The influence of stereotypes on decisions to shoot

Using a videogame to simulate encounters with potentially hostile targets, three studies tested a model in which racial bias in shoot/don’t-shoot decisions reflects accessibility of the stereotype linking Blacks to danger. Study 1 experimentally manipulated the race-danger association by asking participants to read newspaper stories about Black (vs. White) criminals. As predicted, exposure to stories concerning Black criminals increased bias in the decision to shoot. Studies 2 and 3 manipulated the number of White and Black targets with and without guns in the context of the videogame itself. As predicted, frequent presentation of stereotypic (vs. counterstereotypic) targets exacerbated bias (Study 2) and—consistent with our process account—rendered stereotypes more accessible (Study 3). Copyright # 2007 John Wiley & Sons, Ltd. Social psychology has long been concerned with the idea that social categories like race, gender, and social class can influence people’s interpretations of, and reactions to, ambiguous stimuli (Allport, 1954; Campbell, 1967; Duncan, 1976; Kunda & Sherman-Williams, 1993; Wittenbrink, Judd, & Park, 1997). In the domain of criminal justice, category-based judgments can have profound consequences through slow, deliberative judgments like those made by a jury (Eberhardt, Davies, Purdie-Vaughns, & Johnson, 2006; Sommers & Ellsworth, 2000) but also through the spontaneous, split-second reactions of a police officer (Correll, Park, Judd, &Wittenbrink, 2002; Correll, Park, Judd, Wittenbrink, Sadler & Keesee, 2007; Payne, 2001; Plant, Peruche, & Butz, 2005). In particular, research using computer simulations to investigate the influence of race on shoot/don’t-shoot decisions demonstrates a pronounced bias to shoot Blacks. In our paradigm, participants perform a videogame task in which they encounter armed and unarmed targets who are either Black or White. Participants attempt to shoot armed targets and indicate don’t shoot in response to unarmed targets. Though race is irrelevant to this task, participants are faster and more likely to shoot Black targets. They are also faster and more likely to indicate don’t shoot for Whites. In essence, participants respond quickly and accurately when targets nt of Psychology, University of Chicago, 5848 South University, Green 415, hicago.edu s, Ltd. Received 22 November 2006 Accepted 22 May 2007 Stereotypes and decisions to shoot 1103 conform to cultural stereotypes (armed Blacks, unarmed Whites), but respond slowly and inaccurately when targets violate those stereotypes (unarmed Blacks, armed Whites). To account for these effects, we have suggested that stereotypes systematically bias reactions to stimuli in the shoot/don’t-shoot task. It is well established that priming a social category facilitates reactions to stimuli that are congruent with the category’s stereotype relative to stimuli that are stereotype incongruent (Dovidio, Evans, & Tyler, 1986; Fazio, Jackson, Dunton, & Williams, 1995; Wittenbrink et al., 1997). Priming may also bias perception, such that ambiguous stimuli are interpreted in a manner consistent with the primed category (Devine, 1989; Gilbert & Hixon, 1991; Kunda & Sherman-Williams, 1993). In the present research, we examine the role of stereotypes in shaping responses to guns and non-gun objects in the context of a shoot/don’t-shoot task. We are particularly concerned with one component of the stereotype about Blacks in the United States (where nearly all of this research has been conducted), namely, the idea of danger. Though cultural stereotypes about Blacks include a number of other concepts such as poverty and athleticism, danger figures prominently in the Black stereotype (Devine & Elliot, 1995; Wittenbrink et al., 1997), and perceptions of danger bear directly on the decision task in these studies, which involves the detection of a hostile target. To the extent that Black targets seem more dangerous than Whites, they should promote a tendency to shoot, facilitating correct responses for armed targets but inhibiting correct responses for unarmed targets. A core feature of our prediction, then, is that stereotypes linking Blacks to danger promote racial bias in the decision to shoot. Existing research provides some support for this hypothesis. First, participants who report greater awareness of stereotypes about aggression, violence, and danger showmore extreme bias in our task (Correll et al., 2002, Study 3). Second, neuroscientific data suggest that target race affects threat perception, which in turn predicts the magnitude of bias in the decision to shoot. The P200 is a component of the event-related brain potential waveform that has been associated with orientation to evolutionarily significant stimuli, including threats (Carretie, Martin-Loeches, Hinojosa, & Mercado, 2001). Among White participants, P200’s to Blacks are typically more pronounced than P200’s to Whites (Ito & Urland, 2003) and this difference increases as a function of stereotype strength (Correll, Urland, & Ito, 2006). Critically, the magnitude of racial differentiation in the P200 predicts bias in the decision to shoot. Participants who demonstrate a strong P200 threat response to Black targets (relative to Whites) are especially quick to shoot armed Black targets and especially quick to choose don’t shoot for unarmed Whites. The current studies examine whether experimental manipulations designed to increase the accessibility of the Black-danger stereotype exacerbate bias in the decision to shoot. This research borrows from work on the malleability of implicit associations. For example, Blair, Ma, and Lenton (2001) asked participants to imagine stereotypic (or counterstereotypic) women and found that this manipulation increased (or decreased) gender bias on the implicit associations test (IAT; Greenwald, McGhee, & Schwartz, 1998). Similarly, Dasgupta and Asgari (2004), Dasgupta and Greenwald (2001), Karpinski and Hilton (2001), and Han, Olson, and Fazio (2006) manipulated exposure to information that reinforced (or undermined) common associations, including stereotypes, and showed that these manipulations affected performance on implicit measures. We employ similar manipulations to garner evidence that stereotypes about danger causally affect bias in the decision to shoot. Notably, given the public relevance of these effects, we employ manipulations with real-world analogs that may plausibly affect the accessibility of the stereotype of Blacks as dangerous: newspaper articles and base rates in the environment. In Study 1, we expose participants to newspaper stories about violent criminals. Describing those criminals as Black (rather than White) should reinforce stereotypic associations between Blacks and danger and, accordingly, amplify the tendency to shoot Black (rather than White) targets. In Study 2, we manipulate the covariation between race and weapons in the context of the videogame task, such that in one condition Blacks are disproportionately likely to be armed and Whites are Copyright # 2007 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 37, 1102–1117 (2007)