Identifying Characteristics that Influence First-Time, Full-Time Freshmen Persistence and Exploring Effective and Strategic Retention Initiatives for an At-Risk Student Population

The purpose of this research is to understand background and behavioral characteristics that influenced student persistence of first-time, full-time, freshmen at a four-year public institution, The University of Southern Mississippi (USM). This study provided an outline for institutions of higher learning to create a profile assessment on their campus and identify students that were more likely to need additional support in order to be successful. Research has shown that understanding students’ needs can increase student retention on campus (Gerdes & Mallinckrodt, 1994; Briceño-Perriott & Mianzo, 2006; O’Keefe, 2013). Coll and Stewart (2008), explained that research in this field was more reactive because the study was typically conducted after students dropped out of college. Based on the theoretical framework of Vincent Tinto (1975), this study provided an analysis of research regarding student withdrawal and retention. This mixed methods sequential explanatory design consisted of three phases and gained a more comprehensive understanding of student persistence. The first two phases of the study used quantitative data to establish a predictive model and explored variables that influenced the likelihood students would enroll, or not enroll, during their second fall semester. An exploratory and confirmatory factor analyses were conducted to ensure factors or latent variables were valid and reliable. During Phase I, the predictive model primarily focused on a student’s individual background characteristics which included academic preparedness from high school, gender, ethnicity, family background including education level and income, and other variables from a student’s admissions and financial aid application. During Phase II, the background characteristics established during Phase I and the behavioral characteristics were used for a second and final predictive model using logistic regression. A student’s behavioral characteristics included engagement and motivation, goal commitment and procrastination, college choice and institutional commitment, and expectations or adjustment to college. Students with characteristics that were reported as significantly influencing student persistence were considered the at-risk population at USM. Therefore these characteristics could allow administrators, faculty, and staff an opportunity for early intervention. The final phase used qualitative data to further explore the at-risk population at USM and examine how this research could affect university policy and practice