Can teacher feedback be substituted by Gen-AI? A comparative study of writing scores, feedback characteristics, revision and motivation

Feedback is essential in English as a foreign language (EFL) learners’ writing development. However, providing effective feedback remains challenging due to teachers’ time and resource constraints. With the advancement of generative artificial intelligence (Gen-AI), integrating it into writing teaching and learning could be a potential solution. Generative AI-powered writing evaluation can timely deliver individualized, dynamic and convenient feedback to students, addressing the problems that teachers face and supporting students’ learning. Despite this promise, there is limited research exploring the nature of its feedback and whether Gen-AI can match the evaluative quality of teachers. To fill this gap, this study compares the feedback from Gen-AI (GAF) with teacher feedback to explore their respective effects and characteristics. Conducted with 105 EFL students at a Chinese university, the findings reveal that Gen-AI has generated a significantly larger volume of feedback and offers more comprehensive coverage of both local and global issues than teachers. While GAF demonstrates a positive impact on EFL writing, it does not surpass the overall effectiveness of teacher feedback in fostering writing revision and enhancing students’ motivation. The results underscore the unique potential of Gen-AI in complementing teacher feedback, advocating for an integrated feedback approach and shedding light on the further development of the feedback mechanism of Gen-AI.