Fostering Student Engagement and Learning Perception Through Socratic Dialogue with ChatGPT: A Case Study in Physics Education

This classroom-based case study examines how an AI-mediated Socratic dialogue, implemented
through ChatGPT, can support students’ engagement and perceived learning in
undergraduate thermodynamics. Conducted in a first-year engineering physics course at a
private university in northern Mexico, the activity invited small student groups to interact
with structured prompts designed to promote inquiry, collaboration, and reflective reasoning
about the adiabatic process. Rather than functioning as a source of answers, ChatGPT
was intentionally positioned as a mediating scaffold for Socratic questioning, prompting
students to articulate, examine, and refine their reasoning. A mixed-methods approach was
employed, combining a 10-item Likert-scale survey with construct-level statistical analysis
of two focal dimensions: perception of learning and engagement, including an exploratory
comparison by gender. Results indicated consistently high levels of perceived learning
and engagement across the cohort, with average scores above 4.5 out of 5. At the construct
level, no statistically significant gender differences were observed, although a single item
revealed higher perceived learning among female students. Overall, the findings suggest
that the educational value of ChatGPT in this context emerged from its integration within a
Socratic, inquiry-oriented pedagogical design, rather than from the technology alone. These
results contribute to ongoing discussions on the responsible and pedagogically grounded
integration of generative AI in physics education and align with Sustainable Development
Goal 4 (Quality Education).