An investigation of stereotype threat in employment tests
Three experiments explored the existence of stereotype threat in employee selection settings. In Study 1, researchers analyzed performance on a cognitive skills test from a large pool of applicants (N = 49,150) for sales positions at major financial services firm in the United States. Two versions of the test were created so that demographic information was requested either at the beginning or at the end of the test. Effect sizes were small, but performance tended to increase when demographics were collected after the test had been completed. Female ethnic minorities showed the greatest increase in scores when demographics were collected at the end of the testing session. Study 2 served as a replication in a controlled environment with undergraduate students (N = 289) using the same cognitive skills test. Students completed the test under one of three conditions, being asked to indicate: 1) their name and demographics prior to the test; 2) their name prior to the test but their demographics after the test, or; 3) their name and demographic information after the test. There emerged no differences in performance between these three conditions. In Study 3, students (N = 145) from a 2-year community college were given a different cognitive skills employment test in which stereotype threat was manipulated through the timing of demographic collection and the description of the test (either diagnostic or nondiagnostic of ability). Scores improved overall when demographic information was collected at the end of the study, although this effect tended to emerge for all groups. The authors concluded that while these results do not point to stereotype threat specifically since effects did not vary substantially for different ethnic and gender groups who presumably would experience different levels of stereotype threat. Nonetheless, the results indicate that collecting demographic information prior to testing can negatively affect test performance.