Date of Award

2017

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Computer Science

First Advisor

Edmund Lamagna

Abstract

Worldwide, beginning programming has a success rate of 67.7%, which may be a barrier to success for aspiring computer science majors. This may be particularly problematic for women, since only 18% of U.S. graduates with a bachelor degree in computer science are women. The purpose of this study is to identify conceptual predictors of success so those most likely to struggle can be identified. Specifically the focus is on the relationship of math to learning programming. Are math prerequisites helpful or a barrier to success? Can math achievement be used as a predictor of success, and can it be used to detect discrepancies between the success of men and women learning to program? The method for determining the best measure of math achievement is correlational comparison. The method for determining the effect of learning math is comparison of means. The method for determining predictors of success is linear regression.

The results are that the best measure of math achievement is the average of math grades from the year prior to the programming course. Either this average or the grade in the most recent math course can be used to predict whether a student is likely to achieve above a threshold of success. The math average can be used to detect discrepancies between the performance of men and women. For those who have taken a math course in the prior year, the predictor that is most significant is GPA at the time of taking the programming course. For those who start out in precalculus, those who do poorly do not improve their performance by retaking precalculus and doing better. Nor do students who take more math beyond precalculus see significant improvement. But for those who start out in calculus, those who struggle do see improvement when they succeed in learning calculus, and they do benefit from learning more math beyond calculus.

Share

COinS