A paper co-authored by Chaoli Wang, associate professor of computer science and engineering at the University of Notre Dame, received the Kostas Pantazos Memorial Award for Outstanding Paper in Visualization and Data Analysis at the IS&T International Symposium on Electronic Imaging earlier this month. Presented at the conference by Maggie C. Goulden, a student at Trinity College Dublin and former Naughton Research Experiences for Undergraduate student at Notre Dame, the paper demonstrated a new way to visualize and analyze the online learning behaviors of college students in order to increase retention rates and enhance learning outcomes.
Retention rates for students have been linked to student performance. Thus, better understanding learning behaviors is an important topic as higher education across the country continues to face shrinking retention and graduation rates. In fact, in its most recent study, the National Center for Education Statistics cited that the retention rate of U.S. college students [those returning to school from year to year] was 81% while the six-year graduation for those same students was only 60%.
In this paper, Wang and co-authors [Goulden; Eric Gronda of the University of Maryland, Baltimore County; Yurou Yang and Zihang Zhang of Zhejiang University; and Jun Tao, Xiaojing Duan, G. Alex Ambrose, Kevin Abbott, and Patrick Miller of the University of Notre Dame] introduced a novel tool for analyzing course clickstream data. The visual analytics tool they developed, Course Clickstream Visualization (CCVis), goes beyond the tracking of assignments and student behaviors employed in Learning Management Systems (LMS) such as Sakai, Moodle or Blackboard. Specifically, it employs higher-order network and structural identity classifications that enable visual analytics of students’ activity patterns in the online portions of LMS courses to analyze, categorize and summarize those behaviors, including a detailed comparison of group and individual student behaviors. The ultimate goal of CCVis is to help instructors monitor and manage student progress and performance.
The work reported in the paper was supported in part through the National Science Foundation. It targeted an introductory course with more than 2,000 students. The next step, according to Wang, will be to improve the visual encodings of CCVis to be able to collect student writing and instructor feedback for text mining. This would allow instructors to better spot at-risk students. Other refinements could include customization for administrators [to more accurately estimate student retention], as well as students so that they can compare their performance to that of peers [potentially promoting self-motivation].
Wang joined the University in 2014. His research interests include scientific visualization, big data analytics, high-performance visualization, information visualization, user interface and interaction and visualization in education.
Originally published by at conductorshare.nd.edu on January 24, 2019.