Update from 2/1/18 Blackboard Analytics Symposium: Using Bb Predict and UMBC's FYI alert program together is more accurate than using either service alone.
------------------------------------------
Recent high profile articles in both the Baltimore Sun and the New York Times have highlighted the important and growing role of analytics in higher education. UMBC in particular is working to develop accurate means to predict and subsequently support student academic success. Of particular interest is faculty and student Blackboard (Bb) Grade Center use.
Based on a pilot project looking at UMBC Bb data for the past two years, it is possible to predict with 83% accuracy whether or not a student will earn a C or better in any given course before the term begins using prior academic and demographic characteristics. By the start of week 4, however, student grades in a Bb course become the most important predictor throughout the remainder of the term. The model was developed using an 80% random sample of all UMBC Bb courses, tested against the remaining 20%, and included several demographic and academic factors. It was 83% accurate thru week 3, 87% accurate by week 4, and 93% accurate by week 12 (final 20% of a 15-week term).
These findings echo prior research on first year students in UMBC Bb courses. Specifically, when viewing student activity and if faculty enabled the Grade Center, students were up to two and a half times more likely to earn a C or better in Bb courses that used the Grade Center compared to courses that did not. The highest student activity and Grade Center interaction effect occurred among Freshmen (1). These findings also complement research by the Educause Center for Applied Research (ECAR) that students value faculty use of the grade center more than any other LMS function (2). Finally, a recent Bb study found that students’ frequent checking their grades online was highly predictive of their course success (3).
By increasing the use of of the Grade Center, faculty can enhance students’ potential for self-efficacy while furthering the effectiveness of systematic efforts to enhance student engagement.
(1) Fritz, J. L. (2016). Using analytics to encourage student responsibility for learning and identify course designs that help (Ph.D.). University of Maryland, Baltimore County, United States -- Maryland. Retrieved from http://umbc.box.com/johnfritzdissertation, pp. 168-169.
(2) Caruso, J., & Salaway, G. (2007). The ECAR study of undergraduate students and information technology, 2007 (Key Findings) (pp. 1–15). Educause Center for Applied Research. Retrieved from https://net.educause.edu/ir/library/pdf/ers0706/ekf0706.pdf.
(3) Young, J. (2016, September 7). What Clicks From 70,000 Courses Reveal About Student Learning. The Chronicle of Higher Education. Retrieved from http://www.chronicle.com/article/What-Clicks-From-70000/237704/