Chapter 3. Generative AI as a Writing Feedback Companion: Practice, Efficacy, and Student Beliefs
Marila Melnikova & Ashton E. Dawes
Abstract
Corrections that students seek themselves are more significant for language learning, and such feedback allows students to independently and critically analyze their writing. Students need exposure to diverse types of feedback and the power to request feedback whenever necessary. The rise of generative artificial intelligence (AI) presents a new avenue for students to seek feedback. However, the efficacy and quality of feedback depend heavily on students’ awareness of AI’s strengths and limitations and on how they use it. The study described in this chapter investigates the practical value of AI feedback on second-year English-language student writing at a private Japanese university. The researchers individually evaluated AI’s responses to students seeking feedback to see what suggestions for academic improvement it offered. Through interrater reliability, the researchers determined how well AI’s suggestions on grammar, content, structure, APA format, and others followed academic standards. In this way, AI’s strengths and weaknesses in academic writing were determined. Apart from the researcher’s understanding of AI’s reliability, surveys were given at the start and end of the semester to assess students' beliefs about AI’s classroom applicability and accuracy. Students were also asked about their attitudes towards different kinds of feedback. Students completed reflection journals documenting the specific feedback they sought and their subsequent writing choices as directed by AI. Findings highlight shifts in students' attitudes and knowledge, and offer practical uses of AI in writing education, underscoring AI's potential as a feedback companion in the L2 writing classroom. This research contributes insights into integrating AI as a supportive feedback tool for writing, particularly for assistance with grammar and vocabulary. It highlights the importance of varied types of feedback and where technology best fits into the writing process.
About the Contributors
Mariia Melnikova holds a degree in Linguistics from Lomonosov Moscow State University and is currently a lecturer at the English Language Institute at Kanda University of International Studies in Chiba, Japan. She teaches courses in academic writing, reading, and English communication. Her research focuses on English as a foreign language pedagogy, with particular interests in vocabulary acquisition strategies and feedback in academic writing. She has co-authored publications on AI-assisted writing feedback and vocabulary learning practices and has presented her work at international conferences on applied linguistics and language education. Originally from Russia, she has lived and worked in Japan since 2019.
Ashton E. Dawes is a lecturer in the Center for Foreign Language Education and Research, Rikkyo University, Japan. She has nearly a decade of experience in L2 writing instruction, including coordinating writing courses and research. She holds a master’s degree in Applied Linguistics and TESOL from the University of Mississippi, USA. Her research interests include educational design and emerging technologies, as well as peer feedback in writing pedagogy. Her current projects explore collaborative peer review and feedback literacy practices in Japanese higher education, and theorize podcasts as creative outlets for language learners.
Citation
Melnikova, M., & Dawes, A. E. (2026). Generative AI as a writing feedback companion. In R. Dykes, O. Edwards, D. Bollen, & T. S. W. Lin (Eds.), Artificial intelligence in Japan’s language learning classrooms (pp. 54–84). Candlin & Mynard. https://doi.org/10.47908/45/3
Corrections that students seek themselves are more significant for language learning, and such feedback allows students to independently and critically analyze their writing. Students need exposure to diverse types of feedback and the power to request feedback whenever necessary. The rise of generative artificial intelligence (AI) presents a new avenue for students to seek feedback. However, the efficacy and quality of feedback depend heavily on students’ awareness of AI’s strengths and limitations and on how they use it. The study described in this chapter investigates the practical value of AI feedback on second-year English-language student writing at a private Japanese university. The researchers individually evaluated AI’s responses to students seeking feedback to see what suggestions for academic improvement it offered. Through interrater reliability, the researchers determined how well AI’s suggestions on grammar, content, structure, APA format, and others followed academic standards. In this way, AI’s strengths and weaknesses in academic writing were determined. Apart from the researcher’s understanding of AI’s reliability, surveys were given at the start and end of the semester to assess students' beliefs about AI’s classroom applicability and accuracy. Students were also asked about their attitudes towards different kinds of feedback. Students completed reflection journals documenting the specific feedback they sought and their subsequent writing choices as directed by AI. Findings highlight shifts in students' attitudes and knowledge, and offer practical uses of AI in writing education, underscoring AI's potential as a feedback companion in the L2 writing classroom. This research contributes insights into integrating AI as a supportive feedback tool for writing, particularly for assistance with grammar and vocabulary. It highlights the importance of varied types of feedback and where technology best fits into the writing process.
About the Contributors
Mariia Melnikova holds a degree in Linguistics from Lomonosov Moscow State University and is currently a lecturer at the English Language Institute at Kanda University of International Studies in Chiba, Japan. She teaches courses in academic writing, reading, and English communication. Her research focuses on English as a foreign language pedagogy, with particular interests in vocabulary acquisition strategies and feedback in academic writing. She has co-authored publications on AI-assisted writing feedback and vocabulary learning practices and has presented her work at international conferences on applied linguistics and language education. Originally from Russia, she has lived and worked in Japan since 2019.
Ashton E. Dawes is a lecturer in the Center for Foreign Language Education and Research, Rikkyo University, Japan. She has nearly a decade of experience in L2 writing instruction, including coordinating writing courses and research. She holds a master’s degree in Applied Linguistics and TESOL from the University of Mississippi, USA. Her research interests include educational design and emerging technologies, as well as peer feedback in writing pedagogy. Her current projects explore collaborative peer review and feedback literacy practices in Japanese higher education, and theorize podcasts as creative outlets for language learners.
Citation
Melnikova, M., & Dawes, A. E. (2026). Generative AI as a writing feedback companion. In R. Dykes, O. Edwards, D. Bollen, & T. S. W. Lin (Eds.), Artificial intelligence in Japan’s language learning classrooms (pp. 54–84). Candlin & Mynard. https://doi.org/10.47908/45/3
Information About the Book
|
Title: Artificial Intelligence in Japan’s Language Learning Classroom
Editors: Robert Dykes, Oliver Edwards, Dave Bollen, and Tina Shu-wen Lin Publication date: June 2026 Read more... |