Analyzing University Students’Usage Patterns and Learning Effects in ChatGPT-Based Debate Learning: Focusing on AI Teaching Efficacy, Critical Thinking, and Self-Directed Learning Abilities

With the advancement of generative AI, recent studies in the field of education have actively explored its specific applications and attempted to verify its effectiveness. In particular, generative AI is considered to be especially effective when utilized in interaction-based learning environments, such as debate learning. However, there remains a lack of empirical research that applies ChatGPT in actual educational settings and examines its effects there. To address this gap, the present study implemented ChatGPT-based debate learning with university students, analyzed their patterns of ChatGPT usage, and investigated its impacts. The aim was to derive implications for the practical use of ChatGPT in debate learning. The ChatGPT-based debate learning program was conducted over five weeks with 18 university students enrolled in a course on teaching methods and educational technology. The findings revealed that students utilized ChatGPT in various ways during debate learning, including help-seeking, inquiry-based, information-searching, and critical engagement types. Furthermore, ChatGPT-based debate learning was found to improve students’ AI teaching efficacy and self-directed learning ability, although it did not have a statistically significant effect on their critical thinking skills. Based on these results, the following conclusions were drawn. First, students should be guided to use ChatGPT in a multifaceted manner to accomplish academic tasks. Second, sufficient opportunities should be provided for learners to interact with ChatGPT to achieve various learning goals throughout the process. Third, strategies to promote critical thinking need to be integrated into the process of using ChatGPT. Fourth, students should be supported in using ChatGPT autonomously while completing learning tasks.