AI-based educational interventions for enhancing cognitive learning processes in students with disabilities: A meta-analysis

Guided by the Cognitive Learning Processes Adaptation Model (CLPAM), this meta-analysis provides a theory-driven synthesis of how AI-based educational interventions support core cognitive learning processes in students with disabilities. The study aims to clarify the cognitive mechanisms through which AI interventions influence learning by focusing on attention regulation, cognitive load management, and memory storage. Drawing on 42 effect sizes from 20 studies published between 2010 and 2025, random-effects models revealed a significant overall effect of AI-based interventions on cognitive learning processes (g = 0.726). Domain-specific analyses demonstrated robust effects for attention regulation (g = 0.817), memory storage (g = 0.783), and cognitive load management (g = 0.691). Moderator analyses indicated that robotics-based interventions, structured teaching approaches, and moderate levels of AI interactivity were associated with larger cognitive gains. By organizing evidence around theoretically defined cognitive processes, this study advances understanding of AI-supported learning mechanisms and informs theory-driven inclusive educational practice.