Understanding Predictors of Students' Performance in an Online Licensure Test Preparation Course

• In Event: Process Data Modeling and Application in Educational Assessments

Abstract

There is an ongoing “data deluge” as a result of technological trends (e.g., smart devices and computers-based assessment) that is producing large volumes of rich heterogeneous data (e.g., education, health, social network) collected at a velocity (e.g., real-time) that is unprecedented. Data from these technological advances are called big data or process data. This research uses big, process data from an online licensure test preparation course in two states to understand predictors of student performance. Given the volume of the data, this research first employs dimension reduction in each state to reduce the volume of the data. Subsequently, this research uses linear regression and latent change score modeling to explore predictors of student performance in each state.

Authors

• Dakota Wayne Cintron, University of California - San Francisco

• Nina Deng