Cross-validation of a forced-choice personality inventory

Abstract

The Jurgensen Classification Inventory was administered to 261 male college students to test its ability to predict academic achievement. With grade point average as the criterion an item analysis was completed on 100 subjects, composing the lower and upper 28% groups. Two scoring keys were derived and their validities determined on the holdout group of 82 students by computing Pearson r 's between test scores and the criterion. Both keys gave positive validity coefficients but neither of them was statistically significant. They were .12 (unweighted key) and .16 (weighted key). When the weighted key was used to score the papers of the original 100 group a pseudo-validity coefficient of .76 was obtained. (PsycInfo Database Record (c) 2025 APA, all rights reserved)