Search

Emphasizing graduation rate, W. Bowen and D. Bok (1998) argue that race-sensitive admission at selective colleges enhances the educational attainment of underrepresented minority students, and that the effect increases with college selectivity. Focusing on graduation in science, however, R. Elliott and colleagues (1995) conclude that: (1) preferential admission explains lower rates for underrepresented minority groups; and (2) students of given interest and preparation will fare better at less selective colleges. These hypotheses were examined using Bowen and Boks College & Beyond data from 24 institutions. Among students initially intending science, blacks were less likely than whites to graduate in science (40% versus 56%), and women less likely than men (47% versus 61%). Consistent with the first conclusion of Elliot et al., differences in Scholastic Assessment Test mathematics scores fully accounted for the ethnic disparity and reduced the gender disparity. Support was not found for a positive effect of college selectivity. The estimated effect was negative, though this is qualified by the studys restricted selectivity range and correlational design. A prospective science student is advised to choose the college where students with academic qualifications and interests similar to his or her own fare best in science. Four appendixes contain technical notes; a discussion of science, mathematics, and engineering definitions; and tables of study data. (Contains 4 tables, 9 figures, and 52 references.) (SLD)

Benefits and limitations of structural equation models for multivariate experiments with incomplete data are presented. Examples from studies of latent variable path models of cognitive performance illustrate analyses with latent variables, omitted variables, randomly missing data, and nonrandomly missing data. (SLD)