ChatGPT-Supported Teacher Feedback on L2 Students’ Writing: EFL Teachers’ Perspectives
摘要
This study explored an innovative approach to providing written feedback, namely, ChatGPT-supported teacher feedback, in the Chinese tertiary EFL context. Two research questions guided this study: (1) What are EFL instructors’ perceptions of ChatGPT-supported teacher feedback? (2) How do EFL instructors revise ChatGPT output while providing ChatGPT-supported teacher feedback? Specific training was conducted with four course instructors and two classes of students, and the instructors provided ChatGPT-supported teacher feedback on two writing tasks across seven weeks. Perception data were collected from four instructors via individual semi-structured interviews, supplemented with records of ChatGPT feedback, teachers’ responses to ChatGPT output, and students’ uptake of ChatGPT-supported teacher feedback. Qualitative content analysis was conducted, and the constant comparative method was utilized. The findings revealed the instructors’ perceived benefits of ChatGPT-supported teacher feedback and the perceived constraints of ChatGPT output. In implementing ChatGPT-supported teacher feedback, the instructors added new comments, extended the feedback, modified inappropriate suggestions, condensed repetitive comments, and deleted excessive praise. This study provides novel pedagogical insights into the integration of ChatGPT’s capabilities with teachers’ agency when providing written feedback.