These studies focused on whether the subtlety of stereotype activation influences how individuals are affected by positive stereotypes. In Experiment 1, Asian-American students completed a survey about their entertainment preferences (control), one focusing on their ethnic identity without invoking stereotypes (subtle activation), or a questionnaire that asked about their beliefs regarding the commonality and validity of stereotypes associated with Asian-Americans (blatant activation). Performance on a subsequent math task was better in the subtle activation compared with the control and blatant stereotype conditions. Although one might expect better performance when a positive stereotype is invoked, these results suggest that blatant instantiation might increase pressures and thoughts related to meeting positive expectations that interfere with effective performance. In Experiment 2, Asian-American and non-Asian students were exposed either to Asian stereotypes or control words on a computer that were presented either subliminally (without conscious awareness) or supraliminally (with awareness). Asian-American students performed better on a subsequent math test if they had been exposed subliminally to stereotypes rather than to control words. However, their performance was harmed compared with the control condition if the stereotypic words were consciously detected. For non-Asians, performance was improved compared with the control condition if the stereotypic words were presented supraliminally, but not when presented subliminally. These findings suggest that group membership and the manner of stereotype activation interact to affect performance. Whereas group members for whom the stereotype is relevant exhibited assimilation to the stereotype when it was activated in a subtle fashion but contrast away from the stereotype when it was activated blatantly, groups for whom the stereotype was irrelevant showed only assimilation under blatant activation.