The effect of financial aid on student persistence in a two-year institution

Date of Completion

January 1999


Education, Community College|Education, Finance|Education, Guidance and Counseling




This study utilized an ex post facto design to examine the relationship between student persistence and receipt of financial aid in the non-traditional student population at a Southwestern Connecticut community-technical college. The sample population size was 739 and consisted of matriculated freshmen. Institutional databases were used to gather student demographic, academic achievement and financial aid information. ^ This study spanned a four-year period from the Fall 1994 through Spring 1998. A comparison of the persistence rates of non-aided students and that of aided students revealed that financial aid recipients persisted longer, but the differences in persistence rates were not statistically significant in the final study model. ^ An evaluation of aid types along with total aid amounts disclosed that loan recipients persisted at a much higher rate than grant recipients. Total aid amount was significant beyond the .01 level.^ Differences between persisters and non-persisters in relation to various demographic variables (age, gender, income group, and ethnic background); academic achievement variables (college grade point average, college program, placement test scores, and college enrollment status); and financial aid variables (type of aid, receipt of aid, and total amount of aid) were investigated using linear probability models with the maximum likelihood algorithms contained in the GLM Multivariate procedure in SPSS. Those models were further tested using logistic regression analysis. The number of semesters students persisted, along with the cumulative number of credits were also used to examine group differences to obtain a more complete and detailed description. ^ In the final model, the three variables, in order of importance, were cumulative grade point average, total amount of aid received, and enrollment status (full- or part-time). When cumulative number of credits earned was used as a criterion, the model accounted for 33% of the total variability of persistence. ^ A logistic regression showed estimated income to also be a significant predictor. In this sample, the students in the lowest income group had the highest rate of persistence. ^