Hybrid Human-GenAI cognitive apprenticeship: Encouraging pre-service teachers to implement instructional practices to support students' self-regulated learning

To support pre-service teachers in becoming agents of self-regulated learning (SRL), we developed a hybrid Human-GenAI (hH-GenAI) course, which leverages the potential of artificial intelligence within a Cognitive Apprenticeship Model (CAM) through a personalized GenAI-based chatbot ( TeachPal ). Using design-based research, we examined pre-service teachers' implementation of SRL-supporting instructional practices, approach components that affected their self-efficacy judgments, and the role of TeachPal in their professional development. We found that while hH-GenAI CAM provided an effective and scalable approach to support the pre-service teachers in implementing SRL-supporting instructional practices, it fostered engagement in metacognitive discourse to only a limited extent. Furthermore, the interactions and characteristics of the discourse with TeachPal hindered positive experiences and limited realization of the approach's full potential. These findings highlight the importance of the augmented approach that integrates the complementary strengths of human and GenAI while ensuring high-quality discourse in teacher education.