Artificial intelligence in education: effects on motor learning, motivation, and student engagement from an educational psychology perspective

Abstract

The rapid expansion of artificial intelligence (AI) in education has renewed debates about its pedagogical value, particularly in relation to learner-centered approaches and motivational processes. While a growing body of research has examined AI-supported learning in cognitively oriented subjects, empirical evidence remains scarce in practice-based disciplines such as physical education (PE), especially within secondary school contexts in developing countries. This study investigated the effects of integrating AI-supported instructional tools into PE lessons on students' motor learning, intrinsic motivation, and engagement in Tunisian secondary schools. A quasi-experimental pre-test/post-test design involving two intact classes was employed because individual random assignment was not feasible in the school setting (N = 56; age range 15–17 years, M = 16.4, SD = 0.9), including an experimental group receiving AI-supported instruction and a control group following traditional teaching methods over an 8-week intervention period. Motor learning was assessed using the TGMD-3, intrinsic motivation through the Intrinsic Motivation Inventory, and engagement via a multidimensional student engagement scale. Results revealed statistically significant and educationally meaningful improvements in motor skill acquisition, intrinsic motivation, and engagement among students exposed to AI-supported instruction compared with their peers in the control group. Effect sizes indicated strong practical relevance, suggesting that AI-supported visual feedback based on predefined movement indicators was interpreted and contextualized by the teacher to support motor skill acquisition. These findings suggest that AI-supported instruction may enhance motor learning, intrinsic motivation, and student engagement when integrated into teacher-mediated physical education. Given the quasi-experimental design involving intact classes, these findings should be interpreted as evidence of association rather than definitive causal effects. These findings contribute context-sensitive empirical evidence to the field of educational psychology by highlighting the potential of human-centered AI integration in secondary school physical education within a developing country context. The study underscores the importance of pedagogical mediation, learner autonomy, and ethical awareness in the design and implementation of AI-supported learning environments.