Computerized Classification Self-Assessments for Predicting and Supporting Learners' Performance

• In Event: Critical Thinking and Technology-Supported Learning

Abstract

Technology- and analytics-enhanced assessments may provide multiple benefits for stakeholders. However, data analytics approaches currently fail to make full use of educational technology and data for assessment. Therefore, this project thought to implement a computerized classification testing system into a productive learning environment and investigate N = 109 students' usage behaviors as well as possible relations to their final exam performance. An in-depth principal component analysis revealed that specific metrics can explain students' behavioral engagement structure and that the use of the analytics-enhanced self-assessments is related to the final exam performance. As a conclusion, reliable and valid metrics for the design and implementation of computerized classification testing systems are discussed.

Authors

• Muhittin Sahin, University of Mannheim

• Jan Delcker, University of Mannheim

• Dirk Ifenthaler, University of Mannheim