Project-Based Learning with Generative AI as Cognitive Collaborators: A DBR Approach to Higher-Order Thinking

• In Event: Teaching and Learning with AI: Building Literacy, Agency, and Design.

Abstract

In the era of generative AI, educational paradigms must shift to foster higher-order thinking. This study conceptualizes generative AI as a cognitive collaborator and develops a project-based learning model based on Anderson and Krathwohl's taxonomy. Using design-based research (DBR), the model was refined over four iterative cycles incorporating learner feedback. Generative AI played specific cognitive roles supporting students' reflective, critical, and creative thinking. A quasi-experimental design with sixth-grade students compared experimental and control groups using pre- and post-tests measuring reflection, critical thinking, and creativity. Results showed the experimental group achieved statistically significant improvements across all domains. Findings suggest structured integration of generative AI as a cognitive collaborator meaningfully enhances learning outcomes, offering theoretical and practical implications for AI-integrated instruction.

Authors

• Kyungri Kim, Sungkyunkwan University

• Dongho Kim, Seoul National University