Not when, but how: optimizing hospitality case learning beyond AI-Intervention timing

A pedagogical challenge is whether AI-supported hospitality case analysis still depends on students’ capacity. Guided by the 3P model, this study examines two pathways: one linking self-regulated learning to learning outcomes through problem structuring, and another linking epistemic beliefs to learning outcomes through evidence-based reasoning. Questionnaire data were collected from 205 hospitality management students in China who completed tasks at different levels of case uncertainty and at different AI-intervention timings. The findings show that both pathways matter, but the epistemic belief–evidence-based reasoning pathway exerts the stronger influence on learning outcomes. Case uncertainty further strengthens the contribution of problem structuring, whereas AI-intervention timing is not significant under the present design. The study contributes to hospitality education research by specifying dual-process mechanisms in AI-supported case learning and identifying case uncertainty as a meaningful instructional boundary condition. For practice, the results suggest that educators should prioritize evidence use, warranting, and disciplined problem framing.