Investigating the propensity of secondary and tertiary physics teachers for self-regulated learning using ChatGPT

Abstract

Recent studies have shown that artificial intelligence (AI) platforms based on large language models can be utilized for self-regulated learning (SRL). However, the propensity of using AI tools for SRL in physics teaching has not been fully investigated. In terms of physics education, most research on AI gravitates toward student use and consumption, with some studies mostly reporting on the limited accuracy of AI platforms as student tutoring systems in problem-solving. This study investigates the use of AI in physics education, specifically for teachers. Seventy-eight teachers (with physics and non-physics backgrounds) were surveyed about the perceived usefulness (PU), ease of use (PE), behavioral intentions (BI), and actual usage of ChatGPT. Results showed that teachers with non-physics backgrounds rated higher in PU, PE, and BI. Regression analysis indicated PE as a stronger predictor of BI than PU, but BI did not significantly predict actual usage, suggesting barriers like time constraints. Frequent ChatGPT use correlated with improved SRL skills, highlighting its potential to enhance physics education. Future research should investigate additional factors and long-term effects on SRL skills.