A review and meta-analysis of age-based stereotype threat: negative stereotypes, not facts, do the damage

Abstract

Stereotype threat effects arise when an individual feels at risk of confirming a negative stereotype about their group and consequently underperforms on stereotype relevant tasks (Steele, 2010). Among older people, underperformance across cognitive and physical tasks is hypothesized to result from age-based stereotype threat (ABST) because of negative age-stereotypes regarding older adults' competence. The present review and meta-analyses examine 22 published and 10 unpublished articles, including 82 effect sizes (N = 3882) investigating ABST on older people's (Mage = 69.5) performance. The analysis revealed a significant small-to-medium effect of ABST (d = .28) and important moderators of the effect size. Specifically, older adults are more vulnerable to ABST when (a) stereotype-based rather than fact-based manipulations are used (d = .52); (b) when performance is tested using cognitive measures (d = .36); and (c) occurs reliably when the dependent variable is measured proximally to the manipulation. The review raises important theoretical and methodological issues, and areas for future research.

Figure 1

Funnel plot of P1 mean difference effect sizes (“Std diff in means”) plotted against the standard error of the mean difference effect sizes (“Standard Error”). Diagonal lines show 95% confidence limits and a vertical line shows the population effect size. Note: the two largest effect sizes (one positive and one negative) have not been included in the funnel plot due to their larger standard errors requiring a smaller display.