An Investigation into the Effects of Gender on Performance in Computer Related Examinations
This report seeks to examine the hypothesis that “examination result and gender are independent attributes.” In order to investigate the statement, several statistical tests are applied to a large pool of
data. The vast majority of this originates from the records of Leeds University School of Computing.
However, the examination results from other centres are assessed for comparison. A dual approach is used for the analysis section of the project. This entails the use of traditional hand calculation methods as well as computer-based techniques. The main hypothesis quoted above is proved. In other words, it has been demonstrated that the observed very weak link shown between examination marks and gender, illustrated by the cumulative frequency curves and boxplots, is not statistically significant. Although the analysis shows a small trend in favour of males, this is almost insignificant. There is also a trend that suggests a narrowing of the gap between male and female over the last three years. The Chi-squared and t-tests carried out have revealed remarkably consistent figures for both the testing using hand calculations and the computer aided testing using SPSS and MINITAB. They indicate that if there is superiority in the case of males, it is only very marginal. This would appear to be the situation across both the university and A level results. The data for individual modules tends to suggest however that there are slight variations in performance across the range of subjects within a particular course. In essence, the study demonstrates that, in spite of the prevailing low level of participation, women are equally capable of handling the challenges offered by computer science.
Background research, which is also a feature of this project, tends to suggest that if there is a difference between the genders, it could perhaps be explained by other factors other than pure ability, such as opportunity, motivation or confidence.