THE USE OF AI-GENERATED FEEDBACK IN ACADEMIC WRITING: CTU ENGLISH MAJOR STUDENTS’ PERCEPTION

Abstract

This mixed-methods research investigated the perceptions of 100 English-major students at Can Tho University (CTU) concerning the integration of AI feedback tools like ChatGPT and Grammarly into their academic writing practices. Employing surveys and open-ended responses, the study evaluated the perceived advantages, disadvantages, and practicality of these tools through the lens of the Technology Acceptance Model (TAM) and Self-Efficacy Theory. Quantitative findings revealed moderately positive student views on AI's Ease of Use and Perceived Usefulness, in which students particularly valued the real-time grammatical corrections, structural suggestions, and support for idea generation offered by these tools. The immediacy of AI feedback was also considered a significant benefit. However, qualitative data uncovered notable concerns. In addition to this, a substantial majority (66%) of respondents expressed apprehension about potential over-reliance on AI, which they feared could impede the development of their critical thinking and creativity. Additional challenges identified included the limited contextual understanding of AI, inconsistencies in its technical performance across different platforms, and ethical considerations related to data privacy and algorithmic bias. Furthermore, students highlighted the tools' inability to provide nuanced, discipline-specific feedback, especially in tasks demanding rhetorical sophistication or cultural awareness. The study concludes that while AI-generated feedback serves as a valuable supplementary resource for academic writing, its effective implementation necessitates careful integration alongside human instruction to ensure academic rigor and foster critical engagement. The research recommends the adoption of hybrid pedagogical models that combine the efficiency of AI with the mentorship of educators. It also suggests the development of institutional policies addressing the ethical use of AI and the implementation of student training programs to enhance digital literacy skills. Acknowledging limitations such as the study's focus on a single institution and its reliance on self-reported data, the research underscores the need for future investigations in more diverse contexts. This study provides initial, valuable insights into the adoption of AI feedback within the Vietnamese higher education landscape, offering important considerations for educators, policymakers, and students navigating the evolving intersection of technology and pedagogy.