This study reports on the practice of message exchanging activities with Learning Management System (LMS) in the Global Classmate Project. A total of 33 Japanese high school students interacted online with 23 high school students from New York over the eight sessions. The results suggested that the students’ confidence in their English writing increased between pre- and post-exchanges in terms of fluency, accuracy, and expression, although confidence in accuracy decreased between mid- and post-exchanges. Furthermore, the number of words in their posts gradually increased as the session continued. A qualitative analysis, Modified Grounded Theory Approach (M-GTA) of the students’ reflections also revealed that they became less anxious about writing in English, more familiar with writing, more motivated to express themselves, and enjoyed communicating online more.

抄録全体を表示

　今回の部分撤回は、研究倫理委員会による審議で指摘された問題に対処するために行われたものです。
　対象となる箇所については、出版後訂正の国際的な慣行に則って、透明性の確保および学術記録の誠実さを維持する目的で、以下の著者による部分撤回記事において明確に示されています。なお、本論文のその他の部分には影響を及ぼさないことを申し添えます。

『LET九州・沖縄支部紀要』第26号　編集委員会
2026年2月

抄録全体を表示

Since the emergence of ChatGPT in 2022 the rapid development of generative AI has brought significant opportunities and challenges to English writing instruction. While Japan’s Ministry of Education promotes the use of generative AI in schools, and has begun training AI-leader teachers, many educators remain uncertain about how to integrate AI meaningfully into their instruction. This paper examines the pedagogical implications of AI-assisted writing from the perspectives of high school-university collaborative practices. First, it identifies major risks associated with overreliance on AI—including threats to academic integrity, loss of learning opportunities, and students’ inability to critically evaluate AI-generated text. This paper highlights the importance of explicit guidelines, transparent purposes for AI use, and instruction that develops students’ ability to revise, evaluate, and personalize AI outputs. Second, it argues that although AI can support surface-level accuracy and promote autonomous learning, human teachers remain essential for providing content-focused feedback, nurturing students’ thinking, and fostering motivation. Finally, this paper proposes a staged approach to AI integration that prioritizes foundational writing experience before full AI-supported writing. The discussion offers practical considerations for designing writing instruction that makes effective use of AI while safeguarding students’ development as independent writers.

抄録全体を表示