Conditional effects of institutional characteristics on the persistence of traditional-age students at four-year institutions of higher education

Among the many expected outcomes of higher education, many of which are difficult to quantify, there remains the clearly discernible goal of persistence toward degree attainment. If institutional variations have differing effects on traditional-age freshmen of varying racial, gender, or socioeconomic characteristics, that information has implications for program development and policy-making aimed at increasing the levels of degree attainment among selected populations. The question of primary importance was: All else held constant, does institution attended (as differentiated by institutional control, selectivity, proportion of expenditures for academic support, proportion of student enrollment that is undergraduate, and average expenditures per student) affect the likelihood of student persistence during the freshman year? A corollary question was whether institutional characteristics had differing effects on the probability of persistence for students of various genders, racial groups, or socioeconomic levels. Guiding the research was the theoretical framework of Vincent Tinto (1975) who described student departure from college as a longitudinal process of interaction between the individual and the academic and social systems of the college. Multi-stage statistical procedures, including multi-factor analysis of variance and logistic regression were used to estimate the effects of institutional variables on the probability of persistence, controlling for student demographic variables, psychosocial factors and precollege education. Survey data collected by the National Center for Education Statistics (NCES) formed the dataset. Variables included characteristics of 722 four-year postsecondary institutions (360 private and 362 public) from the 1982 HEGIS/IPEDS dataset, matched with student survey and high school information for 2,447 students from the High School and Beyond (1980 Sophomore Cohort) and Postsecondary Educational Transcript Studies. Findings substantiate the hypothesis that institutional effects on persistence vary by gender, race, and socioeconomic status. The likelihood of persistence increases for Caucasian and middle income students attending a private institution. Males, Caucasians, and poor students benefit from attending institutions with a high proportion of graduate students. Implications for practice, policy, and research in higher education, both on the macro- and micro-levels are included