Speaker
Description
This study examines how first-year Japanese university EFL students used generative AI during academic writing tasks. Drawing on prompt logs collected over six weeks, it identifies a functional taxonomy of AI prompt types ranging from command-based prompting to inquiry-based prompting. The findings suggest that command-based use was most common, while inquiry-based prompting aligned more closely with deeper revision and clearer understanding of appropriate AI use.
Abstract section 3: Content/method
This six-week classroom-based mixed-methods study involved 40 first-year Japanese university EFL students across two intact writing classes. Students completed a two-draft writing assignment and timed writing tasks without AI in Weeks 1 and 6. During AI-permitted stages, they submitted prompt logs documenting their ChatGPT use. Prompts were independently coded into seven functional categories.
Abstract section 4: Outcomes/results
Student prompts clustered primarily at the command-based end of the continuum, with grammar correction, translation, and vocabulary requests accounting for most interactions. Inquiry-based prompts, including requests for explanation, evaluation, and organizational feedback, occurred less frequently, but were associated with deeper revisions and clearer student understanding of appropriate AI use. Prompts requesting direct text generation were associated with greater ethical uncertainty. Overall, the findings suggest that prompting behavior shaped the kinds of writing support students received from AI.
Abstract section 1: Relevance
Recent classroom-based studies suggest that structured ChatGPT use can support some aspects of writing development in Japanese university EFL contexts, particularly mechanics and organization, while showing more limited effects on grammar, vocabulary, and content development (Hennessy, 2025; Massoud et al., 2026). At the same time, work in L2 writing suggests that the value of generative AI depends not only on access to the tool itself, but also on how learners prompt it and evaluate its output (Li, 2025; Kim et al., 2025). Prompting behavior remains under-examined in Japanese university EFL writing classrooms.
Abstract section 2: Contribution/research questions
This study addressed three questions: What functional types of AI prompts did first-year Japanese university EFL students use during academic writing tasks? Could these prompt types be reliably classified using a functional taxonomy? How did different prompt types relate to revision depth, surface accuracy, and students’ perceived understanding of appropriate AI use?
Abstract section 5: References
Hennessy, E. (2025). Enhancing L2 writing with ChatGPT for Japanese EFL students. In B. Lacy, M. Swanson, & P. Lege (Eds.), Moving JALT into the future: Opportunity, diversity, and excellence. JALT.
Kim, J., Yu, S., Lee, S., & Detrick, R. (2025). Students’ prompt patterns and its effects in AI-assisted academic writing: Focusing on students’ level of AI literacy. Journal of Research on Technology in Education. Advance online publication. https://doi.org/10.1080/15391523.2025.2456043
Li, S. (2025). Generative AI and second language writing. Digital Studies in Language and Literature, 2(1), 1–31. https://doi.org/10.1515/dsll-2025-0007
Massoud, O. S., Zhang, J., & McQuilling, P. (in press). Examining the effects of structured ChatGPT use on first-year ESL writing development over two semesters. Sophia University Junior College Division Faculty Journal.
| Title | From Commands to Inquiry: AI Prompting in Japanese EFL |
|---|---|
| Teaching Context | College and university education |