<b>Self-Regulated Learning in Mathematical Problem Solving: A Systematic Review of Metacognition, Motivation, Emotion, and Learning Strategies</b>

Source abstract

Mathematical problem solving requires learners to coordinate cognitive, motivational, emotional, and strategic regulation, yet these components are often examined separately. This thematic systematic review aimed to characterize empirical research on self-regulated learning (SRL) in a mathematical problem solving (MPS), map four regulatory domains, and synthesize their interactions. Searches were conducted on 28–30 April 2026 in Scopus, ScienceDirect, SpringerLink, Taylor & Francis Online, and ERIC for English-language primary empirical journal articles published in 2018–2026. Reporting followed PRISMA 2020 and PRISMA-S, methodological quality was appraised using MMAT 2018, and findings were synthesized thematically. The review included 35 reports representing 35 study units; 34 substantive units contributed to domain-frequency analysis. Metacognition appeared in 32 units (94.1%), learning strategies in 31 (91.2%), motivation in 17 (50.0%), and emotion in nine (26.5%). The most consistent cross-domain pattern linked metacognitive monitoring and evaluation with strategy selection and revision, while motivation and emotion shaped engagement and adaptation. These findings support a conditional cyclical account of SRL in MPS and suggest that strategy instruction should be integrated with monitoring, motivational support, and affect regulation. Uneven domain coverage, methodological heterogeneity, and single-reviewer procedures limit certainty. Future process-sensitive, multi-method, and experimental research should especially clarify motivational and emotional mechanisms.