing for a more accurate and un- to identify can now be studied with neural biased assessment of important psychological phenomena. The recent introduction of methods and technology from cognitive neuro- Author’s note science affords social psychologists yet another Address correspondence to Belle Derks, theoretical advancement; we are now able to Leiden University, Institute for Psychological systematically test predictions related to the Research, Social and Organizational mechanisms underlying social psychological Psychology Unit, P.O. Box 9555, 2300 RB processes. This merge of social psychology Leiden, the Netherlands and neuroscience, social neuroscience, has led [email: [email protected]]

Copyright © 2008 SAGE Publications (Los Angeles, London, New Delhi and Singapore) 11:2; 163–181; DOI: 10.1177/1368430207088036 Group Processes & Intergroup Relations 11(2) techniques like functional magnetic resonance threat. We begin by discussing the effects of pre- imaging (fMRI) and electroencephalography judice on targets, including stereotype threat (EEG). Neuroscience techniques allow us to effects and the use of coping mechanisms to explore the role of specifi c neural regions and protect social identity. Then, we describe how systems in complex social psychological phe- these phenomena can be studied with social nomena such as person perception and racial neuroscience techniques. Finally, we examine bias, thereby improving our understanding of the limitations and benefi ts of this work and these phenomena by integrating with what is discuss the value of social neuroscience in already known about these specifi c brain regions relation to more traditional measures. Rather from neuroscience research (see Ochsner & than offering a comprehensive analysis of re- Lieberman, 2001, for a primer on social cognitive search on stereotype threat (see Shapiro & neuroscience). In turn, social neuroscience Neuberg, 2007; Steele, Spencer, & Aronson, informs the fi eld of cognitive neuroscience as 2002), this article offers a glimpse into recent a whole by enriching the current body of know- neuroscience research emerging in the fi eld of ledge with new insights into the intricacies of stigma, detailing how tools borrowed from complex social behaviors and processes. neuroscience and psychophysiology can in- The past few years in particular have seen form our understanding of the target’s per- an explosion of studies using the social neuro- spective. We begin our discussion by providing science approach. As this Special Issue attests, a selective review of stereotype and social iden- work on stereotyping and prejudice is no ex- tity threat. ception. EEG and fMRI methods, for example, have been used in the area of person percep- Stereotype and social identity threat: tion to study the hypothesized relationship be- A brief review tween automatic and controlled processes (e.g. Bartholow, Dickter, & Sestir, 2006; Cunningham Stereotype threat defi ned et al., 2004). To this point, however, most work Stereotype threat has become one of the most on the social neuroscience of intergroup pro- vigorously explored topics of the past decade cesses has focused on the perpetrator’s point in social psychology, with research on the topic of view, focusing, for example, on the mental greeted with much interest within academic processes and brain regions that are activated psychology and among members of the public when people employ stereotypes to perceive and alike. The original Steele and Aronson (1995) interpret their social world (see for examples article is now considered a modern classic (Fiske, Amodio, Devine, & Harmon-Jones, 2007; 2003), and has been cited well over 600 times1. Bartholow et al., 2006; Cunningham et al., Part of the enthusiasm stems from the fact that 2004; Hart et al., 2000; Ito, Willadsen Jensen, the theory of stereotype threat offers a situ- & Correll, 2007; Phelps et al., 2000; Richeson et ational interpretation for the long-lasting and al., 2003; Wheeler & Fiske, 2005). In contrast, seemingly intractable debate over the source surprisingly little social neuroscience research of group differences in academic and cognitive has focused on the target’s point of view. Given performance. This explanation avoids the nature social neuroscience’s role in advancing research vs. nurture trap, by suggesting that situations on stereotyping and prejudice, it is clear that themselves can bring about group differences neuroscience techniques could also advance when threats are ‘in the air’ (Steele, 1997). theory and research on the target’s perspective. Stereotype threat is a situational predica- That is, social neuroscience has great potential ment where individuals are at risk, by dint in fostering a more comprehensive and inte- of their actions or behaviors, of confi rming grated understanding of what it means and negative stereotypes about their group (Steele how it feels to belong to a stigmatized group. & Aronson, 1995). It is the resulting sense that In this article, we review the existing social one might be judged or treated in terms of a neuroscience research on stigma and stereotype negative stereotype. In their original article,

164 Derks et al. neuroscience of stigma

Steele and Aronson (1995) suggested that be- stereotype threat effect to a broad array of groups cause African Americans are well aware of the and domains of activity, both in the real world negative stereotypes impugning their intel- and in the lab. Studies have observed the effect of lectual ability, whenever they are in a situation stereotype threat on performance of women on requiring them to display said ability, they fear tests of political knowledge (McGlone, Aronson, confi rming the stereotype, a process which & Kobrynowicz, 2006), of Latinos (Gonzales, could interfere with optimal performance. In Blanton, & Williams, 2002), and of children of support of their hypotheses, their experiments low socioeconomic status on tests of intelligence indicated that African American college stu- (Croizet & Claire, 1998). However, stereotype dents underperformed when stereotypes about threat does not exclusively affect members of their group became salient, for example, when traditionally devalued or marginalized groups; they took a ‘diagnostic’ test of academic ability indeed, even members of non-stigmatized groups or when they indicated their race on a demo- can experience stereotype threat. Studies have graphic survey prior to taking a test. shown stereotype threat effects with Whites on A speech given by Lawrence Summers, Harvard tests of athletic ability (e.g. Beilock, Jellison, University’s former president, in early 2005 Rydell, McConnell, & Carr, 2006; Stone, Lynch, illustrates that women, too, are exposed to Sjomeling, & Darley, 1999), tests of math ability negative stereotypes. In this speech, Summers (Aronson et al., 1999), or tests of racial bias claimed that women lack innate ability in math (Frantz, Cuddy, Burnett, Ray, & Hart, 2004), and science, and that this accounts for why so as well as with psychology students on tests of few of them are to be found—as either students general knowledge (e.g. Seibt & Förster, 2004). or professors—in math and science classrooms Finally, although there is some controversy over (see Dar-Nimrod & Heine, 2006). Like with whether stereotype threat has any impact out- African Americans, these negative beliefs hurt side of the laboratory (Cullen, Hardison, & and can lead women to feel threatened every Sackett, 2004; Stricker & Ward, 2004), recent time they perform a task that tries their math evidence suggests that it had an effect in state and science ability. In an initial set of studies, middle-school exams in Texas (Good, Aronson, & Spencer, Steele, and Quinn (1999) illustrated Inzlicht, 2003) and in elementary school grades that highly capable women underperformed in Connecticut (Cohen, Garcia, Apfel, & Master, on a math test when they thought the test could 2006). reveal ‘gender differences in ability’ compared The consequence of stereotype threat on to women who thought it was ‘gender-neutral’. performance, then, is widely documented, Importantly, others have shown that subtle with well over 100 separate studies examining manipulations can also activate stereotypes and performance defi cits (Davies, Spencer, & Steele, thus interfere with performance, for example, 2005). In contrast, very few studies have gone changing the gender composition of a group beyond performance to examine some of the (Inzlicht, Aronson, Good, & McKay, 2006; other important consequences of stereotype Inzlicht & Ben-Zeev, 2000), manipulating the threat. More recently, however, researchers gender of a test administrator (Marx & Roman, have recognized this omission and started ad- 2002), or even exposing women to sexist dressing this crucial subject. Work coming out advertisements (Davies, Spencer, Quinn, & of our own lab, as well as that of others (e.g. Gerhardstein, 2002). As an antidote to President Beilock, Rydell, & McConnell, 2007), has now Summers’s biological explanation, these studies indicated that stereotype threat can ‘spillover’ suggest that one reason women are missing and affect behaviors and decisions in domains from science is the situational predicament of unrelated to the threatening stereotype. We stereotype threat. have found, for example, that when female In the years since the publication of Steele and students and students belonging to visible Aronson’s article, many articles have followed minority groups became aware their actions and documented the generalizability of the could confi rm negative stereotypes about their

