AI-MEDIATED FEEDBACK TOOLS FOR WRITING IN NON-NATIVE ENGLISH CONTEXTS

Abstract

With the increasing popularity of AI-driven tools in learning institutions, automated writing feedback as a field of application has become particularly topical in non-native English speakers (NNES) who want to address the requirements of academic and professional communication quality. This research will explore the perspectives and application of AI-mediated writing feedback systems, including Grammarly and Quillbot, by NNES university students to enhance their writing fluency, accuracy, and self-revision of writing. The study seeks to find the pedagogical and perceived limitation of these tools and their behavioral effects on the English as a Second Language (ESL) or English as a Foreign Language (EFL) setting. Mixed-methods case study was developed with 36 undergraduate and graduate NNES students who were taking English academic writing courses in two universities. The pre- and post-intervention writing samples, usage analytics, surveys, and in-depth interviews were used to gather the data. Analytical instruments were frequency of errors, revision, and thematic coding of the qualitative data. The research was based on the assumption that AIs will be used well alongside human instructions, students will differ in their trust and reliance to machine-based suggestions. The findings showed that AI resources had a vastly positive impact on mechanical and grammatical errors reduction and cohesion and lexical diversity, especially among lower-intermediate learners. But there was over-dependence and formulaic sentence formatting and some students complained that they could not make out subtle suggestions. AI-assisted writing sessions showed more writing autonomy and more frequent iteration as compared to traditional teacher feedback alone, but they needed meticulous supervision as they are likely to be misused. The paper concludes that feedback devices mediated by AI can provide valuable support to NNES writers, though their effects are best achieved through the combination of the explicit teaching of writing skills and training access to digital literacy forms of critique.