OpenAlex · Aktualisierung stündlich · Letzte Aktualisierung: 28.03.2026, 12:21

Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.

Enhancing Japanese EFL Students’ Grammar Accuracy and Writing Fluency Using ChatGPT: A Mixed-Method Study

2025·0 Zitationen·Journal of International Education and PracticeOpen Access
Volltext beim Verlag öffnen

0

Zitationen

3

Autoren

2025

Jahr

Abstract

This mixed-methods study investigated whether or not ChatGPT worked as a writing aid to increase the grammar accuracy and writing fluency of Japanese EFL students. Thirty Japanese senior high school students took part in a six-week intervention that included ChatGPT in process-based writing assignments. Three research questions were examined in the study: (1) How much does ChatGPT help learners become more accurate with grammar? (2) What impact does ChatGPT have on writing fluency? (3) What role does ChatGPT play in the development of students' writing? Pre- and post-tests evaluated using analytical rubrics yielded quantitative data. Grammar accuracy (t(29) = 8.62, p < 0.001, d = 1.11) and writing fluency (t(29) = 6.46, p < 0.001, d = 0.84) both significantly improved, according to paired-samples t-tests. Three main themes emerged from qualitative data from learner journals, semi-structured interviews, and classroom observations: ChatGPT as a grammar mentor, fluency development via iterative rewriting, and growing concerns about overreliance. Students reported better coherence and flow in written texts, increased awareness of frequent grammatical errors, and increased confidence in editing sentences. Results show that when ChatGPT is used as a scaffolded, process-oriented tool rather than as a replacement for independent writing, it can support both linguistic accuracy and fluency. Students' worries about dependence, however, emphasize the necessity of clear instruction in digital literacy. The study offers pedagogical implications for integrating AI responsibly in Japanese and wider Asian EFL contexts and adds empirical evidence to the expanding body of research on AI-assisted EFL writing.

Ähnliche Arbeiten