165 Group Processes & Intergroup Relations 11(2) gender and ethnic groups, they had a more biased treatment. This is a broader conception diffi cult time maintaining their attention, phy- of threat because it includes any situation that sically exerting themselves, suppressing their contains the possibility of being marginalized, appetites, and making rational judgments and including not only when stereotypes are ‘in the decisions (Inzlicht, Kang, & Fortune, 2008; air’, but also when cues in the environment Inzlicht, McKay, & Aronson, 2006). Other work send messages that one’s social identity is not has concentrated on the effects of stereotype welcome. threat on the self-concept. Some research has In any given setting, a person can come to suggested that stereotype threat can lead to realize that they could be devalued and discrim- self-uncertainty about one’s abilities and one’s inated against because of their social identity. belonging in a specific setting (Aronson & This awareness could originate from a person’s Inzlicht, 2004; Walton & Cohen, 2007), reducing perceptions of how members of their group are aspirations in stereotyped domains (Davies generally treated (Mendoza Denton, Downey, et al., 2005), while other research indicates that Purdie, Davis, & Pietrzak, 2002) or from cues it can affect the way individuals relate to and in the environment that raise the possibility of identify with their own ingroup (e.g. Pronin, such devaluation (Ellemers, Spears, & Doosje, Steele, & Ross, 2004). Moreover, research study- 2002; Murphy, Steele, & Gross, 2007). Regardless ing how stigmatized people cope with stereo- of the source of realization, the person might type threat shows that chronic threats to social become vigilant to the possibility of identity identity can lead them to disengage from do- threat and sensitive to the cues communicating mains in which their group is negatively stereo- that their group’s stigmatized social status may typed (Major & Schmader, 1998; Major, Spencer, be relevant in the immediate situation (Kaiser, Schmader, Wolfe, & Crocker, 1998; Osborne, Vick, & Major, 2006). Some settings provide few 1995), and compensate for group devaluation cues that stereotypes and prejudice are relevant by claiming their group’s superiority in other and thus result in individuals experiencing little performance domains (Derks, Van Laar, & or no threat to their social identities. These Ellemers, 2006, 2007). The main point here is ‘identity-safe’ environments assure individuals that stereotype threat can affect more than just that their social identities are valued and pose performance, a point to which we will return no barrier (Davies, et al., 2005). Other situations, when describing some of our new lab studies. however, are less hospitable and can become identity threatening by dint of cues within Expansion and refi nement them. A casual chat in one’s dorm room, for In recent years, theorists have both expanded example, may become threatening when a and refi ned the concept of stereotype threat. fi rst-generation college student realizes that all Although stereotype threat is quite pervasive, it his roommates’ parents are college graduates is but one form of social identity threat (Steele et and highly paid professionals. When people al., 2002), a situational predicament occurring are uncertain of their standing, watchful for when individuals become fearful that their stigma-relevant cues, and fi nd themselves in just social category is devalued by others (Tajfel & such an inhospitable environment, the process Turner, 1986; see also Crocker & Major, 1989). of social identity threat is set off. People whose social identity is threatened are At the same time that theorists have elaborated not only worried about what negative stereo- upon the concept of threat, others have tried types say about them, but also about what they to refi ne and categorize the types of threat, say about groups with which they belong and moving the fi eld from social identity threat to identify (Tajfel & Turner, 1986). People may be social identity threats. In a recent review of the threatened by negative stereotypes about their literature, Shapiro and Neuberg (2007) suggested group, but they may also be threatened when that stereotype threat actually means different they suspect that their social identity can put things to different researchers because they often them at risk for social devaluation, exclusion, and use stereotype threat as an umbrella concept

166 Derks et al. neuroscience of stigma without articulating the subconcepts that de- understanding of process has improved greatly fi ne it. To correct for this, Shapiro and Neuberg since those early days. Social identity threat is proposed a multi-threat framework whereby best thought of as a predicament faced by a there is not simply a single stereotype threat, person in a situation; given the range of pos- but rather six qualitatively distinct stereotype sible situations, groups of people, and types of threats. Specifi cally, negative stereotypes can stereotypes, it should come as no surprise that lead individuals to experience different types social identities can be threatened through of threats that arise from the intersection of multiple, interacting pathways. The most pro- two dimensions: the target of the threat (the minent of these are the physiological stress self/one’s group) and the source of the threat response due to increased arousal (Ben Zeev, (the self/outgroup others/ingroup others). So, Fein, & Inzlicht, 2005; Blascovich, Spencer, for example, someone could be threatened if Quinn, & Steele, 2001; O’Brien & Crandall, they worry that the stereotype is true of them 2003; Scheepers & Ellemers, 2005), negative personally or true of their group; they could cognitions that consume executive resources worry that outgroups see them as stereotypic or (Beilock et al., 2007; Cadinu, Maass, Rosabianca, & see their group as stereotypic; and they could Kiesner, 2005; Schmader & Johns, 2003), and worry that ingroups see them as stereotypic or self-regulatory changes in the aims of managing their own group as stereotypic (e.g. Inzlicht & thoughts and emotions (Derks et al., 2007; Ben Zeev, 2003; Marx, Stapel, & Muller, 2005; Seibt & Förster, 2004; Spencer, 2003). Quinn, Kahng, & Crocker, 2004; Schmader, There is support for all of these putative 2002). Similarly, Branscombe and colleagues mechanisms, but what has been missing is (Branscombe, Ellemers, Spears, & Doosje, 1999) an understanding of how these mechanisms defi ned four different types of social identity fit together. Recently, however, Schmader, threat. People can experience threat because Johns, and Forbes (in press) have proposed a their group is devalued (value threat), because comprehensive, integrated model of stereotype they are categorized as members of a group to threat where threat is thought to be produced which they do not want to belong (categoriza- by three distinct, yet interconnected pathways. tion threat), because their group lacks distinct- Placing executive control as the proximal iveness from other groups (distinctiveness mediator, the model implicates a physiological threat), or because they worry whether they are stress response, a performance monitoring accepted within their group (acceptance threat). response, and an affect regulation response. Although there is mounting evidence for many For example, when faced with the possibility of these forms of social identity threat, it is still of confirming a negative stereotype about unclear how these two models converge and their social identity, people may feel anxious how many forms of threat there actually are. and physiologically stressed, they may monitor Nevertheless, it is clear that stereotype threat their performance for what it means for them and social identity threat take on more than and their group’s reputation, and they may try one form, an articulation of which can help us to suppress and deny these emotions so that comprehend one of the more important issues others won’t think they are admitting defeat. All surrounding the threat concept, the issue of three of these processes can tax one’s limited mediation. executive control strength, which are the same cognitive resources needed for intellectual Mediation and process performance. Ultimately, this can result in the Since the publication of the fi rst empirical intellectual performance defi cits seen in studies paper by Steele and Aronson in 1995, researchers of stereotype threat. have asked questions about process, mech- anism, and mediation; fundamentally asking, Limitations ‘What is stereotype and social identity threat?’ While we have learned a great deal about stereo- Although there is still much to discover, our type and social identity threat over the years, some

167 Group Processes & Intergroup Relations 11(2) of the fundamental questions—especially those (e.g. Gaertner & Dovidio, 1986; Oskamp, 2000; related to mechanism—have been more diffi cult Zanna & Olson, 1994). The past two decades, to elucidate. Part of this, we suppose, is because however, have seen a rise in research focusing of the fi eld’s reliance on the measurement of on how prejudice and stereotyping affect its overt behaviors and especially the use of self- victims (e.g. Crocker & Major, 1989; Crocker, report. Let’s take the examination of arousal Major, & Steele, 1998; Ellemers, Spears, & and anxiety in mediating stereotype threat as Doosje, 1999; Jost & Major, 2001; Levin & Van one example. Although Steele and Aronson Laar, 2006; Swim & Stangor, 1998). Examining (1995) suggested that threat was mediated by the existing social neuroscience research on anxiety and emotion in their original paper, group processes and intergroup relations reveals multiple attempts to support this proved futile a similar trend, with most research focused on (e.g. Spencer et al., 1999; Stone et al., 1999). people who stereotype others, rather than on This assumption only gained support once re- the stigmatized themselves (e.g. Amodio et al., searchers started using online measurements of 2007; Bartholow & Dickter, 2007; Ito, Urland, psychophysiology (e.g. Blascovitch et al., 2001) Willadsen Jensen, & Correll, 2006; Ito et al., or indirect manipulations of arousal state (e.g. 2007; Richeson et al., 2003). The introduction Ben-Zeev et al., 2005). And as we will see below, of social neuroscience methods to research on it gained even more traction once researchers stereotyping from the perpetrator’s perspective started measuring direct brain activity with the has yielded insights into the processes that modern tools of neuroscience. underlie prejudice and racial bias. Likewise, The primary objective of this paper is to high- we suspect that bringing a social neuroscience light the important role that social neuroscience approach to social identity and stereotype threat has played—and will soon play—in increasing will lead to similar advancements. our appreciation of stereotype and social iden- In the review that follows, we discuss four tity threat. The use of such tools as fMRI, EEG, lines of social neuroscience research on the and event-related potentials (ERP) allows re- target’s perspective. The fi rst consists of two searchers to answer questions concerning under- fMRI studies examining neural correlates of lying processes. Importantly, using a social neuro- stereotype threat, more specifi cally, showing science approach—which is characterized by a brain activation related to the experience of desire to understand phenomena at the levels being stereotyped. The second is an ERP study of social psychology, cognitive psychology, and examining performance-interfering academic neuroscience (Harmon-Jones & Winkielman, disengagement among African American col- 2007; Ochsner & Lieberman, 2001)—can further lege students. Finally, the third and fourth re- clarify how social identity threat not only affects search lines are ERP studies from our own lab. performance, but also affects behaviors in other The fi rst of these projects reveals the important domains and the self-concept. Ultimately, by role of the neurocognitive confl ict-detection capitalizing on biological concepts and methods, system in stereotype threat effects, especially this type of research offers an integrative analysis as it pertains to stereotype threat spillover into that can inform and refi ne theories related to cognitive control. The second project focuses stereotype threat and stigmatization, more broadly on neural correlates of compensatory ingroup (Cacioppo et al., 2007). bias, a coping mechanism which may be used to alleviate social identity threat. Neuroscience research on stigma Stereotype threat effects examined with fMRI and social identity threat fMRI enables researchers to non-invasively Initially, most social psychological research on study brain activity online with an impressive prejudice and stereotyping examined functions, level of spatial resolution. By tracking the move- characteristics, and consequences for people ment and use of oxygen in the brain, researchers who use stereotypes and hold prejudiced beliefs can identify regions of high or low neural activity,

