Glossary
Stereotype threat
What is stereotype threat?
Stereotype threat concerns the risk of being judged through, or confirming, a negative stereotype about your group. Researchers study whether that situation affects performance.
What the research found
Original verbal-test studies
In Steele and Aronson’s first two studies, participants took a difficult verbal test whose performance was described as either diagnostic or nondiagnostic of ability. With prior SAT scores controlled, Black participants performed worse than White participants in the diagnostic condition, but not in the nondiagnostic condition. This was an adjusted comparison under specific experimental instructions. It does not show that changing instructions eliminates unadjusted population differences.
Later cognitive-testing evidence
Later evidence limits the broader claim. Shewach, Sackett, and Quint’s meta-analysis examined test measures, threat cues, and comparison conditions to distinguish studies more relevant to operational testing. In its focal analysis of 45 samples and 3,532 participants, the estimated effect was d = −0.14, a small standardized performance difference. A smaller set of actual operational studies and studies using motivational incentives produced estimates from 0.00 to −0.14. The authors also found evidence of publication bias. These estimates concern cognitive testing, not the absence of every harmful experience involving stereotypes.
Why it matters
A laboratory finding does not establish the cause of every performance gap in admissions or hiring. A poor interview result also cannot tell us whether stereotype threat affected the candidate.
A proposed mechanism
Working-memory accounts propose that concern about a stereotype can interfere with the mental processing needed for a task. Schmader and Johns (2003) tested this account in three experiments and reported working-memory reductions under their threat manipulations, with mediation of women’s math performance in the third study. This is evidence for a mechanism under those conditions, not a way to infer why a particular applicant performed poorly.