AI IN THE CLASSROOM: CHATGPT'S PERFORMANCE IN INTRODUCTORY PHYSICS

Authors

Shantanu Chakraborty, Valdosta State University, Valdosta, GA, 31698Follow
Bilas Paul, SUNY Farmingdale, Farmingdale, NY, 11735Follow
Ganga Sharma, Fairmont State University, Fairmont, WV, 26554Follow

Abstract

An in-depth evaluation of ChatGPT, a large language model, within the realm of physics education is presented. The study draws on data from several introductory physics courses at SUNY Farmingdale State College, Valdosta State University, and Fairmont State University, covering both algebra-based and calculus-based curricula. By systematically testing ChatGPT with homework assignments, class exams, and final assessments from these courses, the study aimed to evaluate the AI's performance across various physics topics, levels of difficulty, and institutional settings. The selection of these specific courses was driven by the desire to examine how ChatGPT engages with different physics curricula, teaching methods, and academic environments. Through a comparative analysis of ChatGPT's performance, the study provides valuable insights into the role of AI technologies in the field of physics education.

Recommended Citation

Chakraborty, Shantanu; Paul, Bilas; and Sharma, Ganga (2025) "AI IN THE CLASSROOM: CHATGPT'S PERFORMANCE IN INTRODUCTORY PHYSICS," Georgia Journal of Science, Vol. 83, No. 1, Article 67.
Available at: https://digitalcommons.gaacademy.org/gjs/vol83/iss1/67