| This study attempts to add to the college student dropout literature by examining persistence decisions at private, non-selective university using previously unstudied explanatory variables and advanced econometric methods. Three main contributions are provided. First, proprietary data obtained from a type of university that is underrepresented in the literature, a small, multi-campus private university with an open-enrollment admissions policy, are merged with United States census data to empirically examine the relationship between college student dropout and its determinants. While simultaneously controlling for student background, academic, social, financial, and economic characteristics, the researcher scrutinizes the relationship between college student dropout and a wide range of explanatory variables including but not limited to a direct measure of student social integration, zip-code-based poverty rates, the percent of the adult population with a bachelor's degree or higher, location, and major specificity. Finally, to account for the fact that college student departure is a multi-stage, longitudinal event, cross-sectional models are expanded to consider the temporal dimensions of persistence decisions. Emphasis is placed on a statistical approach new to the persistence literature, a discrete-time, multinomial logit event history model. |
