Is High School Performance and Standardized Test Scores as Admission Criteria Enough Considering the Institutional Cost of Misclassification?

Abstract

High school performance and aptitude test scores have been shown to have a marginal relationship with common measures of undergraduate student academic success. This minor relationship suggests a source of admission errors which could contribute to tuition revenue loss. This study’s objective was to answer the following questions: (1) can discriminant functions be constructed that can correctly classify students as individuals obtaining a degree within a reasonable amount of time or individuals that withdrawal early, (2) how efficient are these discriminant functions, (3) do they differ by gender, and (4) what is the estimated institutional cost of misclassifying students. Results indicated that discriminant functions could be developed that correctly classified approximately 60% of students. These discriminant functions were also shown to have similar success rates for males and females. Finally, the estimated institutional cost of misclassifying students along with error rates of occurrence suggest a source of tuition revenue loss and that improved predictability of student potential for academic retention is needed.