168 Derks et al. neuroscience of stigma and make inferences about the involvement of (Moran, Macrae, Heatherton, Wyland, & Kelley, specifi c brain regions in cognitive or behavioral 2006; Somerville, Heatherton, & Kelley, 2006). processes (Bandettini, Birn, & Donahue, 2000). These results suggest that stereotype threat Within the fi eld of stigma, fMRI research allows affects women’s math performance in two ways: us to see fi rst-hand the relationship between the fi rst, it disrupts normal recruitment of cognitive environment and the brain. This type of work areas required for high math performance and, is important, as observations of actual neural second, it increases the recruitment of areas consequences of stigma are a step toward under- which allow for the processing and regulation standing the mechanism underlying com- of emotions. These results converge with and, monly observed behavioral effects of stigma, indeed, largely confi rm, previous behavioral for example, performance decrements or dis- research (Beilock et al., 2007; Schmader & Johns, engagement. Two examples of social neuro- 2003; Schmader et al., in press) suggesting that science research on stereotype threat examine women perform more poorly under stereotype brain activation of women in normal and stereo- threat because valuable cognitive resources are type threat conditions (Krendl, Richeson, Kelley, spent on emotional regulation instead of on & Heatherton, 2008; Wraga, Helt, Jacobs, & the task at hand. The incremental benefi t of the Sullivan, 2007). fMRI work here is in the ability to test a behavioral Using fMRI, Krendl and colleagues (2008) theory at the biological level of action, the brain. sought to understand the neural basis for This can serve as a springboard for further women’s math performance decrements under theory-testing investigations. The fact that these stereotype threat. In both the stereotype threat results converge with the behavioral work of and control conditions, women fi rst completed a others provides consistency across different baseline measure of math performance; women levels of analysis and organization, an important in the threat condition were then reminded of step toward the broad understanding of any the negative stereotype regarding women and complex phenomenon. math performance. Following this, all of the These results were supported and further women completed a second measure of math specifi ed by Wraga and colleagues (2007) in an performance. In addition to the classic stereotype attempt to link brain activation associated with threat performance effect (women in the con- affective processing and reduced performance trol condition displayed improved math per- under stereotype threat. Although their study formance over time, women in the stereotype focused on a different gender stereotype, namely threat condition performed worse over time), the negative stereotype concerning women’s the fMRI results shed light upon the underlying mental rotation performance, the results show mechanisms of this effect. similar consequences of stereotype threat on Not surprisingly, women in the control con- regions associated with emotion regulation. dition showed increased recruitment of neural Female participants were either confronted regions that have been associated with math with a positive stereotype related to women’s learning and performance, including the inferior superior perspective taking skills, a negative prefrontal cortex, left inferior parietal cortex, stereotype related to women’s inferior spatial and bilateral angular gyrus (Dehaene, Spelke, reasoning abilities, or with no stereotype (control Pinel, Stanescu, & Tsivkin, 1999; Delazer et al., condition). fMRI analyses revealed that women 2003). However, women who had been reminded under stereotype threat showed increased activity of their group’s supposed math inferiority did in areas associated with emotional self-regulation not show this enhanced activation in useful, (rostral-ventral anterior cingulate cortex) and math-related brain regions. Instead, they showed social knowledge (right orbital gyrus, Milne & increased activation in the ventral anterior cin- Grafman, 2001). Not only did this study show gulate cortex, an area that has been associated increased activation, it showed that this pattern with emotional self-regulation and the processing of brain activation actually predicted women’s of affective information and social feedback worse performance in spatial rotation.

169 Group Processes & Intergroup Relations 11(2)

Using brain activity to predict actual perform- one or more information processing module ance is a crucial step—it allows us to go beyond (e.g. categorization, evaluation, etc.). The ampli- simply mapping stereotype threat effects in tude of the ERP refl ects the degree to which the the brain to actually predicting behavior based specifi c module is engaged, while the latency on differential activation of brain regions, in refl ects the speed with which this process is this case, brain regions associated with cognit- completed (Cacioppo, Crites, Gardner, & ive processing and emotion regulation. These Berntson, 1994). two studies contribute to our understanding Forbes, Schmader, and Allen (2008) incor- of stereotype threat effects in two ways. First, by porated ERP measures to understand another revealing the disrupted brain activity of women consequence of stigma, academic disengage- under stereotype threat, they provide insight ment. Previous research has shown that con- into the mechanism through which stereotype tinued exposure to stereotype threat and threat can impinge upon women’s cognitive chronic social identity threat can lead ethnic performance. As has been noted by others (e.g. minority group members to disengage from Schmader et al., in press), even though stereo- their academic performance, thereby prevent- type threat is a complex, multi-determined phe- ing poor performance and negative group nomenon, most studies focus on one mediator stereotypes from impinging on their self-worth at a time. The use of a neuroscience approach, (Major & Schmader, 1998; Major et al., 1998). however, can provide a more integrated pic- Stigmatized group members can disengage ture: here it shows that threatened women from performance domains by devaluing the experience reductions in cognitive effi ciency importance of academic achievements, and by and increases in affective processing at the same discounting the diagnosticity of academic per- time. Second, although these studies focused formance feedback (Major & Schmader, 1998). on different gender stereotypes, they both Although domain disengagement alleviates showed increased recruitment of brain areas threats to one’s social identity, it may also con- associated with emotion regulation. This pro- tribute to lower academic performance (Osborne, vides evidence that although gender stereotypes 1995, 1999; Verkuyten & Thijs, 2004). Forbes vary in content, they set in motion the same and colleagues (2008) sought to understand detrimental effects at the neural level. the negative effects of academic disengagement from a neural perspective. An ERP examination of academic In their study, students from ethnic minority disengagement groups performed a confl ict-monitoring task Another neuroscience technique that has (the Eriksen Flanker task, Eriksen & Eriksen, been applied to the study of stigma is the ERP 1974) that was framed either neutrally or as technique. ERPs are derived from continuous diagnostic of intelligence. Participants then measures of electrical activity in the brain, EEG, reported the value they attached to academic by extracting and averaging short epochs (e.g. performance. Next, participants completed 1 second) of data that directly follow an event, the Flanker task while continuous EEG was re- for example, the presentation of a stimulus corded. In the Flanker task, participants must (i.e. stimulus-locked ERP) or a participant’s identify a target stimulus as one of two given response (response-locked ERP). By averaging letters, say M or N, while four congruent fl anker over a large number of epochs, background noise stimuli (MMMMM), or four incongruent in the EEG due to other cognitive processes is fl anker stimuli (NNMNN), are simultaneously cancelled out and one is able to focus precisely presented. Incongruent trials are characterized on the brain activity of interest. Within the ERP by greater response confl ict and require more waveform, researchers have identifi ed different self-control. components that are described in terms of their Performance monitoring on this task was polarity (Positive/Negative) and latency (in ms, measured by a short-latency response-locked e.g. P300, N200). Each ERP component refl ects negative wave component of the ERP called the

