Self-regulated learning intervention in AI-assisted ESP vocabulary learning: Effects on vocabulary achievement and enjoyment

With the rapid advancement of artificial intelligence (AI), intelligent learning tools have become increasingly integrated into English for Specific Purposes (ESP) vocabulary instruction. However, the effectiveness of these tools often depends on the pedagogical framework in which they are embedded. This quasi-experimental study examined the role of a structured self-regulated learning (SRL) intervention in engineering English vocabulary achievement and vocabulary learning enjoyment among Industrial Engineering (IE) undergraduates at a Chinese university. Participants were 266 IE students in pre-existing classes who took part in a seven-week study and were assigned at the class level to an experimental group receiving the SRL intervention or a control group using AI autonomously. The intervention used a structured handbook to scaffold goal setting before AI interaction, real-time self-monitoring during tasks, and genre-reflective evaluation afterward, whereas the control group used AI autonomously without self-regulatory guidance. Results indicated that the experimental group significantly outperformed the control group in post-test vocabulary achievement. The experimental group also demonstrated a significant and meaningful increase in vocabulary learning enjoyment, while the control group showed no significant change. Moderation analyses showed that the between-group difference was larger among students with lower baseline enjoyment, whereas it was not statistically significant among those with higher baseline enjoyment. These findings suggest that structured pedagogical support may contribute to cognitive and affective gains beyond AI access alone. Practically, the findings highlight the value of embedding AI-assisted vocabulary tasks within explicit SRL routines rather than providing AI access as a standalone resource.