Predicting Graduation Rates at Liberal Arts Colleges
Date of Award
2009
Document Type
Thesis
Degree Name
Bachelors
Department
Social Sciences
First Advisor
Hernandez, Sarah
Keywords
SAT, Liberal Arts, College Admissions
Area of Concentration
Social Sciences
Abstract
In this study I investigate whether SAT validity at liberal arts institutions differs from research institutions. I examined the usefulness and fairness of SAT scores at liberal arts and research institutions. For this research I rely on institutional data from education databases, identifying a random sample of 166 institutions (83 liberal arts colleges and 83 research institutions). Variables included average SAT score, high school class rank, class size, financial resource ranking, and graduation rate. I used multiple regression analyses and partial correlation analyses in order to compare the validity of SAT as a predictor of graduation. The findings suggest that SAT scores contribute little to no additional information regarding a student's potential to graduate than what is already provided by high school performance indicators at liberal arts institutions. At research institutions, SAT scores do provide additional information regarding a student's potential to graduate. Overall, SAT scores explained 25% less variance in graduation rates at liberal arts institutions. Given my findings and earlier research reporting social biases in SAT, I call into question the usefulness and fairness of SAT scores as an admission factor at liberal arts institutions.
Recommended Citation
Moore, Bradley, "Predicting Graduation Rates at Liberal Arts Colleges" (2009). Theses & ETDs. 4157.
https://digitalcommons.ncf.edu/theses_etds/4157
Rights
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The New College of Florida, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.