AI Scaffolding for Self-Regulated Learning in Graduate Research: A Mixed-Methods Study

• In Event: AERA Roundtable Session Friday 7:45 am - Gold 2
  In Roundtable Session: Scaffolding Learning with AI

Abstract

This study employs a single-group pre-post design with mixed research methods to explore the value of generative artificial intelligence as cognitive scaffolding in graduate research tasks. Based on Zimmerman's three-phase self-regulated learning model, the research designed a series of coherent sub-tasks for educational technology master's students, providing targeted prompt strategies to stimulate metacognitive abilities. Comprehensive analysis of interaction logs, metacognitive scales, and performance data shows that AI as cognitive scaffolding significantly enhances learners' metacognitive awareness, critical thinking, and reflective dialogue frequency. The study reveals the internal mechanisms and design principles of AI supporting self-regulated learning, providing theoretical foundations and implementation guidelines for effectively integrating AI in graduate research tasks.

Authors

• Jiahui Qi, East China Normal University

• Jun Xu, East China Normal University

• Yutian Gao, East China Normal University