Improving measurement for graduate admissions

In this study we examined alternative item types and section configurations for improving the discriminant and convergent validity of the GRE General Test. A computer-based test of reasoning items and a generating-explanation s measure was administered to a sample of 388 examinees who previously had taken the General Test. Confirmatory factor analyses indicated that three d imensions of reasoning-verbal, analytical, and quantitative-and a fourth di mension of verbal fluency based on the generating-Explanations task could b e distinguished. Notably, generating explanations was as distinct from new variations of reasoning items as it was from verbal and quantitative reason ing. In the full sample, this differentiation was evident in relation to su ch external criteria as undergraduate grade point average (UGPA), self-repo rted accomplishments, and a measure of ideational fluency, with generating explanations relating uniquely to aesthetic and linguistic accomplishments and to ideational fluency. For the subset of participants with undergraduat e majors in the humanities and social sciences, generating explanations add ed to the relationship with UGPA over that contributed by the General Test.