Assessing strategic use of artificial intelligence in self-regulated learning: Instrument development and evidence from Chinese university students

Despite the rapid integration of AI into students’ self-regulated learning, rigorously validated instruments to assess their strategic AI use are critically lacking. Addressing this gap, the present study proposes a conceptual framework of self-regulated use of AI for learning, based on which a strategy inventory was developed and validated in two questionnaire surveys among Chinese university students (with valid n = 488 and 707 respectively). Factor and Rasch analyses of the data identified eight factors underlying the instrument, with four pertaining to the regulation of AI-assisted task processing, and another four associated with the regulation of AI application. The corresponding subscales also demonstrated robust psychometric properties. Quantitative analyses further revealed imbalanced strategy usage. Specifically, output-focused metacognitive regulation was reported to be most frequent, whereas AI-supported self-assessment strategies were underutilized. Results also showed that there were gender and academic disciplinary differences in ethical and tool-focused metacognitive regulation; while degree of AI exposure was positively correlated with most strategy types, it had no significant correlation with ethical regulation. These findings highlight the need for differentiated interventions to address populational gaps in the AI-mediated learning process. This study contributes to a structured approach to understanding AI’s role in self-regulated learning and calls for concerted measures to cultivate students’ metacognitive and ethical awareness in AI application. Pedagogically, the validated instrument can be used by teachers to diagnose students’ strategy use and inform targeted instruction. Students can also use it for self-assessment, enhancing their self-awareness and fostering a more balanced use of different AI strategies.