Reminding people of a negative stereotype reliably lowers their test scores.
Verdict: contested
The original effect is real in some settings but small, inconsistent, and shadowed by publication bias in replications.
What the evidence shows
Steele & Aronson (1995) reported that Black students scored worse on a test framed as measuring ability — presumably burdened by the fear of confirming a stereotype. The concept became hugely influential in education and diversity policy.
Hundreds of studies followed, but replications are uneven. A meta-analysis of the effect in girls' maths performance (Flore & Wicherts, 2015) found a small average effect and clear signs of publication bias, and some large pre-registered studies find little or nothing. The phenomenon likely exists under specific conditions, but it is smaller and more fragile than the confident classroom version suggests; treating it as a large, universal cause of achievement gaps is contested.
Sources
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
Black participants scored lower when a test was described as diagnostic of ability than when it was not.
DOI: 10.1037/0022-3514.69.5.797 →Flore, P. C., & Wicherts, J. M. (2015). Does stereotype threat influence performance of girls in stereotyped domains? A meta-analysis.
Journal of School Psychology, 53(1), 25–44
The average effect on girls' maths performance was small, with evidence of publication bias inflating the literature.
DOI: 10.1016/j.jsp.2014.10.002 →