Detecting pre-knowledge via classification tree-based modeling of responses, response times, and visual fixations

• In Event: Process Data and Data Quality

Abstract

Technology-enhanced learning system (TELS) has drawn much attention in educational assessment recently. To gain behavioral inferences by properly modeling multi-source data collected from TELS, this study proposes an innovative structural classification tree-based model for jointly modeling responses, response times, and visual fixations to detect aberrant test-taking behavior like preknowledge responding.

Authors

• Kaiwen Man, University of Alabama

• Qipeng Chen, University of Alabama

• PEIDA ZHAN