Modeling student success: A longitudinal study of Black students and White students

Community Colleges are enrolling an increasingly diverse student population, representing more minority group members and students with English as a second language. One minority group, American Black students, are disproportionally enrolled in community colleges and yet do not succeed at nearly the rates of groups of students. This study was designed to identify factors that are associated with student success and to use this information to predict which students are most likely to be at-risk students, that is are least likely to succeed. In keeping with this objective, American Black students' and American White students' progress and degree of success were assessed at a large East Coast community college. In the aggregate, the background of Black students was often disadvantaged, compared to White students, with lower family incomes, less parental education, and less college preparedness. In college, Black students collectively attended more frequently on a part-time basis, less frequently entered with college-level English or math skills, and more often had need of financial assistance than did White students. Black students succeeded at each of three levels of success, as defined within this study, less frequently than did White students. However, controlling for family educational levels, statistically significant differences between the performance of Black students and White students at Levels I and II were negated. An iterative chi-squared technique (CHAID) was used to identify potential independent variables to be used in constructing a predictive model. Logistic regression techniques were utilized to identify students that were most likely to be non-successful. Intervention strategies were recommended to improve the success rates of these students. The research methodology employed in this study may be custom fitted to other educational institutions to provide administrators and decision-makers with a sophisticated methodology that is both fairly inexpensive and yet is capable of identifying potentially at-risk students. Future research is needed to establish more refined measures of success and to further develop and generalize the methodological approach developed in this study.