Impact of AI-Assisted Peer Feedback on EFL Students’ Self-Regulated Learning, Self-Efficacy, and Motivation in English Academic Writing

Abstract

Although the demand for English academic writing talents is growing, the teaching of English academic writing in China faces challenges such as students’ lack of self-regulated learning, self-efficacy, and learning motivation. Therefore, this study aims to explore the impact of AI-assisted peer feedback with DeepSeek on Chinese EFL undergraduates’ self-regulated learning, self-efficacy, and learning motivation. The researchers conducted a 16-week experiment on 60 third undergraduates majoring in English at a university in northern China, who were divided into an experimental group (EG, N=30) receiving AI-assisted peer feedback with DeepSeek, and a control group (CG, N=30) receiving teacher feedback. Mixed-design ANOVA results showed that both EG and CG students showed an upward trend in self-regulated learning, self-efficacy, and learning motivation, but EG students made greater progress than CG students. In addition, paired sample t-test results showed that EG students had significant self-regulated learning, self-efficacy, and learning motivation before, during, and after the questionnaire, and the effect size was medium to high. However, a small number of CG students were significant and the overall effect size was small. The results of the semi-structured interviews further verified the quantitative research findings that AI-assisted peer feedback was considered to be more helpful than traditional teacher feedback in improving different levels students’ self-regulated learning, self-efficacy, and learning motivation. Implications and recommendations were proposed of this study.