Course reassignment and learning engagement: the intervention strategy of AI-attributed reassignment

Course reassignment is a common practice in university registration systems, yet reassigned students often exhibit lower learning engagement than those in self-selected courses, and the underlying psychological mechanism remains unclear. This study examined the impact of course reassignment on learning engagement, the mediating role of self-efficacy, and the buffering effect of AI-attributed course reassignment.

Two scenario-based online experiments were conducted with university students randomly assigned to conditions. Experiment 1 (N = 393) used a single-factor between-subjects design comparing reassignment vs. self-selection. Experiment 2 (N = 293) added an AI-attributed reassignment condition in which an AI engine matched students' interests, traits, and abilities with the most suitable available courses (participants in the AI condition read a simulated AI-generated learning analysis report.). Participants completed a learning task and reported learning engagement and self-efficacy. Mediation analyses were conducted using bootstrapping.

Experiment 1 showed that course reassignment was associated with lower learning engagement compared with self-selection, and this effect was mediated by self-efficacy (indirect effect B = 0.198, SE = 0.051, 95%CI [0.096, 0.299]). Experiment 2 demonstrated that AI-attributed reassignment attenuated this negative effect. Learning engagement in the AI-attributed condition did not differ from self-selection and was significantly higher than in random reassignment. Moreover, the indirect effect through self-efficacy became non-significant (B = 0.140, SE = 0.113, 95% CI [-0.077, 0.362]).

These findings identify self-efficacy as a key psychological mechanism through which course reassignment impairs learning engagement, and demonstrate that AI-attributed reassignment can restore both self-efficacy and engagement by signaling person-course fit. This study extends self-efficacy theory to the course selection context and offers a feasible, technology-driven intervention for mitigating the negative consequences of course reassignment in higher education.