Can AI Teach Like Socrates? An Exploratory Study of AI-Student Dialogue in Secondary Physics Education



Author Information

Krishan Sanjay Mahesh, Whitley Secondary School, Singapore

Abstract

While generative artificial intelligence (GenAI) holds considerable promise for education, concerns remain that when used primarily for answer retrieval, it may encourage cognitive offloading rather than meaningful learning. This classroom-based study explores an alternative application of GenAI: whether a custom Retrieval-Augmented Generation (RAG) chatbot can engage students in Socratic dialogue to support deeper conceptual understanding in secondary physics. Nineteen Secondary 4 students in Singapore interacted with a RAG chatbot during a one-hour thermal physics lesson as part of their preparation for the GCE Cambridge O-Level examination. A thematic analysis of the chat transcripts uncovered three prominent patterns of interaction: identifying and addressing student misconceptions, scaffolding support for students with lower conceptual readiness, and challenging more advanced learners to deepen their learning. The chatbot delivered immediate and individualised feedback at scale, enabling dialogic interactions that are often unfeasible for teachers to conduct with multiple students in a typical classroom setting. However, the quality of these interactions was highly dependent on student input. In cases where students provided vague or tangential responses, the chatbot tended to shift towards more explanation-driven replies, limiting opportunities for sustained Socratic dialogue. These findings highlight the potential of AI tutoring systems to facilitate Socratic dialogue in secondary physics education. They also underscore the need for AI design principles that incorporate diagnostic questioning and adaptive scaffolding, as well as instructional strategies that equip students to engage meaningfully with AI tools in education.


Paper Information

Conference: ACE2025
Stream: Design

This paper is part of the ACE2025 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window


To cite this article:
Mahesh K. (2026) Can AI Teach Like Socrates? An Exploratory Study of AI-Student Dialogue in Secondary Physics Education ISSN: 2186-5892 – The Asian Conference on Education 2025: Official Conference Proceedings (pp. 1677-1691) https://doi.org/10.22492/issn.2186-5892.2026.127
To link to this article: https://doi.org/10.22492/issn.2186-5892.2026.127


Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Powered by WP LinkPress

Share this Research

Posted by James Alexander Gordon