Name: S. Anne Josephine Mary & M. Vakkil
Country: India

Year: 2026
Volume: Vol-16, Issue-2, Year-2026 (July-December)
Page Number: 7-14
DOI: https://doi.org/10.5281/zenodo.21371054

ABSTRACT
Artificial intelligence (AI) tools are increasingly proposed as a means of improving conceptual understanding and measurable achievement in physics, a subject that higher secondary students consistently find demanding because of its abstract theoretical content and heavy dependence on mathematical reasoning. This study investigates the effectiveness of an AI-supported instructional intervention in improving physics learning outcomes among 45 higher secondary students. A single-group pretest-posttest pre-experimental design was adopted, in which the sample was taught a defined physics unit using an AI-based learning tool (an adaptive tutoring or conversational AI system providing stepwise problem-solving support, instant feedback, and personalised practice) over a six-week period, with achievement measured before and after the intervention using a teacher-constructed, content-validated Physics Achievement Test. The study is grounded in constructivist learning theory, Vygotsky'