Validity of a Battery of Experimental Tests in Predicting Performance of Navy Project 100,000 Personnel

摘要

This report summarizes results of a four-phase study that originated as part of the Project 100,000 research effort. The purpose of the study was to develop 'culture fair' aptitude tests that would permit the Navy to identify potentially successful recruits from those who scored low on conventional tests. Nineteen experimental tests/questionnaires were developed to measure practical (as opposed to academic) mental abilities. The experimental instruments were divided into four batteries, each of which was administered to a separate sample ranging in size from 5,000 to 12,000 recruits. The instruments were validated against supervisory performance ratings, rating progression, and retention criteria for sample members. Separate analyses were done for Mental Level IVs, Blacks, and for apprenticeship level (nonrated, undesignated strikers) and technical rating groups. The findings were generally negative. With only a few exceptions, the experimental tests were not valid predictors of on-job performance for any of the subgroups studied, and were less valid than the conventional tests for predicting either job performance or rating advancement. Also, because of the wide variety of 'culture fair' tests evaluated, it is unlikely that paper-and-pencil tests can be found that will identify previously overlooked aptitudes in low-ability populations.