Abstract:

An on-going study and data reporting of the admissions, retention, graduation, and program completion of an academic program is critical in student success and retention in higher education. One means of accepting the responsibility and accountability for consistently measuring evidence of student success is to make predictions using analytical tools to track the overall performance of students. When students are provided with resource information and access to support services, successful engagement and retention increases.

This study will examine a model of students’ academic performance based on retention and graduation rates at ten Historically Black Colleges and Universities (HBCUs) in the southeastern region. These predictors variables used are retention and graduation rates. Administrators in higher education can use these results to make informed and intentional data-driven decisions to address policies.

Students at HBCUs succeed at a higher rate than at non HBCUs. HBCUs teach a large portion (20 percent) of the “first-time, full-time Black students” that attend four-year institutions, and they “play a critical role in providing Black students with access to four-year, post-secondary opportunities,” according to several research studies.

Data analytics can be used to track these performance rates. The methodology we are using is DBScan and K-means clustering techniques to identify patterns in student learning outcomes, retention rates, and graduation rates at Historically Black Colleges and Universities. Predicting students’ academic performance has long been an important research topic in many academic disciplines.

To this end, we are very interested in using the knowledge based and aid of effective tools in CDS&E to discover unknown trends and patterns in higher education.

Keywords:

Cluster, retention rates, graduation rates, HBCU, K-Means, DBScan, CDS&E, Higher Education, Algorithms, Data Analytics.