Abstract
In this study, the assumption that good performance in mathematics in the final school year could be used as a pre-entry requirement to programming courses at universities in South Africa, is challenged. The extant literature reports positive relationships between mathematics performance and success in programming courses. As computer programming modules in higher education institutions (HEIs) are typically characterised by low success rates, it becomes important to eliminate potentially erroneous entry requirements. The low success rate in programming modules is ascribed to the abstract nature and content of programming courses, and the inadequacy of pre-university education to prepare students for the cognitive skills required for success in such programmes. This paper reports on a single independent variable, ‘performance in high school mathematics’, and its relationship to performance in two computer programming courses. The dataset comprised the school marks of four cohorts of students who were enrolled for the programming modules between 2012 and 2015. Firstly, we computed the point-biserial correlation between a dichotomous variable that indicated whether students had mathematics as a subject in Grade 12 or not, and their performance in the programming modules. Once we established that a relationship existed, the marks achieved in the final school year for mathematics, and performance in two programming modules were correlated. Results indicated that the school mathematics marks correlate only marginally, and that correlations were not significant, with performance in the two programming courses. We also correlated the school mathematical literacy marks with performance in the two programming courses, and found that a strong positive correlation that was significant existed with the second semester programming course. We conclude that the mark achieved for school mathematics cannot be considered as a valid admission criterion for programming courses in the South African context.