170 Derks et al. neuroscience of stigma error-related negativity (ERN). The amplitude of outcome is not achieved. Distinguishing these the ERN is larger after errors than after correct processes with traditional behavioral mea- responses, and is therefore interpreted as a sures is diffi cult; however, neurophysiological neural indicator of performance monitoring data might help by revealing online brain and error detection (Gehring, Coles, Meyer, & processes. Donchin, 1995; Gehring, Goss, Coles, Meyer, & In our experiment, we induced stereotype Donchin, 1993). Forbes et al. showed that the threat by giving our male and female partici- degree to which members of ethnic minority pants a ‘diagnostic’ math test. We then had par- groups devalue academics predicted how much ticipants complete the Stroop color-naming they monitored their own performance on the task, a task that requires attentional control and Flanker task (as indicated by ERN amplitude). self-regulation (Ellis, Rothbart & Posner, 2004). This was only the case, however, when the task To examine whether women under stereotype was framed as diagnostic of intelligence: threat either misregulate or underregulate their whereas devaluing the academic domain was self-control compared to men, we measured the unrelated to ERN amplitude when the task amplitude of medial-frontal potentials after was presented neutrally, academic devaluing participants responded. These medial-frontal negatively predicted ERN amplitude when the waves refl ect activation of the preconscious task was presented as diagnostic of intelligence. confl ict monitoring system and constitute the This indicates that when minority students earliest indication that the confl ict monitoring no longer care for academic performance, they system has detected a mismatch between in- cease to monitor their performance, especially tended and actual responses or, in essence, self- when negative stereotypes are ‘in the air’. Again, control failure (Gehring et al., 1993; Falkenstein, the results here converge with other behavioral Hoormann, Christ, & Hohnsbein, 2000). In tasks work (Major & Schmader, 1998) and provide like the Stroop where participants are required consistency across different levels of analysis to make choices quickly, this waveform can be and organization. seen after incorrect responses, but also after correct responses on high-confl ict trials (e.g. Stereotype threat and misregulation of ‘RED’ in blue font). They are normally not seen, self-control however, on low-confl ict trials (e.g. ‘RED’ in red As mentioned earlier, negative stereotypes can font) (Yeung, Botvinick, & Cohen, 2004). impair effortful self-regulation and spillover to By examining this ERP after errors and high- domains unrelated to threatening stereotypes and low-confl ict correct responses, we can deter- (Beilock et al., 2007; Inzlicht et al., 2006). But mine whether stereotype threat spillover is due how does stereotype threat spillover? Emerging to underregulation or misregulation. Recent work from our lab suggests that the neurally research suggests that neural mechanisms under- based self-control system plays a role (Inzlicht lying confl ict-monitoring are weakened among et al., 2008). underregulated individuals (Inzlicht & Gutsell, According to Baumeister and Heatherton 2007). Therefore, if stereotype threat results (1996), there are at least two distinct causes in underregulation, neural responses to errors for self-control failure: underregulation and should be dampened. If stereotype threat results misregulation. Underregulation is said to in misregulation, neural responses should be occur when an individual fails to exert self- heightened whereby everything, even insignifi - control because he or she is unwilling or simply cant low-confl ict trials, is fl agged as attention- unable to do so (i.e. ego-depletion, Muraven & worthy. Baumeister, 2000; Muraven Tice & Baumeister, Our results suggest a role for misregulation. 1998). Misregulation, on the other hand, is Male participants showed larger waves to in- said to occur when self-control strength is ap- congruent than congruent trials, indicating plied in a misguided or counterproductive way that they were correctly detecting which trials such that despite one’s best efforts, the desired required their attention. Female participants,

171 Group Processes & Intergroup Relations 11(2) in contrast, displayed higher amplitude waves 1984; Scheepers, Spears, Doosje, & Manstead, overall, ruling out the possibility that they were 2006; Tajfel & Turner, 1986). In an ERP study, underregulated. They also displayed higher we sought to gain more insight into this compen- waves for congruent than incongruent trials. satory ingroup bias by mapping out automatic Because these medial-frontal waves indicate (ERP) and controlled (self-report) forms of that events are of signifi cance and require at- ingroup evaluations. tention, correction, and strategic adjustment Researchers examining the relationship be- (Bartholow et al., 2005; Holroyd & Coles, 2002), tween threats to group value and compensatory these results suggest that our female participants ingroup bias have often encountered diffi culties were misregulated, such that low confl ict events in measuring social identity threat itself. Most (congruent trials) were unnecessarily reacted to of this research has attempted to assess threat with equal or greater vigilance than high con- and evaluations of the ingroup with self-report fl ict trials (incongruent trials). Importantly, this measures of, for example, anxiety, collective misregulation, or overregulation of easy low- self-esteem, or evaluations of the ingroup and confl ict trials, signifi cantly mediated the effect outgroup on specifi c dimensions. However, of stereotype threat on cognitive control on the several researchers have suggested that people Stroop task. under identity threat might be unable, or even The insight gained with this study is that unwilling, to accurately report their complex stereotype threat spillover is due to a misregu- and varied feelings toward their own ingroup lation of the confl ict-monitoring system rather (Bettencourt, Miller, & Hume, 1999; Branscombe than underregulation. Threatening math et al., 1999; Scheepers & Ellemers, 2005). Re- settings leave women in a state where their cently, however, researchers have measured neural systems show increased responsivity to social identity threat more reliably by looking both signifi cant and non-signifi cant events. This at cardiovascular indices of arousal (Blascovich supports the assertion that women devote a et al., 2001; Scheepers & Ellemers, 2005). In our large amount of effort and resources to control study, we assessed group value threat even outcomes and attend more carefully to the vari- more specifi cally with ERP measures that indi- ous consequences of their actions (Jamieson cate stigmatized group members’ automatic & Harkins, 2007; Murphy et al., 2007). This (vs. controlled) evaluation of their ingroup heightened, but unnecessary, vigilance might and outgroup. then lead the confl ict-monitoring system to be The starting point of our study was ERP re- non-optimally responsive to confl ict in other search conducted by Ito and her colleagues domains and ultimately lead to self-control showing that White individuals automatically failure later on. evaluate racial ingroup members more positively than racial outgroup members (Ito,Thompson, & Neurological correlates of compensatory Cacciopo, 2004; Ito & Urland, 2005; Ito et al., ingroup bias 2006). This research focused on a specifi c com- In a second project, we examined the underly- ponent of the ERP termed the late positive ing neural processes of compensatory forms of potential (LPP), which is a positive-going ingroup bias (Derks & Inzlicht, 2008). In addition defl ection that peaks between 350 to 900 ms to the domain devaluing strategy examined by after presentation of a stimulus (Donchin, Forbes and colleagues (2008), stigmatized group 1981; Goldstein, Spencer, & Donchin, 2002; members can cope with negative stereotypes Nieuwenhuis, Aston Jones, & Cohen, 2005). and social identity threat by explicitly favoring LPP amplitude varies with the degree to which their own group over a higher status outgroup a stimulus is evaluated differently from the in performance domains unrelated to the context in which it is presented, thus refl ecting negative stereotype (Brewer, 1979; Cadinu & automatic evaluation of stimuli (Cacioppo Cerchioni, 2001; Derks et al., 2007; Ellemers & et al., 1994). Ito and colleagues (2004) utilized Van Rijswijk, 1997; Mummendey & Schreiber, this characteristic of the LPP to measure White

172 Derks et al. neuroscience of stigma participants’ automatic evaluation of racial 2001; Ellemers, Van Dyck, Hinkle, & Jacobs, ingroup and outgroup members in an ‘oddball 2000; Ellemers & Van Rijswijk, 1997). task’. Results revealed that White faces elicited This study presents a new way of measuring higher LPPs than Black faces, indicating that a specifi c form of social identity threat, namely within the context of negatively valenced pic- threat to group value (Branscombe et al., 1999). tures, ingroup faces were seen as more evaluatively Our results indicate that this type of social incongruent than outgroup faces. Importantly, identity threat affects the automatic associations this automatic form of ingroup bias measured women have with their gender group. By meas- with the LPP was positively related to explicit uring people’s neural responses to in- and out- measures of modern racism. group related stimuli, social identity effects can In our study we adjusted this ERP paradigm be assessed more directly than was previously to measure how female participants automatic- possible with self-report measures. These ERP ally evaluated pictures of men and women when measurements thus allow us to differentiate the group value threat was present or absent. We automatic associations group members have hypothesized that automatic evaluations of with their ingroup from their controlled evalu- female targets would be lower when female ations of this group. Moreover, by relating the participants were confronted with negative neural measures to controlled reports of com- stereotypes about their group than when they pensatory ingroup bias, we not only revealed were confronted with positive stereotypes about the differential results obtained with implicit their group. Moreover, by also administering and explicit measures, but also underlined self-report measures of compensatory ingroup the compensatory nature of reporting ingroup bias, we aimed to show that lower automatic bias. evaluations of the ingroup following a negative stereotype would be related to higher explicit Strengths and limitations of social compensatory ingroup bias. neuroscience Our results showed that stereotype activation predicted LPP amplitude to female faces. Par- With this review of social neuroscience research ticipants confronted with a negative stereotype on stigma, we highlighted some examples of showed smaller LPPs to female faces than research using techniques derived from neuro- participants confronted with a positive stereotype. science to contribute to our understanding of This indicates that in the negative stereotype stereotype threat and social identity processes. condition these oddballs stood out less in the Although some may argue that it is too early to negative context. In addition to this ‘ingroup summarize the social neuroscience work that devaluation effect’, the results also revealed has taken the target’s perspective, this review the differences between neural and self-report offers a glimpse of how an integrated, biological measures: while implicit ingroup evaluation (as approach can further our understanding of indexed by the LPP) suggested that threatened complex social psychological phenomena. We women maligned their own ingroup, self-report conclude this article by discussing the incremental of ingroup evaluation suggested that threatened value of social neuroscience in understanding women favored the ingroup. Moreover, LPP stigma and social identity threat. amplitude to female targets was negatively re- The fMRI studies we reviewed (Krendl et al., lated to this explicit ingroup bias. That is, lower 2008; Wraga et al., 2007) revealed the patterns of automatic evaluations of the ingroup were brain activation that result from the experience associated with higher explicit ingroup bolstering of stereotype threat. These studies lend support when women were confronted with negative to previous behavioral theories which posited stereotypes about their gender. We suggest that dealing with the emotions aroused by that this could be interpreted as an overt com- stereotype threat lead to decrements in cognitive pensation attempt (see also Cadinu & Cerchioni, performance. Indeed, these studies showed

