ReducingStereotypeThreat.org

Sekaquaptewa & Thompson, 2003

This experiment compared performance of women under solo versus non-solo status conditions under conditions designed to create or undermine stereotype threat. White male and female undergraduates learned math-related information that was later to be shared with a group. Students were told either that the material was "traditional math" (stereotype threat for women) or were assured that "men and women perform equally well on this type of material" (control). After studying the information, students then were asked to take turns answering questions about the information in the presence of three men or three women. Only women performed more poorly when they were solo (i.e., reported their answers to three men) than when they non-solo (i.e., they reported their answers to women). Only women also performed more poorly when the task was described as related to math ability compared with when there were supposedly no gender differences. Although solo status appeared to reduce performance by decreasing women's expectations of success, reduced expectations did not appear to account for poorer performance under stereotype threat. These results show that solo status and stereotype threat both contribute to reduced performance of women compared with men. Women's performance was poorest when both factors were present and best when both were absent.