Beyond the Immediate Task: Durability and Everyday Relevance of Process and Person Praise in Children: A Systematic Review and Stratified Meta-Analysis
Does praising effort instead of ability build lasting motivation in children's daily lives? Or does the famed benefit depend on scripted laboratory setbacks pitted against fragile ability labels? Adhering to PRISMA 2020 standards, we reconciled 36 legacy bibliographic exports (1,870 unique APA records, with 789 topic candidates adjudicated at full text) against a six-database harvest across OpenAlex, Crossref, Semantic Scholar, ERIC, Europe PMC/PubMed, and arXiv (2,495 deduplicated records) plus first-hand inspection of 74 primary reports and online supplements spanning 1998 to 2026. Cross-study family linkage resolved 42 quantitative study units into 37 study families and k = 41 strictly independent cohorts (N = 5,088 analytic participants; 5,030 unique youths; 133 primary effect sizes alongside cluster-confounded sensitivity rows). Restricted maximum likelihood (REML) random-effects models separated temporal and ecological distance (Subgroup A), comparator wording (Subgroup B), and endpoint realism (Subgroup C). Incompatible estimands were never pooled. During same-session laboratory tasks after staged failures, process or nongeneric praise beat person or generic praise (k = 23, g = +0.68, 95% CI [+0.43, +0.93], I² = 85%), yet funnel asymmetry was pronounced (Egger p = .050; PEESE-adjusted g = +0.26), preregistered replication was null (g = +0.05), and older youths occasionally showed attributional or behavioral backfire. Against neutral or objective feedback, immediate process praise yielded no gain across 16 cohorts (N = 2,229, g = +0.11, 95% CI [−0.04, +0.27], p = .160; mixed-effects comparator test QM(2) = 11.39, p = .0034, R² = 23.7%). Person-praise harm drove those three-arm gaps. Post-withdrawal behavioral follow-ups 10 to 21 days later rested on two small samples marred by asymmetric memory exclusions (g = +0.46, 95% CI [−0.01, +0.94], p = .058). In naturalistic caregiver speech (k = 10, N = 1,295), overall correlations were weak (r = +0.10, Knapp–Hartung 95% CI [−0.001, +0.204]) and dropped to zero across all four multi-wave prospective home and diary cohorts spanning 6 months to 7 years (N = 525, r = +0.05, 95% CI [−0.04, +0.14], p = .289), whether predicting growth mindset questionnaires (k = 4, N = 722, r = +0.01, p = .775) or standardized achievement and IQ (k = 3, N = 310, r = +0.04, p = .548). Adults should drop fixed ability labels during setbacks rather than expecting effort praise alone to lift long-term trajectories.
Authors
- Alexander Kuznetsov (ORCID: https://orcid.org/0009-0002-6652-3544)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-20
- DOI
- https://doi.org/10.21203/rs.3.rs-11795652/v1
- Primary Topic
- Education, Achievement, and Giftedness
- Type
- preprint