conceptual understanding

Guided ChatGPT inquiry accompanied stronger thermodynamics scores, with teacher support built into the intervention

Tarik El Fathi, Aouatif Saad, Hayat Larhzil, Driss Lamri, El Mehdi Al Ibrahmi

Disciplinary and Interdisciplinary Science Education Research

A quasi-experiment with 120 first-year engineering students compared conventional thermodynamics tutorials with guided ChatGPT inquiry. The AI-supported group achieved higher conceptual post-test scores and reported positive experiences, but the intervention also included prompt training, peer discussion and instructor feedback. Difficulties remained on quantitative topics. The study supports evaluating a complete instructional design; it does not isolate the chatbot's contribution or establish durable gains beyond one semester.