Students’ perceptions of an eight-step writing model incorporating machine translation and writing support tools: a narrative case study in a Japanese EFL context

Abstract

Despite the increasing availability of AI-assisted writing tools, the pedagogical frameworks that guide the ethical use of these tools in L2 writing instruction remain underdeveloped. This qualitative narrative case study examines the perceptions of Japanese university students regarding English as a Foreign Language (EFL) writing within a structured eight-step writing model in which AI-assisted writing tools, specifically machine translation and writing support tools, are systematically integrated. Nine students participated in a semester-long writing course in which they progressed through the stages of independent drafting, AI-assisted revision, back-translation for meaning verification, peer review, oral presentation, and reflection. Semi-structured interviews revealed that learners’ perceptions of these tools shifted over time. Initial resistance rooted in ethical concerns gradually shifted toward an understanding of AI-assisted writing tools as legitimate mediational resources. The structured procedure supported learner agency and authorship by requiring students to articulate ideas independently before selectively engaging with AI-assisted writing tools. Back-translation helped learners evaluate whether the AI-generated output aligned with their intended meaning. Peer review and presentation fostered social accountability and audience awareness. Overall, the findings suggest that embedding AI-assisted writing tools within a process-oriented instructional sequence can mediate learners’ behavioral, affective, cognitive, and social engagement. This structured approach may help lower-proficiency EFL writers to maintain ownership of meaning while producing more communicatively effective texts.