Shin, D., & Lee, J. H. (2025). Leveraging LLM-based chatbots for interactional grammar feedback in L2 writing: Opportunities and challenges. Innovation in Language Learning and Teaching, 1–13. (100번째 Journal Paper) https://doi.org/10.1080/17501229.2025.2549037 등록전
Published online: 28 Aug 2025
https://www.tandfonline.com/eprint/SIJZFKP32RBPPCQT8KCK/full?target=10.1080/17501229.2025.2549037
As ChatGPT is used for many purposes in L2 classrooms, recent studies have focused on employing it to provide grammar feedback on students’ writing. However, the feedback explored has tended to be unidirectional and thus limited in terms of engaging L2 learners in revision or eliciting error detection and correction. In light of the principle of dynamic assessment in a Vygotskian framework, this study showcases an innovative practice based on a chatbot capable of providing interactional writing feedback for L2 writing instruction. It is designed to allow learners to identify their errors and provide progressively more explicit feedback. A total of 89 second-year pre-service teachers at a university of education in South Korea participated in this study, in which they attended a chatbot-building workshop, built a grammar feedback chatbot, and received feedback from it on their English essays. Our findings show that the target chatbot offers several advantages as a feedback provider in L2 writing classrooms, and that it could be a complementary resource for human teachers rather than a replacement.
KEYWORDS: Chatbot ChatGPT grammar feedback L2 writing LLM