Early Predictors of Student Completion at a Two-Year Institution

Abstract

This quantitative, correlational dissertation research examined if, and to what extent, student early academic success, student socioeconomic status, and student classification as a member of an underrepresented minority predicted student completion at a southeastern Washington two-year institution. The theory of behavioral economics, the iPASS model, and the SSIPP framework served as the foundation for the research. Archival data from 1939 certificate or degree-seeking students first enrolled in higher education at a southeastern Washington two-year institution in the summer or fall quarter of 2016 represented the sample. Research question one asked if student early academic success, operationalized as student first-term GPA, predicted student completion at a two-year institution. Research questions two and three asked if socioeconomic status, operationalized as eligibility to receive the Pell Grant or the Washington College Grant, or both, and classification as a member of an underrepresented minority, operationalized as ethnicity, predicted student completion at a two-year institution. Results from the binomial logistic regression analysis indicate the model was statistically significant, χ²(13) = 230.420, p < .001. The model explained between 11.2% (Cox & Snell R²) and 15.3% (Nagelkerke R²) of the variance in student completion and correctly classified 66.1% of the cases. Sensitivity was 28%, specificity was 89.2%, positive predictive value was 61%, and negative predictive value was 67%. Student first- term cumulative GPA and SES were statistically significant in predicting student completion; student ethnicity was not statistically significant in predicting student completion.