The Impact of ChatGPT-Assisted Learning on Students' Understanding of Circular Motion in Grade 10 Physics

Abstract

This study investigated whether a generative artificial-intelligence learning assistant (ChatGPT) improves tenth-grade students' conceptual understanding of circular motion. A quasi-experimental pre-test–post-test control-group design was employed at one Indonesian private senior high school (N = 46). Class X-2 (n = 25) received two weeks of ChatGPT-supported instruction, while Class X-1 (n = 21) followed conventional teacher-led lessons. Conceptual knowledge was assessed with a 15-item multiple-choice test (KR-20 = 0.82) administered before and after the intervention. The normalized gain (g) of the experimental group (M = 0.54, SD = 0.18) exceeded that of the control group (M = 0.48, SD = 0.17). An independent-samples t-test confirmed that the difference was statistically significant, t(44) = 2.11, p = .041, d = 0.62 (medium effect). Teachers can leverage ChatGPT prompts to provide on-demand explanations, contextual examples, and individualized practice that complement textbook explanations of abstract physics concepts. The short intervention period and single-school sampling limit generalizability; future work should employ longer treatment durations and more diverse contexts. This article offers one of the first controlled studies in Indonesia investigating large-language-model support for senior high school physics learning.