A Case Study: Exploring Student Academic Performance Data for Actionable Knowledge

For decades, university and college administrators have been concerned with student academic performance regarding time-to-degree and time-to-attrition as well as determining the reasons why students take too long to graduate or leave school before receiving their degree. Although this problem has been researched extensively, the concerns remain to this day. With the advent of big data, data science, and sophisticated visualization approaches, there are more tools at our disposal to analyze and explore student data in the hopes of better identifying or predicting which students are at-risk. However, determining which features to mine remain a challenge for this type of analysis. In this case study, we analyzed 10 years of student data in the College of Computing and Informatics (CCI) of a large state university. Using EventFlow, a tool that analyzes event sequences, we developed a novel knowledge generation approach that can be used to test hypotheses about reasons for student attrition.