Improving Pedagogical Agents for Scaffolding Using Complex Systems Theory: Analysis and Visualization of Trace Data (Poster 6)

• In Event: AERA Poster Session Four
  In Poster Session: Digital Learning and Development

Abstract

Intelligent tutoring systems (ITSs) embed pedagogical agents to scaffold learners’ self-regulated learning (SRL) strategies but do not account for learners’ balance between novel and repetitive SRL behaviors. This paper investigates two participants’ SRL behaviors who vary in their learning outcomes using auto-Recurrence Quantification Analysis (aRQA) and grounded in complex systems theory (CST). Participants were extracted from a larger study (N=59) which collected undergraduate students’ log-file data as they learned about the human circulatory with MetaTutor, a hypermedia-based ITS, while being prompted to deploy SRL strategies. Results found differences in participants’ visualized SRL behaviors. This study establishes CST and aRQA as valid metrics of SRL, revolutionizing how ITSs can capture, analyze, and provide information to pedagogical agents for scaffolding SRL.

Authors

• Daryn Dever, University of Central Florida
  Presenting Author

• Megan Wiedbusch, University of Central Florida
  Non-Presenting Author

• Roger Azevedo, University of Central Florida
  Non-Presenting Author