Revisiting Rydell, McConnell, & Beilock (2009) : a case for the inclusion of non-targets of stereotype threat

This study sought to examine the role of multiple identities as a possible protective factor against stereotype threat for females taking a difficult math test. Specifically, it sought to replicate the findings of Rydell, McConnell, and Beilock (2009), who found that making a positively stereotyped identity salient (college) at the same time a negatively stereotyped identity (female) was salient, buffered the effects of stereotype threat. This study also attempted to evaluate the validity of a common experimental stereotype threat manipulation, which is to make explicit statements about performance which remind test subjects of existing stereotypes. Using a quantitative experimental design, and replicating the methodology used in the 2009 study, math-identified college students were randomly assigned to take a difficult math test under circumstances which varied salient identities. For the experimental conditions, an explicit statement was made about prior performance by either females or females and college students. For math-identified females, the statement about female performance was believed to invoke a negative stereotype about math ability and thus stereotype threat. However, when the statement was about both their gender as well as their college identity (thought to be positively stereotyped), this would cause the females to suppress their gender identity in order to maintain positive self-esteem and thus would be protected from stereotype threat effects. It was also predicted that non-targets of threat (males) would not be affected by the manipulations, as according to the theory of stereotype threat, a stereotype has to be self-relevant to become a threat. Results failed to replicate the findings of the previous study. While not significant, females actually trended towards better performance when reminded of the negative stereotype about females, as compared to a control group. More importantly, this type of manipulation was shown to significantly affect non-targets of threat, which is a violation of stereotype threat theory. When reminded of the negative stereotype about females, males performed significantly worse than a control group. This evidence supports the idea that making explicit statements about ability is an invalid method of invoking stereotype threat in an experimental setting.