Uncovering Students’ Processing Tactics Towards ChatGPT’s Feedback in EFL Education Using Learning Analytics

Hanyu Su,Yixin Tong,Xinyi Zhang,Yizhou Fan

DOI: https://doi.org/10.1007/978-981-97-4442-8_18

2024-01-01

Abstract:ChatGPT, a promising chatbot program based on generative artificial intelligence (GAI), shows significant promise for its potential use in English as a Foreign Language (EFL) education. By engaging in dialogue with students, it provides personalized support for their reading and writing, which are fundamental aspects of EFL instruction. Although research on students’ interaction with ChatGPT is emerging, there is a lack of focus on their Processing Tactic towards ChatGPT’s Feedback (PTCF) and its impact on learning outcomes. To mitigate this gap, we employed a learning analytic method combining cluster analysis and process mining to analyze multimodal data generated during their reading and writing tasks. The analysis identified three clusters of students’ PTCF: (1) Silence with no response, (2) Passive dependence, and (3) Active construction. By conducting sequence analysis on these behavior patterns, it was found that learners adopted three learning modes when using ChatGPT, which are Comprehension-focused, Tool-inspired, and Thinking-alienator. The findings also revealed differences in learning gains among groups of students with varying learning modes, specifically in the improvement of domain knowledge. Implications of these findings for both practical application and future research were also discussed later.