AI-mediated learning: ChatGPT explanation quality as a predictor of students’ confidence in physics learning

The integration of artificial intelligence (AI) tools, such as ChatGPT, in education has the potential to support students’ understanding of complex subjects, such as physics. However, empirical evidence regarding how the quality of AI-generated explanations influences students’ learning confidence remains limited. This study investigates whether the quality of explanations generated by ChatGPT predicts students’ confidence in learning physics among high school students. A total of 167 high school students in Indonesia participated in the study. Data were collected using a structured questionnaire that measured students’ perceptions of ChatGPT's explanation quality and their confidence in learning physics. The data were analyzed using descriptive statistics and multiple linear regression. The results indicated that female students rated ChatGPT’s explanations more positively than male students. Regression analysis revealed that explanation quality dimensions related to understanding basic principles and accurate application of physics concepts were significant predictors of students’ confidence in physics learning, particularly in relation to exam readiness. These findings indicate that certain dimensions of ChatGPT explanation quality, particularly those related to conceptual clarity and accurate application of physics concepts, are associated with higher levels of students’ perceived confidence in learning physics. The study contributes to the emerging literature on artificial intelligence in science education by identifying specific dimensions of AI-generated explanation quality that are associated with students’ perceived confidence in learning physics.