Evaluating AI-Assisted Instructional Systems in Secondary Physics: A Quasi-Experimental Study From the United Arab Emirates

Objectives This study investigates the effects of Artificial Intelligence (AI)-assisted teaching on student outcomes in Grade 11 Advanced Physics at the Applied Technology School (ATS) in the United Arab Emirates (UAE). The project is in line with UAE’s strategic vision for Smart Learning Education aimed at discerning whether AI-assisted teaching is an effective teaching method relative to traditional teaching by teachers in enhancing student performance. Methods A quasi-experiment using a pre-test and post-test control group design was used to compare AI-supported learning to traditional teaching, as per the Physics curriculum standards of the School. The AI-supported teaching included adaptive scaffolding, feedback, and inquiry-based guiding, whereas the traditional teaching adhered to a standard lecture-based teaching methodology. The pre-test comparisons between groups were established using means (AI group = 72.2%, Traditional group = 71.7%). Results Results of the post-test showed a pronounced performance benefit for students in the AI-assisted learning group (Mean = 18.00) than those in the traditional learning group (Mean = 14.25). The calculated value of the size of the effect (Cohen’s d = 1.21) showed a large result, which is useful in estimating its effectiveness in learning concepts in upper secondary Physics classes using AI-enabled feedback tools. Conclusions AI-assisted learning supported strong pedagogical values in this UAE secondary Physics learning environment, raising levels of achievement as well as fostering increased learning regularity. This study is placed in context by the state of current K-12 learning in STEM-related topics, reflecting upon opportunities as well as challenges in this area, focusing accordingly upon an implementation-aimed integration of teachers’ support in future-ready educational approaches in Gulf-based science-education institutions.