Gaining Insights on Student Success using Attendance Data

Findings from Two Terms of Qwickly Pilot

During the FA18 and SP19 semesters, DoIT partnered with Math/Stat to assess the relationship between student attendance and academic outcomes leveraging laptop and mobile technology. Using a tool provided by a third-party vendor, Qwickly Attendance, faculty were able to keep attendance within their courses in the learning management system (Blackboard). More than 8,000 students, across 239 Blackboard courses participated during the FA18 term. The SP19 included 125 courses that made use of the tool for attendance, totaling 4,487 records. Sixteen of these courses were in Math/Stat, and represented 1,181 records.

There was a wide range of attendance use, from instructors using the tool only once, to others making use of it consistently throughout the term. The findings from this initial term indicated that students who attend class more regularly were less likely to be identified as at-risk at the semester midpoint and less likely to earn a DFW in their respective courses.

Overall, there was a statistically significant, negative relationship between absenteeism and receiving an FYI during both terms. As attendance increased, the chance of receiving an FYI decreased. During the FA18 term, of the students enrolled in courses that used Qwickly attendance, those who received an FYI attended on average about 8% fewer sessions than those students who did not receive an FYI (78.8% vs 86.6%). There was a similar pattern when looking only at the Math/Stat courses using the tool, although with slightly lower attendance rates. In these courses, students who received a DFW attended class almost 30% less than their peers (47% vs 76%).

There was a minimal reduction in class-level DFW rates for those sections participating in the pilot compared with the same courses taught by the same instructors between the FA17 and SP18 terms (22.24% vs. 23.65%), while overall DFW rates for all Math/Stat courses across these terms held steady during this same time period (19.73% vs. 19.52%). In turn, there was not strong evidence to suggest taking attendance on its own impacts students’ academic outcomes. However, since poor attendance is positively correlated with adverse academic performance, the signal from students’ attendance might be used to identify and remediate risk-behavior patterns, particularly since there remain courses that do not make use of the FYI system.

The data resulting from this pilot indicate absenteeism is associated with poorer academic outcomes, and suggest the signal from this behavior might be leveraged along with existing messaging intervention techniques to remediate the pattern during a semester to positively impact students’ chances of academic success. Kathleen Hoffman, Associate Dean of the College of Natural and Mathematical Sciences, notes how the use of the tool is beneficial to "quickly and easily track student attendance” and can help support student self-efficacy since "data gathered in UMBC math classes shows students the direct relationship between their grade in the course and their attendance in class."

Posted: September 3, 2019, 9:27 AM