The Impact of AI-Assisted Learning Engagement on English Proficiency Gains: The Serial Mediating Role of Intrinsic Motivation and Academic Self-Efficacy in Chinese Higher Education Contexts

Abstract

This study examined how AI-assisted learning engagement influences English proficiency gains through the serial mediation of intrinsic motivation and academic self-efficacy among Chinese university students. A cross-sectional survey design was employed. A total of 456 undergraduate students were recruited from three comprehensive universities in eastern China. The hypothesized serial mediation model was tested using PROCESS model 6 program. The findings indicated that the influence of learning engagement with AI assistance on the enhancement of English proficiency was significant (c' = 0.18, p < 0.05), with a significant serial indirect effect through intrinsic motivation and academic self-efficacy (indirect effect = 0.10, 95% CI [0.039, 0.171]). The psychological mediators accounted for 55% of the total relationship, while the final model accounted for 52% of the variance in English proficiency outcomes. Overall, the results indicated that learning tools facilitated by AI contributed to educational effectiveness via autonomous motivation, which further enhanced learners' self-efficacy beliefs. This study makes an important contribution to the technology-mediated learning literature by demonstrating that the effectiveness of learning tools facilitated by AI depends on the learners' psychological processes rather than their technological features, which is particularly important for educational settings in higher education contexts.