#97: Higher-Order Interactions in Learning: Extending Cognitive Load and Self-Regulated Theories through a Network Approach

While learning network research focuses on pairwise relationships, we applied higher-order network analysis to three-body interactions among 39 variables (PISA 2015, n=2,189). Our approach, combining network estimation with hierarchical screening and Stability Selection, revealed emergent phenomena: interaction thresholds where two-body effects saturate, competition among multiple regulation processes, and context-dependent motivational reversals. These discoveries extend Cognitive Load and Self-Regulated Learning theories, demonstrating how higher-order analysis can refine learning theories.