The Differential Aptitude Tests as Predictors of Success in a High School for Engineering Program

Abstract

This study examined (1) the predictive validity of the Differential Aptitude Tests (DAT) relative to achievement criteria for seniors in a high school for engineering as measured by (a) the students' ( N = 89) grades in engineering, mathematics, and science courses and (b) cumulative grade point average (CGPA) and (2) the differential validity of the 8 DAT scales in comparison of high ( N = 35) and low ( N = 54) achievers. The following selected results became apparent: (1) Numerical Ability (NA) scores were correlated .39 and .31, respectively, with performance in science courses and CGPA at the .001 level of significance and .25 with performance in mathematics courses ( p < .01); (2) Abstract Reasoning (AR) scores were correlated .46, .47, and .49, respectively, with performance in engineering, mathematics, and science courses ( p < .001) and .48 with CGPA ( p < .001); and (3) corresponding correlations of .38, .37, .49, and .44—all significant beyond the .001 level—were found between Space Relations (SR) and each of the same measures cited in the previous clause. In fulfilling the second purpose, a two-group discriminant analysis involving the 8 DAT scales yielded a canonical correlation of .88 ( p < .001) corresponding to a chi-square of 126.67 ( df = 5). The Verbal Reasoning (VR), NA, Clerical Speed and Accuracy (CSA), SR, Spelling (SP), and Language Usage (LU) scales were each valid discriminators of the high-achieving and low-achieving students (all p < .01). It was concluded that the results supported continued use of the AR scale as an admissions criterion. It was recommended that the study be replicated with similar populations and that follow-up investigations involving cross-validation be conducted for the students who decide to further their engineering programs at the college level.