Department of Psychology
Master of Arts
Since little research has addressed the question, the present study sought to determine if potential psychological problems, as measured by the College Adjustment Scales, predict college attrition. For Hypothesis One, 423 subjects were used to assess the predictability of the nine College Adjustment Scales. Using logistic regression, Substance Abuse problems was the only scale found to significantly predict attrition, p <.033. Hypothesis Two examined the incremental predictive validity of the College Adjustment Scales when supplemented with the College Student Inventory, a nineteen-scale instrument designed to predict college attrition. Although the College Student Inventory significantly predicted attrition, R = -.09, p < .05, the College Adjustment Scales could not significantly incrementally predict student departure. It was concluded that the lack of representativeness of the samples was the reason for the inability of the College Adjustment Scales to predict attrition. In both samples, subjects were substantially different from the larger samples from which they were derived in that they were more female, had higher ACT scores, and had higher high school GPAs.
Mattis, Christopher, "A Further Exploration of College Student Attrition: The Predicative Validity of the College Adjustment Scales" (1998). Masters Theses & Specialist Projects. Paper 267.