Much ado about grit: A meta-analytic synthesis of the grit literature.

584 effect sizes from 88 independent samples representing 66,807 individuals

Middle-school age and above, predominantly US academic and military samples.

Psychometric meta-analysis with correction for unreliability, followed by hierarchical regressions on the meta-analytic correlation matrix to estimate incremental validity. Every grit measure in the corpus is self-report; the criteria are registrar GPA and retention records.

Grit correlates with conscientiousness at rho = .84 (k = 22, N = 18,826) - the two constructs are not distinguishable. Grit correlates with cognitive ability at rho = .05 and with academic performance at rho = .18. The facets diverge sharply - perseverance of effort correlates .26 with academic performance while consistency of interest correlates .10, which undermines treating them as one trait. The decisive numbers are the incremental-validity increments over conscientiousness - overall grit adds .000 to overall academic performance, .002 to high-school GPA and .004 to college GPA. Reversing the order, conscientiousness added after grit contributes .180, .160 and .170. The authors conclude that interventions designed to enhance grit may only have weak effects on performance and success.

High for the underlying correlations, which are all observational. The meta-analysis is a synthesis of confounded associations and inherits the confounding.

Replicated by Lam and Zhou 2022 on a fifteen-times-larger corpus (r = .19 overall, perseverance .21, consistency .08) and converged on by Schmidt et al. 2018 using confirmatory factor analysis rather than meta-analysis.