The effect of AI-assisted problem-based learning strategies on civic education learning outcomes

Background-This study examines the effectiveness of ChatGPT-assisted Problem-Based Learning (PBL) in improving student learning outcomes in civic education. Traditional PBL often relies on tutor facilitation, which can result in inconsistent support and limited access to learning resources. Objectives-This study addresses this limitation by integrating ChatGPT as a support tool within the PBL process. Method-A quasi-experimental pretest–posttest control group design was used, involving 70 undergraduate students divided into an experimental group (ChatGPT-assisted PBL, n = 35) and a control group (conventional PBL, n = 35). Student learning outcomes were measured using a 25-item multiple-choice test. Data were analyzed using ANCOVA to control for pretest differences. Results-The results showed a significant effect of the instructional method on posttest scores, F (1,67) = 22.13, p < .001. The adjusted mean score of the experimental group (M = 86.02) was higher than that of the control group (M = 78.58), with a large effect size (partial η² = .29). Conclusions-These findings indicate that integrating ChatGPT into PBL can improve student learning outcomes by providing consistent support during problem-solving activities. This study contributes empirical evidence on the use of AI in civic education and supports its integration into problem-based learning environments.