Extracting, Structuring, and Analyzing Log File Data to Make Different Thinking Profiles Visible

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

Abstract

This study explores the possibilities of using log data analysis in an educational context to make thinking and learning differences visible and to gain a better insight into students’ thinking in a dynamic problem-solving environment. The sample for the study was drawn from first-year university students (N=2211). Time-on-task and latent class analysis was employed to find students whose thinking strategies show similar patterns and who have thus proved to be qualitatively similar thinkers during the problem-solving process. Results confirmed that educational data mining and log file analysis can be used to make students’ thinking visible and to quantify qualitative differences in thinking. This study contributes to an understanding of how students interact with and think while solving unknown and complex problems.

Authors

• Gyongyver Molnar, University of Szeged

• Beno Csapo, University of Szeged