Content area
Full Text
Stereotype lift occurs when non-targets perform better in a stereotype-relevant testing situation compared to a testing situation that is less stereotype-relevant. The boundary conditions of this effect, however, are not well understood. To this end, the present research focuses on one critical moderator of stereotype lift, namely activation of non-targets' social self in stereotype-relevant (i.e., diagnostic) testing situations. Results showed that simply describing a test as diagnostic of ability did not lead to strong lift effects unless the social self was also activated (either indirectly or directly): non-targets need more "pushing" to activate their social self in diagnostic testing situations because they are not threatened by a negative stereotype. In contrast, we found that the test diagnosticity manipulation was enough to cause stereotype threat because targets need less pushing to activate their social self and the associated negative stereotype in diagnostic testing situations.
Considerable research has focused on the adverse effects of negative stereotypes on targets' test performance in stereotype-relevant situations (i.e., stereotype threat; Steele, 1997; Steele & Aronson, 1995). In contrast, far less work has focused on the test performance boost of non-targets in those same settings, for example, when male participants' math test performance is higher in a diagnostic testing situation compared to a nondiagnostic one (i.e., stereotype lift; Walton & Cohen, 2003). Although both stereotype threat and stereotype lift effects may occur within the same testing situations little is known about how and when lift occurs relative to what is known about stereotype threat. Moreover, it is interesting to note that lift effects are not always found in experiments examining stereotype-based performance (e.g., Gonzales, Blanton, & Williams, 2002; Steele & Aronson, 1995), and when lift is found, some experiments show stronger lift effects than others. Thus, an important question is why do some experiments find stronger lift effects (e.g., Spencer, Steele, & Quinn, 1999, Experiment 2) than other experiments using similar participant populations (e.g., Gonzales et al., 2002)?1 In the present research we argue that activation of one's social identity (i.e., the social self; see also Brewer & Gardner, 1996; Bromgard, Trafimow, & Bromgard, 2006; Stapel & Koomen, 2001)-particularly the social self or social identity that is most relevant to the current situation and task-plays a central role in determining how...