Integrating Technology and Culture through AI-Driven Personalization: A Mixed-Methods Study of Learner Engagement in Malaysian Chinese Language Education

Abstract

Artificial intelligence (AI) has increasingly enabled personalized teaching in language education, yet its effectiveness in multicultural contexts remains underexplored. This study investigates how AI-driven personalization influences learner engagement in Malaysian Chinese language education. Using an explanatory sequential mixed-methods design, the quantitative phase (N = 368) tested relationships among perceived AI personalization, self-efficacy, cognitive load, and engagement. Structural equation modeling showed that AI-driven personalization positively predicted learner engagement, both directly and indirectly through enhanced self-efficacy, while excessive cognitive load weakened this relationship. Cultural congruence strengthened the positive effects of personalization.Follow-up interviews and classroom observations revealed that culturally responsive task design and teacher mediation enhanced learners’ autonomy and trust in AI systems, whereas culturally misaligned feedback reduced motivation.Overall, the study highlights AI-driven personalization as a culturally situated pedagogical process with implications for AI design and teacher development in multilingual educational settings.