Stereotype Threat

Definition

Stereotype threat is defined as a situational predicament: the risk of confirming, as self-characteristic, a negative stereotype about a group one belongs to.

The claim is situational rather than dispositional. The same person performs differently depending on whether the stereotype is made relevant, which is what made the idea attractive: it located a cause of performance gaps in the environment rather than in the individual.

History

Steele and Aronson's 1995 studies with African American college students on a verbal test are the origin. Spencer, Steele and Quinn extended it to women and mathematics in 1999.

The idea moved into education policy, diversity training and public discourse very quickly and on a very thin evidential base by current standards.

Theory

Proposed mechanisms include working memory depletion by intrusive worry, physiological stress responses, monitoring for evidence of the stereotype, and shifts in motivation toward avoiding failure.

The working memory account is the most developed and predicts that threat effects should be largest on tasks with high executive demand, which some studies support.

A theoretical weakness is that the mechanism is usually inferred from the performance effect rather than measured independently.

Research

The current evidential picture is much weaker than the reputation.

• Meta-analyses that adjust for publication bias find substantially reduced effects, in some analyses close to zero.
• Large preregistered replications, including multi-lab efforts, have largely failed to find the effect in the women-and-mathematics paradigm.
• The original Steele and Aronson design used covariate adjustment for prior SAT scores; the unadjusted difference is much less striking, and this feature of the analysis is rarely reported in secondary accounts.
• Some field studies of interventions derived from the theory have shown effects, and some have not. Values-affirmation interventions in schools have a mixed record with several well-powered nulls.

The defensible summary: the effect may exist under some conditions in some populations, and the confident general claim made throughout the 2000s is not supported.

Applications

Educational assessment design, identity-safe classroom practice, and interventions intended to narrow achievement gaps.

The design implications that remain reasonable — not making group membership salient immediately before an assessment, providing accurate normative information — are cheap and low-risk regardless of the size of the underlying effect.

Criticism

• Replication failures are substantial and include high-powered preregistered attempts.
• Publication bias is demonstrable in the literature.
• Policy adoption ran far ahead of evidence, and the idea was taught as settled while the evidence base was small and selectively reported.
• Small original samples with flexible analysis choices.
• The construct risks becoming unfalsifiable when nulls are explained by unmeasured moderators such as domain identification.

References

• Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69(5), 797-811.
• Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and women's math performance. Journal of Experimental Social Psychology, 35(1), 4-28.
• Flore, P. C., & Wicherts, J. M. (2015). Does stereotype threat influence performance of girls in stereotyped domains? Journal of School Psychology, 53(1), 25-44.
• Shewach, O. R., Sackett, P. R., & Quint, S. (2019). Stereotype threat effects in settings with features likely versus unlikely in operational test settings. Journal of Applied Psychology, 104(12), 1514-1534.