Applying Text Mining to Identify Critical Factors Contributing to the Retention Rate of First-Year Students at Historically Black Colleges and Universities

Between 2007 and 2022, four-year Historically Black Colleges and Universities (HBCUs) consistently exhibited lower first-year student retention (FYSR) rates compared to four-year non-profit higher education institutions that are not HBCUs. The average difference over this 14-year period was approximately 14% between 2011 and 2022. The lower FYSR rates at HBCUs may indicate administrative challenges, potentially impacting their ability to provide educational services for Black students effectively. To accomplish this, we analyzed 11,993 students’ online comments from 2010 to 2022 across 53 four-year HBCUs with low FYSR rates. We utilized latent Dirichlet allocation to construct topics and n-grams to identify positive and negative word sequences from each constructed topic. Furthermore, we conducted a university-level regression analysis to examine the relationship between student satisfaction ratings on each topic and FYSR rates. Finally, we discuss strategies to improve the FYSR rates at these HBCUs.