173 Group Processes & Intergroup Relations 11(2) that the experience of stereotype threat leads In this review, we also discussed how ERP tech- to heightened activity in emotion-regulation niques have been used to study coping strategies centers of the brain, while simultaneously low- used by targets of stereotyping. These measures ering activation in brain regions associated with suggested that when ethnic minority students high math or spatial performance. Although disengage from the academic domain, they are previous research has shown that stereotype less likely to automatically monitor their per- threat increases cardiovascular threat responses formance in that domain (Forbes et al., 2008). (Blascovich et al., 2001) and taxes verbal working Although previous research had suggested that memory (Beilock et al., 2007; Schmader & Johns, devaluing would lead to lower performance 2003), to our knowledge, these are the fi rst stu- among stigmatized individuals (Crocker et al., dies to document these effects at the same time. 1998; Steele 1992, 1997), this research further Examining stereotype threat with neuroimaging elucidates this process by showing why this is techniques, in other words, offers the chance the case. The fi nding that domain devaluation to measure multiple mechanisms as they actually leads to lower performance monitoring occur and interact. This helps achieve a richer provides researchers with an additional tool to understanding of how individuals cope with examine people’s investments in specifi c per- stereotype or social identity threat. formance domains in an implicit way. Being We also discussed research illustrating how able to study this process outside of participants’ ERP techniques can improve our understanding awareness can be valuable, for example, when of stereotype threat spillover effects (Inzlicht, participants are unlikely to correctly report their Kang, & Fortune, 2008). ERP techniques enabled investment in certain domains due to demand us to measure how women under stereotype characteristics. Moreover, self-report measures threat monitor their performance, revealing of domain investment can be problematic in studies that focus on changes in domain invest- that self-control failure under stereotype threat ment due to experimental manipulation, as it is is not a matter of underregulation due to cog- unclear whether participants have insight into nitive exhaustion or ego-depletion (Inzlicht & changes in domain investment and are able to Gutsell, 2007), but rather misregulation of exe- report their investment accurately. cutive functions that lead women to vigilantly In the same vein, our own work (Derks & monitor their performance, even on easy tasks Inzlicht, 2008) showed that threats to group where this vigilance is unnecessary. This distinc- value, which is diffi cult to assess directly using tion is crucial as it suggests that the reduction traditional self-report measures, does show up of stereotype threat effects is not a matter of in ERP measures that indicate automatic rather increasing women’s motivation or self-control than controlled devaluation of the ingroup. strength, but rather, it is a matter of teaching Moreover, we were able to show that automatic women to reappraise or reduce their anxiety ingroup evaluations were negatively correlated in order to maintain a normal pattern of atten- with compensatory ingroup bias in response tional regulation (Ben Zeev et al., 2005; Cohen to threat. Although this study affi rmed what et al., 2006; Martens, Johns, Greenberg, & previous research had already assumed about Schimel, 2006; Sherman, Kinias, Major, Kim, & the mediating role of social identity threat in Prenovost, 2007). This ERP paradigm improves compensatory ingroup bias, using the LPP our understanding of the effects of inappropriate to assess automatic evaluation of the ingroup vigilance as it provides us with a straightforward and the outgroup does provide researchers with and clean way to measure different patterns of a new tool to assess the magnitude of social self-regulation (normal vs. misregulation vs. identity threat in different settings. Assessing underregulation) and determine how they relate the automatic association people have with to actual performance. Thus, ERP techniques social categories is not easy to do with behavioral offer an opportunity to directly test relationships measures. This is because even relatively implicit between self-regulation and performance. measures of these evaluations, such as the

174 Derks et al. neuroscience of stigma

Implicit Associations Test (Greenwald, McGhee, minds of those who are stigmatized. Although & Schwartz, 1998), require participants to give social neuroscience techniques expand our ex- a response, making it hard to distinguish be- perimental toolbox, one should not conclude tween automatic and controlled processes on that social psychological phenomena are not these outcomes (Conrey, Sherman, Gawronski, real until their neural correlates have been Hugenberg, & Groom, 2005). ERP measures identifi ed. Neuroscience techniques are not enable us to assess implicit associations more some ‘pipeline’ to the mind, offering evidence reliably, by giving us more direct insight into that is somehow ‘better’ than behavioral evi- automatic social identity processes. Being able dence (Kihlstrom, 2006). These techniques have to measure changes in social identity directly their advantages, but ultimately any explanation can be valuable in studies examining, for will need to address brain and behavior. When example, which contextual cues elicit identity deciding to use neurological measures in re- threat or the effectiveness of experimental search, therefore, one should always evaluate manipulations designed to alleviate the negative whether the same insights could not be gained effects of social identity threat on stigmatized using simpler measures. One should also be group members. careful of being too easily swayed by activity in We believe that social neuroscience gains us the brain. A recent study suggests that people insight into social psychological processes by are easily convinced of the veracity of a bad ex- integrating the fi ndings in our own fi eld with planation when it is accompanied by irrelevant that of cognitive neuroscience. In our opinion, neuroscientifi c information (Weisberg, Keil, this integration is social neuroscience’s greatest Goodstein, Rawson, & Gray, 2008). We hope, strength. Social psychological theories can only however, that the many excellent examples gain what E. O. Wilson calls consilience when of previous social neuroscience research can they are explained and described at multiple inspire stigma researchers to fi nd ways to ad- levels of analysis and found to be consistent dress their own research questions with this with one another (Wilson, 1998; see also Snow, new approach (for examples see Cacioppo 1959). Only by extending our theories and et al., 2007; Cacioppo & Berntson, 2004, 2005; fi ndings to other social and biological levels Cacioppo, Visser, & Pickett, 2006; Harmon-Jones of analysis, can we ensure convergence of data & Winkielman, 2007). and explanation (Cacioppo, 2004). In other One often heard criticism in discussions words, peering into the brain adds an extra about the usefulness of social neuroscience dimension of analysis and offers the possibility measurements (especially concerning brain of gaining a multilevel, integrative analysis of imaging techniques such as fMRI) that social complex social phenomena, such as stereotype neuroscience is simply a new form of phrenol- threat. ogy (Uttal, 2002; but see Landreth & Richardson, 2004). Although we think that there is much Limitations and recommendations for future to learn from brain imaging in itself—as it research allows us to link social psychological processes We hope that this review of current social neuro- to knowledge about specifi c functions of brain science stigma research will encourage other regions—we think it is important for researchers social psychologists working on this important to move beyond simply showing effects on topic to consider using social neuroscience in fMRI, EEG, or ERP measures. Instead, we sug- their own research. Research on stereotype threat gest that researchers relate measures of brain and social identity protection has sometimes functioning to behavioral measures. The fi eld been limited by diffi culties in reliably assessing of social neuroscience does not limit itself to the emotions, cognitions, and motivations of questions involving localization and neural our subjects. Techniques developed in neuro- systems, but actually provides measures that science can be an additional tool for gaining allow us to directly assess the mediating role more direct insight into what goes on in the of social cognition on behavior. By adopting

