Development of AI-based learning media in improving the problem-solving ability of middle school students' moments of inertia

Abstract

Problem solving ability in physics, particularly on the topic of moment of inertia, remains a challenge for secondary school students because the concept is abstract and highly mathematical. This study aims to develop an artificial intelligence (AI)-based learning media and to examine its feasibility and effectiveness in improving students' problem solving ability on moment of inertia material. The research employed a Research and Development (R&D) design following the ADDIE model (Analysis, Design, Development, Implementation, Evaluation). The subjects were 64 students of grade XI Senior High School selected through purposive sampling.

Data were collected using a problem solving ability test, a media feasibility validation sheet, and a student response questionnaire, then analyzed descriptively and through a paired-sample t-test. The results show that the AI-based learning media was categorized as "very feasible" based on expert validation with an average score of 89.2%, and there was a significant increase in students' problem solving ability after using the media, with an N-gain score of 0.71 (high category). It can be concluded that the developed AI-based learning media is effective and feasible to improve students' problem solving ability on moment of inertia material.