| This study addressed student retention at Historically Black Colleges and Universities. It was designed to determine to what extent academic performance, engagement, and student gender predict retention at a private rural historically Black university. Archival data were collected from the university, including grade point average, Beginning College Survey of Student Engagement and National Survey of Student Engagement results, and gender for 119 students. The theoretical framework was comprised of William Spady's sociological model of the dropout process, Vincent Tinto's model of student interaction, and Alexander Astin's theory of student involvement. Engagement data was collected from the Beginning College Survey of Student Engagement and the National Survey of Student Engagement results. Four major research questions and hypotheses guided this quantitative study: To what extent do academic performance, engagement, and student gender predict retention? To what extent do engagement and gender predict retention? To what extent do gender and academic performance predict retention? To what extent do engagement and GPA predict retention? The analysis involved a series of binary logistic regressions conducted to assess the extent to which academic performance, engagement, and gender predict retention. The results showed that freshman GPA during the second semester was the only significant predictor of sophomore retention: One-unit increase in Spring GPA increased the odds of returning for sophomore year by a factor of 3.315. This research may be useful to college administrators trying to improve retention.;Keywords: historically Black college or university, grade point average, student engagement, gender, African American, student retention. |