175 Group Processes & Intergroup Relations 11(2) this mediational approach, social psychologists control and medial frontal negativity: Beyond can contribute back to cognitive neuroscience, errors and response confl ict. Psychophysiology, as results from studies in our fi eld will help to 42, 33–42. uncover the complex functions of specifi c brain Baumeister, R. F., & Heatherton, T. F. (1996). Self-regulation failure: An overview. Psychological regions. Inquiry, 7, 1–15. Beilock, S. L., Jellison, W. A., Rydell, R. J., Note McConnell, A. R., & Carr, T. H. (2006). On the causal mechanisms of stereotype threat: Can TM 1. According to the ISI Web of Science , Steele skills that don’t rely heavily on working memory and Aronson’s (1995) article has been cited still be threatened? Personality and 609 times as of September 12, 2007. Social Psychology Bulletin, 32, 1059–1071. Beilock, S. L., Rydell, R. J., & McConnell, A. R. References (2007). Stereotype threat and working memory: Mechanisms, alleviation, and spillover. Journal of Amodio, D. M., Devine, P. G., & Harmon-Jones, Experimental Psychology: General, 136, 256–276. E. (2007). Mechanisms for the regulation of Ben Zeev, T., Fein, S., & Inzlicht, M. (2005). intergroup responses: Insights from a social Arousal and stereotype threat. Journal of neuroscience approach. In P. Winkielman & Experimental Social Psychology, 41, 174–181. E. Harmon-Jones (Eds.), Social neuroscience: Bettencourt, B. A., Miller, N., & Hume, D. (1999). Integrating biological and psychological explanations Effects of numerical representation within of social behavior (pp. 353–375). New York: cooperative settings: Examining the role of Guilford. salience in in-group favouritism. British Journal Aronson, J., & Inzlicht, M. (2004). The ups and of Social Psychology, 38, 265–267. downs of attributional ambiguity: Stereotype Blascovich, J., Spencer, S. J., Quinn, D., & Steele, vulnerability and the academic self-knowledge of C. (2001). African Americans and high blood African American college students. Psychological pressure: The role of stereotype threat. Science, 15, 829–836. Psychological Science, 12, 225–229. Aronson, J., Lustina, M. J., Good, C., Keough, Branscombe, N. R., Ellemers, N., Spears, R., & K., Steele, C. M., & Brown, J. (1999). When Doosje, B. (1999). The context and content of White men can’t do math: Necessary and social identity threat. In N. Ellemers, R. Spears, suffi cient factors in stereotype threat. Journal of & B. Doosje (Eds.), Social identity: Context, Experimental Social Psychology, 35, 29–46. commitment, content (pp. 35–58). Oxford UK: Bandettini, P. A., Birn, R. M., & Donahue, K. Blackwell. M. (2000). Functional MRI: Background, Brewer, M. B. (1979). In-group bias in the minimal methodology, limits, and implementation. In intergroup situation: A cognitive-motivational G. G. Berntson, J. T. Cacioppo, & L. G. Tassinary analysis. Psychological Bulletin, 86, 307–324. (Eds.), Handbook of psychophysiology (2nd ed., Cacioppo, J. T. (2004). Common sense, intuition, pp. 978–1014). New York: Cambridge University and theory in personality and social psychology. Press. Personality and Social Psychology Review, 8, Bartholow, B. D., & Dickter, C. L. (2007). Social 114–122. cognitive neuroscience of person perception: Cacioppo, J. T., Amaral, D. G., Blanchard, J. J., A selective review focused on the event-related Cameron, J. L., Carter, C. S., Crews, D. et al. brain potential. In P. Winkielman & E. Harmon- (2007). Social neuroscience: Progress and Jones (Eds.), Social neuroscience: Integrating implications for mental health. Perspectives on biological and psychological explanations of social Psychological Science, 2, 99–123. behavior (pp. 376–400). New York: Guilford. Cacioppo, J. T., & Berntson, G. G. (2004). Essays in Bartholow, B. D., Dickter, C. L., & Sestir, M. A. social neuroscience. Cambridge, MA: MIT Press. (2006). Stereotype activation and control of Cacioppo, J. T., & Berntson, G. G. (2005). Social race bias: Cognitive control of inhibition and its neuroscience: Key readings. New York: Psychology impairment by alcohol. Journal of Personality and Press. Social Psychology, 90, 272–287. Cacioppo, J. T., Crites, S. L., Gardner, W. L., & Bartholow, B. D., Pearson, M. A., Dickter, C., Sher, Berntson, G. G. (1994). Bioelectrical echoes K. J., Fabiani, M., & Gratton, G. (2005). Strategic from evaluative categorizations: I. A late positive

176 Derks et al. neuroscience of stigma

brain potential that varies as a function of trait professionally. Personality and Social Psychology negativity and extremity. Journal of Personality and Bulletin, 28, 1615–1628. Social Psychology, 67, 115–125. Davies, P. G., Spencer, S. J., & Steele, C. M. (2005). Cacioppo, J. T., Visser, P. S., & Pickett, C. L. (2006). Clearing the air: Identity safety moderates Social neuroscience: People thinking about thinking the effects of stereotype threat on women’s people. Cambridge, MA: MIT Press. leadership aspirations. Journal of Personality and Cadinu, M. R., & Cerchioni, M. (2001). Social Psychology, 88, 276–287. Compensatory biases after ingroup threat: ‘Yeah, Dehaene, S., Spelke, E., Pinel, P., Stanescu, R., but we have a good personality’. European Journal & Tsivkin, S. (1999). Sources of mathematical of Social Psychology, 31, 353–367. thinking: Behavioral and brain-imaging Cadinu, M. R., Maass, A., Rosabianca, A., & Kiesner, evidence. Science, 284, 970–974. J. (2005). Why do women underperform Delazer, M., Domahs, F., Bartha, L., Brenneis, under stereotype threat? Evidence for the role C., Lochy, A., Trieb, T. et al. (2003). Learning of negative thinking. Psychological Science, 16, complex arithmetic—an fMRI study. Cognitive 572–578. Brain Research, 18, 76–88. Cohen, G. L., Garcia, J., Apfel, N., & Master, A. Derks, B., & Inzlicht, M. (2008). Neural correlates (2006). Reducing the racial achievement gap: of compensatory ingroup bias following social A social-psychological intervention. Science, 313, identity threat. Unpublished manuscript: Leiden 1307–1310. University. Conrey, F. R., Sherman, J. W., Gawronski, B., Derks, B., Van Laar, C., & Ellemers, N. (2006). Hugenberg, K., & Groom, C. J. (2005). Striving for success in outgroup settings: Effects Separating multiple processes in implicit social of contextually emphasizing ingroup dimensions cognition: The quad model of implicit task on stigmatized group members’ social identity performance. Journal of Personality and Social and performance styles. Personality and Social Psychology, 89, 469–487. Psychology Bulletin, 32, 576–588. Crocker, J., & Major, B. (1989). Social stigma and Derks, B., Van Laar, C., & Ellemers, N. (2007). self-esteem: The self-protective properties of Social creativity strikes back: Improving stigma. Psychological Review, 96, 608–630. motivated performance of low status group Crocker, J., Major, B., & Steele, C. (1998). Social members by valuing ingroup dimensions. stigma. In D. T. Gilbert, S. T. Fiske, & G. Lindzey European Journal of Social Psychology, 37, 470–493. (Eds.), The handbook of social psychology (4th ed., Donchin, E. (1981). Surprise! Surprise? Vol. 2, pp. 504–553). Boston: McGraw-Hill. Psychophysiology, 18, 493–513. Croizet, J. C., & Claire, T. (1998). Extending the Ellemers, N., Spears, R., & Doosje, B. (Eds.). concept of stereotype and threat to social class: (1999). Social identity: Context, commitment, content. The intellectual underperformance of students Oxford, UK: Blackwell. from low socioeconomic backgrounds. Personality Ellemers, N., Spears, R., & Doosje, B. (2002). Self and Social Psychology Bulletin, 24, 588–594. and social identity. Annual Review of Psychology, Cullen, M. J., Hardison, C. M., & Sackett, P. R. 53, 161–186. (2004). Using SAT-grade and ability-job Ellemers, N., Van Dyck, C., Hinkle, S., & Jacobs, performance relationships to test predictions A. (2000). Intergroup differentiation in derived from stereotype threat theory. Journal of social context: Identity needs versus audience Applied Psychology, 89, 220–230. constraints. Social Psychology Quarterly, 63, 60–74. Cunningham, W. A., Johnson, M. K., Raye, C. L., Ellemers, N., & Van Rijswijk, W. (1997). Identity Gatenby, J. C., Gore, J. C., & Banaji, M. R. needs versus social opportunities: The use (2004). Separable neural components in the of group-level and individual-level identity processing of Black and White faces. Psychological management strategies. Social Psychology Science, 15, 806–813. Quarterly, 60, 52–65. Dar-Nimrod, I., & Heine, S. J. (2006). Exposure Ellis, L. K., Rothbart, M. K., & Posner, M. I. to scientifi c theories affects women’s math (2004). Individual differences in executive performance. Science, 314, 435. attention predict self-regulation and adolescent Davies, P. G., Spencer, S. J., Quinn, D. M., & psychosocial behaviors. Annals of the New York Gerhardstein, R. (2002). Consuming images: Academy of Sciences, 1021, 337–340. How television commercials that elicit stereotype Eriksen, B. A., & Eriksen, C. W. (1974). Effects threat can restrain women academically and of noise letters upon the identifi cation of a

