Key points are not available for this paper at this time.
The effectiveness of precollege and college achievement measures in predicting persistence for men and women in computer-related fields was examined. Persistence rates were similar for men and women in computer technology, electrical/computer engineering, and industrial engineering. However, fewer women than men persisted in computer science. Discriminant function analysis was conducted separately for men and women in each field and was equally effective for men and women in correctly classifying persisters and nonper-sisters (64–72%) in each field except computer technology where the classification rate was considerably lower for women (58%). GPA was generally the most important variable followed by a measure of math ability. High school science grades and number of semesters were often selected for the discriminant function for men, but not for women. The potential role of nonachievement factors in persistence such as peer and faculty support and expectancies are also discussed.
Jagacinski et al. (Sun,) studied this question.