Investigating Collaborative Knowledge Construction Among Pre-Service Teachers Supported by ChatGPT in a Learning Community

Leveraging artificial intelligence technologies to enhance collaborative knowledge construction in teacher research activities has become essential for improving instructional competence and fostering professional development. To investigate the effectiveness and optimization strategies of teacher research empowered by ChatGPT, this study conducted a quasi-experimental study, subsequently employed Chi-square tests and Lag Sequential Analysis to perform in-depth analyses of discourse patterns in teachers' collaborative knowledge construction within ChatGPT -empowered contexts. The findings revealed that ChatGPT -empowered contexts facilitated more practice-oriented and coherent collaborative knowledge construction among teachers. Participants not only shared more constructive information but also demonstrated enhanced fluidity and iterative refinement in social interactions. Furthermore, teachers with higher prior knowledge levels exhibited superior collaborative knowledge construction performance compared to their less experienced counterparts. Based on these discoveries, this study proposes three strategic directions for optimizing ChatGPT applications in teacher research: 1) Refining technical implementation scenarios of ChatGPT systems, 2) Prioritizing quality enhancement of collaborative knowledge construction processes, and 3) Addressing the moderating effects of teachers' prior knowledge levels. These recommendations aim to advance the intelligent transformation of teacher professional development through targeted technological empowerment.