Human-AI Collaborative Learning and Critical Thinking Development: Evidence from a Technology-Enriched Graduate Course

• In Event: AERA Roundtable Session Sunday 1:45 pm - Gold 3
  In Roundtable Session: AI in Student Learning: Programming, Writing, and Engagement

Abstract

With the rise of Generative AI (GAI), human-AI collaboration is increasingly integrated into higher education, offering new learning opportunities while posing risks to critical thinking. This quasi-experimental study involved 28 graduate students in a 12-week course using a multi-agent AI system. Pre/post questionnaires and dialogue coding examined changes in critical thinking disposition, ability, and interaction patterns across different skill levels. Findings show the AI-supported environment significantly improved students’ critical thinking disposition but had limited effect on ability. Distinct dialogue behavior patterns emerged among groups and across learning phases. The study provides empirical insights into AI’s role in enhancing critical thinking and offers guidance for designing effective human-AI collaborative learning environments.

Author

• Yinuo Li, university of bristol