An application of a model of student attrition at a public urban community college

Susan Cook Mulligan, Fordham University


This study investigated the extent to which Tinto's comprehensive model of traditional student attrition, revised to include external variables, could contribute to the understanding of nontraditional student attrition. The revised model was tested on a sample of 256 full-time and 107 part-time freshmen at an urban community college. A locally designed questionnaire was used to measure the model's core constructs. Students completed the questionnaire during their second semester of attendance. Nonpersisters were defined as students who voluntarily left college prior to their third consecutive semester, before completing their formally declared program of study. Data were analyzed for the pooled full- and part-time samples as well as subgroups based on gender and age. The primary statistical procedures were hierarchical regression and path analysis. The revised model explained from 22% to 79% of the variance in persistence for various subgroups. For only one subgroup, part-time 24 years or less, did the addition of the external variables produce a significant increase in the explained variance. Across groups, one academic system variable, Grade Point Average in the First Year, was the most consistent predictor of attrition. Institutional Commitment, defined as Intent to Leave, was the best single predictor of persistence for three of the four part-time subgroups. Social Integration did not have a significant effect on persistence. With the exception of Social Integration, the core constructs of Tinto's model were supported though the patterns of influence for subgroups varied from theoretical expectations. These patterns were unique for each subgroup, suggesting institution-specific influences that could only be identified through a single-institution study. The results also suggest that pooling groups may mask important differences in attrition patterns. Therefore, attrition studies, particularly at the two-year level, should incorporate subgroup analysis based on status, age, and gender.

Subject Area

Academic guidance counseling|Community colleges

Recommended Citation

Mulligan, Susan Cook, "An application of a model of student attrition at a public urban community college" (1989). ETD Collection for Fordham University. AAI9007186.