Using trace data to enhance Students' self-regulation: A learning analytics perspective

Abstract

The purpose of this study was to investigate whether students' self-reported SRL align with their digital trace data collected from the learning management system. This study took place in an upper-level college agriculture course delivered in an asynchronous online format. By comparing online students' digital trace data with their self-reported data, this study found that digital trace data from LMS could predict students' performance more accurately than self-reported SRL data. Through cluster analysis, students were classified into three levels based on their self-regulatory ability and the characteristics of each group were analyzed. By incorporating qualitative data, we explored possible explanations for the differences between students' self-reported SRL data and the digital trace data. This study challenges us to question the validity of existing self-reported SRL instruments. The three-cluster division of students' learning behaviors provides practical implications for online teaching and learning.