Technological compensation or pedagogical awakening? A longitudinal RCT of AI and expert feedback in pre-service physical education

A 16-week longitudinal cluster-randomized trial (N = 356) evaluated the comparative impacts of artificial intelligence feedback, human expert feedback, and self-guided reflection on pre-service physical education teachers. Findings establish an operational trade-off between efficiency and depth: AI feedback excelled at boosting immediate student moderate-to-vigorous physical activity and alleviating novice survival anxiety through low-latency spatial monitoring, whereas human expert feedback exhibited distinct, sustained developmental contributions.

This study demonstrates that artificial intelligence and human expertise serve fundamentally distinct, complementary functions in professional development and practicum training. Clarifying that automated systems excel at rapid metric optimization and anxiety mitigation enables institutions to deploy hybrid training models rather than pursuing complete automation. Incorporating AI into supervisory workflows protects expert capacity for nuanced, sustained pedagogical mentorship.

A 16-week cluster-randomized controlled trial with 356 pre-service teachers identified an efficiency-versus-depth trade-off between AI feedback and human expert feedback.

Forward consideration, not a verified fact.