177 Group Processes & Intergroup Relations 11(2)

target letter in a nonsearch task. Perception and Harmon-Jones, E., & Winkielman, P. (2007). Social Psychophysics, 16, 143–149. neuroscience: Integrating biological and psychological Falkenstein, M., Hoormann, J., Christ, S., & explanations of social behavior. New York: Guilford. Hohnsbein, J. (2000). ERP components on Hart, A. J., Whalen, P. J., Shin, L. M., McInerney, reaction errors and their functional signifi cance: S. C., Fischer, H., & Rauch, S. L. (2000). A tutorial. Biological Psychology, 51, 87–107. Differential response in the human amygdala Fiske, S. T. (2003). The discomfort index: How to to racial outgroup vs. ingroup face stimuli. spot a really good idea whose time has come. Neuroreport: For Rapid Communication of Psychological-Inquiry, 14, 203–208. Neuroscience Research, 11, 2351–2355. Forbes, C. E., Schmader, T., & Allen, J. J. B. (2008). Holroyd, C. B., & Coles, M. G. H. (2002). The role of devaluing and discounting in performance The neural basis of human error processing: monitoring. Unpublished manuscript, University Reinforcement learning, dopamine, and the of Arizona. error-related negativity. Psychological Review, Frantz, C. M., Cuddy, A. J. C., Burnett, M., Ray, H., 109, 679–709. & Hart, A. (2004). A threat in the computer: Inzlicht, M., Aronson, J., Good, C., & McKay, L. The race implicit association test as a stereotype (2006). A particular resiliency to threatening threat experience. Personality and Social Psychology environments. Journal of Experimental Social Bulletin, 30, 1611–1624. Psychology, 42, 323–336. Gaertner, S. L., & Dovidio, J. F. (1986). Prejudice, Inzlicht, M., & Ben-Zeev, T. (2000). A threatening discrimination, and racism: Problems, progress, intellectual environment: Why females are and promise. In J. F. Dovidio & S. L. Gaertner susceptible to experiencing problem-solving (Eds.), Prejudice, discrimination and racism defi cits in the presence of males. Psychological (pp. 315–332). New York: Academic Press. Science, 11, 365–371. Gehring, W. J., Coles, M. G. H., Meyer, D. E., Inzlicht, M., & Ben Zeev, T. (2003). Do high- & Donchin, E. (1995). A brain potential achieving female students underperform manifestation of error-related processing. In in private? The implications of threatening G. Karmos, M. Molnar, V. Csepe, I. Czigler, environments on intellectual processing. J. E. Desmedt (Eds.), Perspectives of event-related Journal of Educational Psychology, 95, 796–805. potentials research. Journal of Electroencephalography Inzlicht, M., & Gutsell, J. N. (2007). Running on and Clinical Neurophysiology (Suppl. 44), 287–296 empty: Neural signals for self-control failure. Gehring, W. J., Goss, B., Coles, M. G. H., Meyer, Psychological Science, 18, 933–937. D. E., & Donchin, E. (1993). A neural system for Inzlicht, M., Kang, S., & Fortune, K. (2008). error detection and compensation. Psychological Stereotype threat spillover: How threats to identify Science, 4, 385–390. affect eating, decision making, and the brain. Goldstein, A., Spencer, K. M., & Donchin, E. Unpublished manuscript, University of Toronto. (2002). The infl uence of stimulus deviance Inzlicht, M., McKay, L., & Aronson, J. (2006). and novelty on the P300 and novelty P3. Stigma as ego depletion: How being the target of Psychophysiology, 39, 781–790. prejudice affects self-control. Psychological Science, Gonzales, P. M., Blanton, H., & Williams, K. J. 17, 262–269. (2002). The effects of stereotype threat and Ito, T. A., Thompson, E., & Cacioppo, J. T. (2004). double-minority status on the test performance Tracking the timecourse of social perception: of Latino women. Personality and Social Psychology The effects of racial cues on event-related Bulletin, 28, 659–670. brain potentials. Personality and Social Psychology Good, C., Aronson, J., & Inzlicht, M. (2003). Bulletin, 30, 1267–1280. Improving adolescents’ standardized test Ito, T. A., & Urland, G. R. (2005). The infl uence of performance: An intervention to reduce the processing objectives on the perception of faces: effects of stereotype threat. Journal of Applied An ERP study of race and gender perception. Developmental Psychology, 24, 645–662. Cognitive, Affective and Behavioral Neuroscience, 5, Greenwald, A. G., McGhee, D. E., & Schwartz, 21–36. J. L. K. (1998). Measuring individual differences Ito, T. A., Urland, G. R., Willadsen Jensen, E., & in implicit cognition: The implicit association Correll, J. (2006). The social neuroscience of test. Journal of Personality and Social Psychology, stereotyping and prejudice: Using event-related 74, 1464–1480. brain potentials to study social perception.

178 Derks et al. neuroscience of stigma

In J. T. Cacioppo, P. S. Visser, & C. L. Pickett Marx, D. M., & Roman, J. S. (2002). Female (Eds.), Social neuroscience: People thinking about role models: Protecting women’s math test thinking people (pp. 189–212). Cambridge, performance. Personality and Social Psychology MA: MIT Press. Bulletin, 28, 1183–1193. Ito, T. A., Willadsen Jensen, E., & Correll, J. (2007). Marx, D. M., Stapel, D. A., & Muller, D. (2005). We Social neuroscience and social perception: New can do it: The interplay of construal orientation perspectives on categorization, prejudice, and and social comparisons under threat. Journal of stereotyping. In P. Winkielman & E. Harmon- Personality and Social Psychology, 88, 432–446. Jones (Eds.), Social neuroscience: Integrating McGlone, M. S., Aronson, J., & Kobrynowicz, D. biological and psychological explanations of social (2006). Stereotype threat and the gender gap behavior (pp. 401–421). New York: Guilford. in political knowledge. Psychology of Women Jamieson, J.P., & Harkins, S.G. (2007). Mere effort Quarterly, 30, 392–398. and stereotype threat performance effects. Journal Mendoza Denton, R., Downey, G., Purdie, V. J., of Personality and Social Psychology, 93, 544–564. Davis, A., & Pietrzak, J. (2002). Sensitivity to Jost, J. T., & Major, B. (2001). The psychology of status-based rejection: Implications for African legitimacy: Emerging perspectives on ideology, justice, American students’ college experience. Journal and intergroup relations. New York: Cambridge of Personality and Social Psychology, 83, 896–918. University Press. Milne, E., & Grafman, J. (2001). Ventromedial Kaiser, C. R., Vick, S. B., & Major, B. (2006). prefrontal cortex lesions in humans eliminate Prejudice expectations moderate preconscious implicit gender stereotyping. Journal of attention to cues that are threatening to social Neuroscience, 21, 1–6. identity. Psychological Science, 17, 332–338. Moran, J. M., Macrae, C. N., Heatherton, T. Kihlstrom, J. F. (2006). Does neuroscience F., Wyland, C. L., & Kelley, W. M. (2006). constrain social-psychological theory? SPSP Neuroanatomical evidence for distinct cognitive Dialogue, 21, 26–27. and affective components of self. Journal of Krendl, A. C., Richeson, J. A., Kelley, W. M., Cognitive Neuroscience, 18, 1586–1594. & Heatherton, T. F. (2008). The negative Mummendey, A., & Schreiber, H. J. (1984). consequences of threat: An fMRI investigation ‘Different’ just means ‘better’: Some obvious and of the neural mechanisms underlying women’s some hidden pathways to in-group favouritism. underperformance in math. Psychological Science, British Journal of Social Psychology, 23, 363–368. 19, 168–175. Muraven, M., & Baumeister, R. F. (2000). Self- Landreth, A., & Richardson, R. C. (2004). regulation and depletion of limited resources: Localization and the new phrenology: A review Does self-control resemble a muscle? Self, 126, essay on William Uttal’s the new phrenology. 247–259. Philosophical Psychology, 17, 107–123. Muraven, M., Tice, D. M., & Baumeister, R. Levin, S., & Van Laar, C. (Eds.). (2006). Stigma and F. (1998). Self-control as limited resource: group inequality: Social psychological perspectives. Regulatory depletion patterns. Journal of Mahwah, NJ: Erlbaum. Personality and Social Psychology, 74, 774–789. Major, B., & Schmader, T. (1998). Coping with Murphy, M. C., Steele, C. M., & Gross, J. J. (2007). stigma through psychological disengagement. Signaling threat: How situational cues affect In J. K. Swim & C. Stangor (Eds.), Prejudice: The women in math, science, and engineering target’s perspective (pp. 219–241). San Diego, CA: settings. Psychological Science, 18, 879–885. Academic Press. Nieuwenhuis, S., Aston Jones, G., & Cohen, J. D. Major, B., Spencer, S., Schmader, T., Wolfe, C., (2005). Decision making, the P3, and the locus & Crocker, J. (1998). Coping with negative coeruleus-norepinephrine system. Psychological stereotypes about intellectual performance: The Bulletin, 131, 510–532. role of psychological disengagement. Personality Nieuwenhuis, S., Ridderinkhof, K.R., Blom, J., and Social Psychology Bulletin, 24, 34–50. Band, G.P.H., & Kok, A. (2001). Error-related Martens, A., Johns, M., Greenberg, J., & Schimel, brain potentials are differentially related to J. (2006). Combating stereotype threat: The awareness of response errors: Evidence from an effect of self-affi rmation on women’s intellectual antisaccade task. Psychophysiology, 38, 752–760. performance. Journal of Experimental Social O’Brien, L. T., & Crandall, C. S. (2003). Stereotype Psychology, 42, 236–243. threat and arousal: Effects on women’s math

