Teacher-related factors and students' academic performance: a multilevel analysis.

BACKGROUND: This study examines the influence of teacher-related factors on students' academic performance in the post-COVID-19 context using a multilevel framework, drawing on Social Cognitive Theory and Self-Determination Theory. METHODS: Data were collected from 40 teachers and 418 students across 10 universities in Hainan, China. Confirmatory factor analysis (CFA) and common method variance (CMV) tests were conducted. Hierarchical linear modeling (HLM) was employed to examine the cross-level relationships among teacher-student relationships, teaching innovation, teacher professional identity, student self-efficacy, and academic performance. RESULTS: The results indicate that teacher-student relationships and teacher professional identity are significantly associated with students' academic performance, whereas teaching innovation shows no significant association. Furthermore, student self-efficacy positively predicts academic performance and mediates the relationship between teacher-student relationships and academic performance. CONCLUSION: These findings highlight the critical role of relational and psychological mechanisms in post-pandemic educational recovery.