Exploring AI-Assisted Self-Regulated Learning Profiles and the Predictive Role of Academic Appraisals: A Control-Value Perspective on Chinese EFL University Students

Abstract

As artificial intelligence technology becomes more integrated with foreign language education, understanding how learners regulate their engagement with these technologies is critical. Grounded in Control-Value Theory, this study investigates Chinese university students' AI-assisted self-regulated learning practice in the context of English as a foreign language (EFL) acquisition. Latent Profile Analysis was conducted on a dataset of 551 Chinese university EFL students to identify distinct self-regulated learning profiles based on six dimensions: goal setting, environment structuring, task strategies, time management, help seeking and self-evaluation. Three learner profiles emerged: Disengaged Learners, Partially Engaged Learners and Proactive Self-Directed Learners. Subsequent multinomial logistic regression revealed that academic appraisals (i.e., academic control and value) significantly predicted profile membership, with higher levels of both appraisals associated with a greater likelihood of being in the Proactive group. The findings highlight the heterogeneity of learners' AI use and the pivotal role of motivation in shaping effective self-regulation. The study extends the application of Control-Value Theory to AI-enhanced learning contexts and underscores the need to foster learners' sense of agency and task value.