Abstract Embedding smart learning companions in small-group collaboration can provide sustained questioning and diagnostic feedback,creating productive cognitive conflict and adaptive scaffolding that supports the internalization of scientific argumentation thinking from external dialogic norms to learners’ metacognitive monitoring. To address the challenges of cognitive detachment,inefficient collaboration,and insufficient personalized scaffolding in current science classrooms,this study proposes a deep collaborative teaching model supported by smart learning companions to cultivate scientific argumentation. The model features a teacher-AI-student triad synergy and follows a four-stage process based on argumentation logic: conjecture inquiry,scheme co-creation,empirical investigation,and debate consensus. Relying on mechanisms of cognitive offloading,adaptive scaffolding evolution,and adversarial collaborative internalization,the model utilizes the external heterogeneous intervention of smart learning companions to disrupt group cognitive inertia. This approach drives substantive evidence scrutiny and logical repair among students. Practical application in the case “changes in producing gas” demonstrates that this model effectively overcomes superficial collaboration and facilitates the transition from intuitive experience to standardized argumentation. This study offers an actionable scheme for science education in the intelligent era and delineates the direction of teacher transformation empowered by technology.
TIAN Sai-Qi, GUO Chen-Hui, LIU Xiao-Xuan, ZHENG Yi-Zhuo. Collaborative Teaching Model Supported by Smart LearningCompanions for Scientific Argumentation: Changes in Producing Gas[J]. Chinese Journal of Chemical Education, 2026, 47(15): 72-79.
