AI-supported collaborative learning to reduce EFL college students’ anxiety in story writing

Abstract

In traditional collaborative learning, English as a Foreign Language (EFL) students with similar levels of foreign language writing anxiety (FLWA) frequently showed limited knowledge organization and teamwork effectiveness. This study proposed AI-supported collaborative learning to reduce EFL college students’ FLWA in story writing. The 36 students participated, with one class receiving AI-supported collaborative learning (experimental group) and the other receiving traditional collaborative learning (control group). Data were from pre- and post-FLWA scales and story writing tests. Results revealed that the experimental group reported significantly lower FLWA and higher story writing proficiency than the control group. Different from traditional collaborative learning, AI-supported collaborative learning provided automated feedback (story chain, word choices, grammar checks, and evaluation) that helped the students overcome challenges (e.g., idea generation and lexical usage). This AI support, with positive feedback among peers, initiated positive emotional responses. Under the broaden-and-build theory, these positive emotions broadened their thoughts, built knowledge, and ultimately reduced FLWA and improved English story writing.