Abstract:

Artificial Intelligence can be integrated into physical sciences teaching by employing interactive language model chatbots like ChatGPT to improve critical educational outcomes. The study investigated the effects of self-efficacy, motivation, satisfaction, and learner engagement on conceptual understanding when ChatGPT is integrated into physical sciences lessons. The self-regulatory learning model was the theoretical framework for the study. A cross-sectional survey design was employed in this quantitative study. Four schools were randomly selected from 54 secondary schools in Vembe East District of Limpopo in South Africa. A questionnaire designed to measure the learners’ self-efficacy, motivation, satisfaction, engagement and conceptual understanding was used in data collection. Data collection occurred after learners were exposed to learning physical sciences with ChatGPT integrated into their lessons. The sample size from the four schools was 398 learners, and all of them responded to the questionnaire. Structural equation modelling was used in data analysis. The study has implications for teachers’ integration of artificial intelligence tools like ChatGPT in teaching physical sciences. The findings extend our understanding of the practical implications of the self-regulatory learning model when integrating ChatGPT into instructional practices.

Keywords:

Artificial Intelligence, ChatGPT, Physical sciences, self-regulatory learning model, Self-efficacy, Motivation, Satisfaction, Learner engagement, conceptual understanding.