Impacts of generative AI on student teachers' task performance and collaborative knowledge construction process in mind mapping-based collaborative environment

Collaboration has long been recognized as an efficacious pedagogical strategy. With the visual pedagogical scaffold, for instance, mind maps, learners can externalize their comprehension and engage in meaning negotiation. However, given the limitations posed by the conventional collaborative environment such as the difficulty for groups to achieve deep discussions and generate group intelligence, the generative artificial intelligence (GAI) tool was introduced. The efficacy of the GAI tool in boosting task performance is still contested and its influence on collaborative knowledge construction remains unclear. Therefore, this study was conducted in an elective course at a university. A total of 30 student teachers were the participants, including 10 groups of 3 students. In the subsequent quasi-experimental study, 5 experimental groups employed the GAI tool and mind mapping based collaborative environment (GMMCE) and 5 control groups employed the conventional mind mapping based collaborative environment (MMCE). The results showed that the experimental groups outperformed the control groups on both collaborative tasks in the course. According to Ordered Network Analysis (ONA), the experimental groups developed a collaborative knowledge construction process in cognitive dimension, i.e., progressive interaction from individual to peer to group. On the contrary, the control groups were mainly reflected in the transition between learners' individual expression and peer interaction since they were more centered on the regulative dimension. We also supplemented students’ perception of using the GAI tool from the interview data. Finally, we highlighted the implications for pedagogy and described the research limitations as well as future directions.