Key points are not available for this paper at this time.
Abstract The surging popularity of generative AI, especially ChatGPT, has evoked both enthusiasm and caution within the education community. To effectively harness the full potential of ChatGPT in educational contexts, it is crucial to thoroughly analyze its impact and suitability for different educational purposes. The current paper aims to contribute to this understanding by investigating the impact of ChatGPT on student interactions in a computer-supported collaborative learning (CSCL) environment when deployed as a conversation agent. Furthermore, we present a concise overview of our iBot system, which seamlessly incorporates ChatGPT's capabilities into CSCL, providing technical guidelines for potential future adoptions in similar educational contexts. Through extensive statistical analysis, we numerically examine how including ChatGPT affects the dynamics of student interactions. Our findings reveal a significant shift in the patterns of student interactions, with a notable increase in active learning interactions occurring predominantly between individual students and ChatGPT, even in the presence of other students. While recognizing that the observed changes may not be universal in every educational context, our quantitative analysis emphasizes the need to consider the potential impact of integrating ChatGPT into collaborative learning scenarios. By shedding light on the effects of ChatGPT in CSCL, this study offers valuable insights for educators and researchers to better understand its role as a conversation agent in collaborative learning settings. It encourages thoughtful consideration of the appropriate use case of ChatGPT and emphasizes the importance of further exploring its potential to enhance collaborative learning experiences.
Kim et al. (Thu,) studied this question.