Curricular Choice and the Differential Prediction of the Tertiary-Academic Performance of Men and Women

Abstract

American research findings suggest that the tertiary-academic performance of men and women is predicted differentially in the sense that for the same high-school and aptitude-test performance, the actual grades achieved by women are higher than those predicted for them, and that for men, the situation is reversed. In the present study, three hierarchical regressions were performed to predict the extent of differential prediction for 329 men and 470 women who registered as first-year students at the University of the Orange Free State in 1992. With the two scales of the GSAT, matriculation symbol point total (MSPT), and a combination of these two predictors, gender (intercept) differences accounted for 4.55%, 2.36% and 2.18% of variance in accumulated mean curriculum percentage marks (MCPMs) over a three-year period. Next, hierarchical multiple regressions, in which the categorical variable of curricular choice was specified to enter the equation ahead of gender, were performed to predict the deviations of students' predicted MCPMs from their actual MCPMs (as indices of differential prediction). When cur-ricular choice was controlled through this semi-partial correlational approach, the percentage of these indices' variance that was explained by gender was reduced by between 53% and 78%. In the process the percentage of variance explained by gender was virtually wiped out: When the index of differential prediction was based on MSPT and a combination of the GSAT and MSPT, gender explained less than one per cent of the variance of these indices.