Praise for Intelligence Can Undermine Motivation and Performance (with the Li & Bates 2019 replication attempt)
Mueller, C. M., & Dweck, C. S. (1998); Li, Y., & Bates, T. C. (2019) · 1998
grade Creplicationdeveloper-involvedmixed
Sample
6 lab experiments (~400 children); replication 3 preregistered studies (n=623)
Population
US 5th graders (original); Chinese 9-13-year-olds (replication).
Design
Intelligence-praise vs effort-praise after success, then failure, then retest. Li & Bates: 3 preregistered direct replications; Yeager & Dweck's reanalysis contests the failure; Bates rebuts. Unresolved.
Key findings
Person/intelligence praise is at best useless and plausibly harmful — but the famous demonstration failed its only large direct replication, and the dispute is unresolved. Treat as a zero-cost 'do not' (praise process/work, not the person), not as a robust effect size.
Genetic confound
Randomized; the null mindset-attainment association is consistent with attainment tracking stable heritable ability rather than malleability beliefs.
Replication notes
The praise-backfire direction is plausible (converges with Kluger & DeNisi praise cue d=0.09) but the magnitude is fragile and replication-contested.
DOI / URL
10.1037/0022-3514.75.1.33 ; 10.1037/xge0000669
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Post-failure performance: intelligence vs effort praise (original) | direction | intelligence-praised performed worse, chose easier tasks, lied about scores more | researcher-designed | immediate; lab | active-alternative | end-of-treatment | non-cognitive |
| Direct replication (n=623) | per-study significance | no significant effect in each study as analyzed; aggregate reanalysis contested | researcher-designed | immediate; lab | active-alternative | end-of-treatment | non-cognitive |
| Mindset beliefs -> attainment/IQ/failure response (replication) | association | null | mixed | cross-sectional | none | not-applicable | attainment |
Cited by
- Does teaching a growth mindset raise achievement?mixedconf: highgc: low
- Feedback and formative assessmentmixedconf: highgc: low