Prior Knowledge, Confidence, and Their Effects on STEM Performance and Self-Regulated Learning Patterns

• In Event: AERA Poster Session Thursday 2:15pm
  In Poster Session: Assessing and Supporting Self-Regulated Learning Across Learning Contexts

Abstract

This study investigated relationships between prior knowledge, confidence, and self-regulated learning (SRL) in engineering design tasks among 167 middle and high school students. Hierarchical clustering identified three profiles: confident beginners (high confidence, low knowledge), cautious beginners (low confidence, low knowledge), and knowledgeable skeptics (high knowledge, low confidence). Both beginner groups achieved significantly higher learning gains than knowledgeable skeptics, challenging traditional assumptions about prior knowledge and learning outcomes. Epistemic network analysis revealed knowledgeable skeptics had weaker goal-evaluation connections but stronger ineffective adaptation links compared to beginners. Cautious beginners showed the strongest goal-oriented connections, suggesting that knowledge-confidence alignment promotes effective SRL. Findings validate the SRL model in engineering design contexts and highlight the importance of confidence calibration in STEM education.

Authors

• Juan Zheng, Lehigh University

• Shiyi Liu, Lehigh University

• Shan Li, Lehigh University

• Charles Xie, Institute for Future Intelligence