Artificial Intelligence‐Supported Adaptive Flipped Learning and Changes in Autonomy, Competence, Relatedness and Novelty in Higher Education: A Quasi‐Experimental Study
Alberto Ferriz‐Valero, Ricardo Sánchez‐Gil‐Machín, Salvador García‐Martínez, Nuria Molina‐García, Alberto Real‐Fernández, Salvador Baena‐MoralesABSTRACT
This quasi‐experimental study examined whether AI‐supported Adaptive Flipped Learning (AFL) was associated with psychological need and novelty‐related experiences in 311 university students (AFL: 164; comparison: 147) across 16 classes. The Psychological Need Satisfaction Scale and Psychological Need Thwarting Scale assessed autonomy, competence, and relatedness satisfaction and frustration, respectively; the Novelty Need Satisfaction and Frustration Scale assessed both novelty dimensions. Pre–post change models adjusted for instructor‐year blocks, with wild‐cluster bootstrap inference and false‐discovery‐rate correction across eight outcomes. AFL showed more favourable changes in autonomy, competence, and relatedness satisfaction and lower novelty frustration. Novelty satisfaction was supported after correction only in baseline‐adjusted sensitivity analysis (primary q = 0.076). These findings provide qualified evidence for the integrated AFL model; non‐random allocation limits causal inference, and suboptimal factor fit warrants caution about the psychological need satisfaction scores and novelty.