Self-regulation and metacognition correlate strongly with mathematics achievement (r = .42 and .37)¶

Claim · All claims
Evidence · 1 study · review r3 · q3 · i3 large · n=30 meta-analyses, 66 correlations from 13,853 effect sizes (~4.66 million participants)
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Self-regulation correlates r = .42 with mathematics achievement in children (Robson et al. 2020, k = 22) and metacognition .37 in adolescents (Muncer et al. 2022, k = 74). These are correlations compiled across meta-analyses: they describe how strongly the characteristic goes with mathematics achievement, not what causes it.

Subclaims¶

q3 i3 Self-regulation correlates r = .42 with mathematics achievement in children (Robson et al. 2020, k = 22) and metacognition .37 in adolescents (Muncer et al. 2022, k = 74). → Breit et al. 2025

Evidence¶

Breit et al. 2025¶

Breit, M., Schneider, M., & Preckel, F. (2025). Mathematics achievement and learner characteristics: A systematic review of meta-analyses. Learning and Individual Differences, 118, 102621. doi:10.1016/j.lindif.2024.102621

q3 · systematic review of meta-analyses · i3 · large effect, r = .42 and .37 · n=30 meta-analyses, 66 correlations from 13,853 effect sizes (~4.66 million participants) · review · r3

The review compiles zero-order correlations from published meta-analyses (each with at least ten effect sizes, the largest one kept where several covered a variable), found by a systematic PsycINFO search in July 2024. It pools nothing further and controls for nothing: many of these characteristics correlate with one another, so a correlation here says how strongly the characteristic goes with mathematics achievement, not that it causes it. Self-regulation r = .42 (Robson et al. 2020, k = 22, n = 7,076 children), stronger in younger samples; metacognition r = .37 (Muncer et al. 2022, k = 74, n = 570,575 adolescents, I² = .99). Both correlate more strongly with mathematics than in Hattie's (2023) cross-domain synthesis.

Discussion¶

The authors call these promising targets because they are malleable, citing a separate meta-analysis of self-regulation interventions (Pandey et al. 2018, d = 0.42); that intervention evidence is not part of this review.

Learner Variables¶

• Metacognition — predictor: learners who differ on it differ in outcomes

Related Claims¶

• Mathematics achievement correlates more strongly with learners' knowledge, skills and cognitive abilities than with their personality traits, goals or emotions
• Teacher-observed emotion regulation correlates with academic outcomes more strongly than performance-tested ERA does — related
• Numerical metacognition, but not emotion metacognition, predicts children's standardized mathematics achievement — a narrower finding that bears on this claim
• Metacognition is significantly positively correlated with math performance in adolescence (pooled r = .37) — possibly the same claim (merge candidate)
• Mathematics self-concept correlates strongly with mathematics achievement (r = .43), while verbal self-concept barely does (r = .12) — related