How teacher-GenAI integrated feedback affects undergraduate academic writing: A quasi-experimental study

Integrating teacher feedback with GenAI feedback has attracted growing interest, yet how different teacher feedback types interact with GenAI feedback to shape undergraduate students’ academic writing remains underexplored. This study adopted a quasi-experimental design to explore the impact of three feedback types on undergraduate students’ academic writing. The results indicated that the combination of teacher cognitive feedback and GenAI feedback most effectively enhanced academic writing performance; teacher affective feedback paired with GenAI promoted more extra revisions, whereas GenAI feedback alone often led to non-implementation. Semi-structured interviews revealed markedly distinct perceptions across three dimensions: the gap between feedback recognition and implementation, emotional responses to feedback, and developmental trajectories of academic writing strategies. This study advocates for a human-centred approach to human-AI feedback integration, offering insights into leveraging their complementary strengths for meaningful learning.