This study evaluates a ChatGPT-based interactive learning model with the aim of improving the reading skills of English Education students at STKIP Kumala Lampung. Using a quasi-experimental design method, 50 students were divided into an experimental group (with ChatGPT) and a control group (conventional) for eight weeks. The research instruments included comprehension tests, questionnaires, interviews, and observations. The results of the pre-test showed initial equality between the groups. However, post-test results revealed a significant improvement in the experimental group (M = 74.56) compared to the control group (M = 65.24), with a large effect size (Cohen's d = 1.30; t(48) = 4.69, p <.001). Observational data showed higher involvement in participation and discussion in the experimental group. Qualitative findings confirm that ChatGPT improves vocabulary comprehension, motivation, and active learning, despite the risk of reliance on AI. Overall, the findings indicate that the ChatGPT-based interactive learning model effectively improves students’ reading comprehension and engagement. The study implies that ChatGPT can support adaptive, interactive, and student-centered EFL learning in higher education.
Copyrights © 2026