Can Learners’ Use of GenAI Enhance Learning Engagement?—A Meta-Analysis

The integration of generative artificial intelligence (GenAI) into educational contexts has sparked significant interest in its potential to enhance learning engagement. However, empirical findings remain inconsistent, and a systematic synthesis of its effects across distinct engagement dimensions is lacking. This meta-analysis synthesizes evidence from 31 empirical studies (91 effect sizes), with the core aim of investigating the relationship between learners’ GenAI use and learning engagement, alongside the role of key moderating variables. Results indicate that GenAI exerts a significant positive effect on overall learning engagement, demonstrating the strongest impact on cognitive engagement, followed by affective and behavioral engagement. Moderator analyses reveal nuanced, sub-dimension-specific effects: the positive influence is most pronounced in higher education, shows significant benefits across all three sub-dimensions in basic education, and is non-significant in continuing education; medium-duration interventions (1 day–1 month) yield the largest effects; and teacher intervention significantly amplifies gains in cognitive engagement. Both learning mode and interaction approach exert significant positive effects on overall learning engagement, while their impacts on the sub-dimensions did not show significant heterogeneity. This study enriches the theoretical system of educational technology integration by clarifying the directional effect and moderating mechanisms of GenAI on learning engagement and provides a nuanced evidence base for designing context-sensitive implementations, offering valuable insights for fostering personalized and engaging learning experiences.