179 Group Processes & Intergroup Relations 11(2)

performance. Personality and Social Psychology Schmader, T., Johns, M., & Forbes, C. (in press). Bulletin, 29, 782–789. An integrated process model of stereotype threat Ochsner, K. N., & Lieberman, M. D. (2001). The effects on performance. Psychological Review. emergence of social cognitive neuroscience. Seibt, B., & Förster, J. (2004). Stereotype threat and American Psychologist, 56, 717–734. performance: How self-stereotypes infl uence Osborne, J. W. (1995). Academics, self-esteem, and processing by inducing regulatory foci. Journal of race: A look at the underlying assumptions of Personality and Social Psychology, 87, 38–56. the disidentifi cation hypothesis. Personality and Shapiro, J. R., & Neuberg, S. L. (2007). From Social Psychology Bulletin, 21, 449–455. stereotype threat to stereotype threats: Osborne, J. W. (1999). Unraveling Implications of a multi-threat framework for underachievement among African American causes, moderators, mediators, consequences, boys from an identifi cation with academics and interventions. Personality and Social Psychology perspective. Journal of Negro Education, 68, Review, 11, 107–130. 555–565. Sherman, D. K., Kinias, Z., Major, B., Kim, H. S., & Oskamp, S. (Ed.). (2000). Reducing prejudice and Prenovost, M. (2007). The group as a resource: discrimination. Mahwah, NJ: Erlbaum. Reducing biased attributions for group success Phelps, E. A., O’Connor, K. J., Cunningham, W. A., and failure via group affi rmation. Personality and Funayama, E. S., Gatenby, J. C., Gore, J. C. Social Psychology Bulletin, 33, 1100–1112. et al. (2000). Performance on indirect measures Snow, C. P. (1959). Two cultures. Cambridge, UK: of race evaluation predicts amygdala activation. Cambridge University Press. Journal of Cognitive Neuroscience, 12, 729–738. Somerville, L. H., Heatherton, T. F., & Kelley, Pronin, E., Steele, C. M., & Ross, L. (2004). Identity W. M. (2006). Anterior cingulate cortex bifurcation in response to stereotype threat: responds differentially to expectancy violation Women and mathematics. Journal of Experimental and social rejection. Nature Neuroscience, 9, Social Psychology, 40, 152–168. 1007–1008. Quinn, D. M., Kahng, S. K., & Crocker, J. (2004). Spencer, S. J. (2003). Media images and stereotype Discreditable: Stigma effects of revealing a threat: How activation of cultural stereotypes can mental illness history on test performance. undermine women’s math performance. Paper Personality and Social Psychology Bulletin, 30, presented at the 4th Annual Meeting of the 803–815. Society of Personality and Social Psychology, Richeson, J. A., Baird, A. A., Gordon, H. L., Los Angeles, CA. Heatherton, T. F., Wyland, C. L., Trawalter, S. Spencer, S. J., Steele, C. M., & Quinn, D. M. et al. (2003). An fMRI investigation of the (1999). Stereotype threat and women’s math impact of interracial contact on executive performance. Journal of Experimental Social function. Nature Neuroscience, 6, 1323–1328. Psychology, 35, 4–28. Scheepers, D., & Ellemers, N. (2005). When the Steele, C. M. (1992). Race and the schooling of pressure is up: The assessment of social identity Black Americans. Atlantic Monthly, 269, 68–78. threat in low and high status groups. Journal of Steele, C. M. (1997). A threat in the air: How Experimental Social Psychology, 41, 192–200. stereotypes shape intellectual identity and Scheepers, D., Spears, R., Doosje, B., & Manstead, performance. American Psychologist, 52, 613–629. A. S. R. (2006). Diversity in in-group bias: Steele, C. M., & Aronson, J. (1995). Stereotype Structural factors, situational features, and threat and the intellectual test performance social functions. Journal of Personality and Social of African Americans. Journal of Personality and Psychology, 90, 944–960. Social Psychology, 69, 797–811. Schmader, T. (2002). Gender identifi cation Steele, C. M., Spencer, S. J., & Aronson, J. (2002). moderates stereotype threat effects on women’s Contending with group image: The psychology math performance. Journal of Experimental Social of stereotype and social identity threat. In M. Psychology, 38, 194–201 P. Zanna (Ed.), Advances in experimental social Schmader, T., & Johns, M. (2003). Converging psychology (Vol. 34, pp. 379–440). San Diego, evidence that stereotype threat reduces working CA: Academic Press. memory capacity. Journal of Personality and Social Stone, J., Lynch, C. I., Sjomeling, M., & Darley, J. Psychology, 85, 440–452. M. (1999). Stereotype threat effects on black and

180 Derks et al. neuroscience of stigma

white athletic performance. Journal of Personality monitoring and the error-related negativity. and Social Psychology, 77, 1213–1227. Psychological Review, 111, 931–959. Stricker, L. J., & Ward, W. C. (2004). Stereotype Zanna, M. P., & Olson, J. M. (Eds.). (1994). The threat, inquiring about test takers’ ethnicity and psychology of prejudice: The Ontario symposium gender, and standardized test performance. (Vol. 7). Hillsdale, NJ: Erlbaum. Journal of Applied Social Psychology, 34, 665–693. Swim, J. K., & Stangor, C. (Eds.). (1998). Prejudice: The target’s perspective. San Diego, CA: Academic Biographical notes Press. belle derks is an assistant professor in social Tajfel, H., & Turner, J. C. (1986). The social and organization psychology at Leiden identity theory of intergroup behavior. In S. University. Her research focuses on the effects Worchel & W. G. Austin (Eds.), The psychology of of stigma and social identity threat on the well- intergroup relations (pp. 7–24). Chicago: being, performance motivation and mediating Nelson-Hall. physiological responses (EEG, ECG, ICG) of Uttal, W. R. (2002). Precis of the new phrenology: The members of socially devalued groups. She limits of localizing cognitive processes in the brain. was recently awarded the 2007 Social Issues Netherlands: Kluwer Academic Publishers. Dissertation Award from the Society for the Verkuyten, M., & Thijs, J. (2004). Psychological Psychological Study of Social Issues (SPSSI). disidentifi cation with the academic domain among ethnic minority adolescents in the michael inzlicht is an assistant professor in Netherlands. British Journal of Educational the department of psychology at University of Psychology, 74, 109–125. Toronto Scarborough. He completed his BSc in Walton, G. M., & Cohen, G. L. (2007). A question anatomical sciences at McGill University, his PhD of belonging: Race, social fi t, and achievement. in experimental psychology at Brown University, Journal of Personality and Social Psychology, 92, and his postdoctoral fellowship in applied 82–96. psychology at New York University. He conducts Weisberg, D. S., Keil, F. C., Goodstein, J., Rawson, basic research on prejudice, stigma, and self- E., & Gray, J. R. (2008). The seductive allure of regulation, and has begun using methods and neuroscience explanations. Journal of Cognitive techniques borrowed from neuroscience and Neuroscience, 20, 470–477. psychophysiology to understand them more Wheeler, M. E., & Fiske, S. T. (2005). Controlling fully. He was recently honored by being named racial prejudice: Social-cognitive goals affect the 2006 Louise Kidder Early Career Award amygdala and stereotype activation. Psychological recipient. Science, 16, 56–63. Wilson, E. 0. (1998). Consilience: The unity of sonia kang is a PhD student in the department knowledge. NewYork: Knopf. of psychology at the University of Toronto. She Wraga, M., Helt, M., Jacobs, E., & Sullivan, K. completed her BSc (Hons) at the University (2007). Neural basis of stereotype-induced shifts of Alberta. She conducts basic and applied in women’s mental rotation performance. Social research on the experience of prejudice and Cognitive and Affective Neuroscience, 2, 12–19. discrimination across the lifespan from early Yeung, N., Botvinick, M. M., & Cohen, J. D. (2004). childhood to old age. The neural basis of error detection: Confl ict

181