Formulating a Prediction Model of Retention Rate in the University of North Carolina System

Attrition;Persistence;Retention rate;Retention model

Ukpabi, Chinasa Victor ; Robert Serow, Committee Chair,Duane Akroyd, Committee Co-Chair,Dennis Daley, Committee Member,Troy Chen, Committee Member,Ukpabi, Chinasa Victor ; Robert Serow ; Committee Chair ; Duane Akroyd ; Committee Co-Chair ; Dennis Daley ; Committee Member ; Troy Chen ; Committee Member

University:North Carolina State University

关键词: Attrition;    Persistence;    Retention rate;    Retention model;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/3014/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

Institutional effectiveness has become an issue in American Higher Education as governing bodies now require evidence of quality in accountability and funding issues. One measure widely used today to assess effectiveness is retention rate. In response to the above as well as to demographic changes, increasing costs, and the intense competition for new students, educational institutions are seeking new methods to increase the retention rate of their institutions.Most research on retention focused on individual student- level variables, which only predict persistence. There is an obvious need to understand the impact of some uncontrollable external influences on retention rates. In light of the importance that retention rate has assumed, this study sought to develop a predictive model of retention rate in the 16-campus University of North Carolina System.In an effort to develop a comprehensive model, this study employed selected institution-level variables. The study will fill a void, as the UNC system does not currently have a general statistical model for predicting retention rates in its multi-campus system.The central research question for the study is whether an institution's retention rate is a function of the demographic characteristics, economic conditions, college management, and fiscal policy of the state in which the institution resides?Pooled cross-sectional time series technique was employed and the method of Ordinary Least Squares (OLS) was used in the estimation of the regression equation. Data was pooled for ten years to provide greater number of data points to overcome a potential degree of freedom problem that would arise studying only 15 institutions. Four themes emerged from the analysis: headcount enrollment, amount of education and general expenditure on instruction and academic support, the county population where institution is located and the rate of unemployment in the county, are significant predictors of retention rate for an institution.Future research including developing prediction models for minority institutions and employing only external variables will perhaps provide additional insights in our understanding of retention rate behavior.