Examining Learning Engagement Through Sequential Pattern Mining (Poster 15)

• In Event: AERA Poster Session Four
  In Poster Session: Digital Learning and Development

Abstract

This study examined college students’ behavioral and cognitive engagement in an online class through a learning analytics approach. Self-report and log data from 254 undergraduate students were used to model students’ engagement quantity and quality. The study findings highlight the importance of learning analytics in modeling students’ behavioral and cognitive engagement in online learning. With two iterations of modeling, our results suggest that combining log data as indicators of behavioral engagement with self-reported cognitive engagement yielded more meaningful clusters as compared to clustering with log data only. The resulted three clustering groups showed significant features in both frequency and time-on-task of behavioral engagement, deep learning strategies, as well as behavioral transition sequences.

Authors

• Kui Xie, Michigan State University
  Non-Presenting Author

• Zilu Jiang, The Ohio State University
  Non-Presenting Author

• Zilong Pan, Lehigh University
  Presenting Author

• Qiwei Men, The Ohio State University
  Non-Presenting Author

• Jingwen He, Michigan State University
  Non-Presenting Author