Speaker
Description
As generative artificial intelligence (GenAI) tools have gained increasing attention in the field of L2 writing instruction, understanding how learners with different English proficiency levels engage with AI-generated feedback is essential for developing AI literacy and learner agency. This qualitative case study explores the behavioral, cognitive, and affective engagement of six Vietnamese EFL students, representing different English proficiency levels who participated in four writing tasks facilitated by ChatGPT. The data for analysis consisted of successive draft essays, screen recordings of the revision process and stimulated recall interviews, and were analyzed based on within-case and cross-case analyses. The results showed differences in the engagement of the learners according to their levels of proficiency. The high proficiency level students engaged critically with the ChatGPT suggestions as they evaluated and adapted the suggestions to their revisions. It was found that the intermediate level students demonstrated moderate engagement, while the low-level students relied more on the feedback provided by the AI and accepted the suggestions provided by it without criticizing it. This shows that English proficiency not only impacts how students make use of GenAI feedback but also their critical engagement with it. The results of this study highlight the need for fostering critical engagement of the learners with AI-generated feedback in addition to developing their literacy skills.
Keywords: generative artificial intelligence, ChatGPT, engagement, English proficiency, Vietnamese EFL students
