Asosiy kontentga oʻtish
Maqola

CHATGPT-GENERATED FEEDBACK IN EFL WRITING: EFFECTIVENESS, LIMITATIONS, AND LEARNER REVISION

E’zoza IslomovaSecond-Year Undergraduate Student Faculty of Foreign Languages and Literature (English) Uzbekistan State World Languages University, Tashkent, Uzbekistan Narrative Review
2026en
ABI

Annotatsiya

Generative artificial intelligence has expanded access to immediate and interactive feedback on second- and foreign-language writing. ChatGPT can comment on grammar, vocabulary, organisation, and content; however, receiving feedback does not necessarily result in accurate revision or lasting learning. This narrative review examines research on ChatGPT-generated feedback in English as a Foreign Language (EFL) and English as a Second Language (ESL) writing, focusing on feedback functions, learner engagement, feedback uptake, revision quality, and the distinction between immediate text improvement and independent writing development. The discussion connects foundational feedback frameworks with empirical studies and recent reviews. The literature suggests that ChatGPT can provide timely explanations and support repeated revision, and that guided use may improve selected aspects of writing. At the same time, AI-generated suggestions can be inaccurate, generic, or insufficiently sensitive to context, and learners may accept them without adequate evaluation. Research on learner engagement further indicates that cognitive processing, observable revision behaviour, and affective responses all shape the use of feedback. Evidence from Uzbekistan is also beginning to emerge, although locally situated studies have examined broader AI-assisted writing practices rather than ChatGPT feedback alone. The review argues that ChatGPT is best treated as a supplementary resource within teacher-guided instruction. It recommends more transparent research designs, closer analysis of revision decisions, and assessments of transfer to new writing tasks.

Hali tarjima qilinmagan

Iqtiboslar va manbalar