There has been a focus on gender differences in education by the media who have claimed there is a ‘boy crisis’, this is the idea boys are underperforming in school compared with females. Our data shows that females outperformed their male counterparts in most samples supporting (Voyer and Voyer, 2014) who stated “A female advantage in school marks is a common finding in education research, and it extends to most course subjects”. However, it’s unclear from looking at the raw data weather the female advantage is consistent across all subjects and weather it remained this way throughout education. Here, we chose to look into these two queriers by analysing the data to determine if there is a significant inequality in scholastic performance between the genders.

Our hypotheses read, females outperform males in social science and females outperform males in science. If these are both true the difference will be greater in social science than science. The null states the difference won’t be greater in social science than science. This was based on the idea males do better in STEM subjects, with girls often being told this is a male field. Girls in the UK are found to have low self esteem when it comes to science (Adams, 2015). We analysed the datum to see if this affected their performance compared to other subjects. For our second querier we hypothesise; females will outperform males across the whole of education and their outperformance will decrease as they progress through education from elementary to undergraduate. The null states females outperformance of males won’t decrease as they progress through education. This hypothesis was formed to see if the female advantage is constant, and if when the two genders matured and progressed into more advanced education females would still outperform males.

Methods and results:

We received the data set from our tutor, which included 182 different results. The values we were interested in were: the subject, school type and female bias indicator (d). The female bias indicator is the standardised difference between two means, which in this datum indicates the difference in male and female test scores.

To analyse the data for our first hypothesis, we used the mean to determine if females outperform males in social science and science. Then to see if their is a difference in results between the two subjects we used Cohen’s d which produces an effect size indicating the standardised difference between two means. Therefore it will show there is a strong difference between the two sets of data, which in this case is Social science and Science. Then to test if the difference is statistically significant we used the statistical test, the t-test.

Using these methods of data analysis we found females outperformed males in both subjects as the means of the female bias indicator were 0.88 and 0.47 (Fig. 1). Cohen’s d found a value of 0.81, if the result is above 0.8 there is a large effect size and shows there is a difference between the two sets of data. This was supported by the T-test where a p-value of 0.01 was attained, this is less than the critical value at 5% 0.05, so the null hypothesis is rejected. Females outperform males in both social science and science but the difference is greater in social science.

The methods of analysis used for the second hypothesis were the mean and a One-Way Analysis of Variance. The mean was calculated for the female bias indicator of each education level to see if females did outperform males across education and to see if there was a visible difference between them when presented on a bar chart. We then used the ANOVA as a statistical test, due to it being able to determine if there are any statistically significant differences between more than two groups.

The results of the mean and bar chart, presented females to be outperforming males across all the educational levels (Fig. 2) with high school being the education level where females on average most outperform males, however, there were not large visible differences between the school types. This was supported by the ANOVA, the test produced a p-value of 0.22, this is larger than the critical value at 5% 0.05, therefore the null hypothesis is accepted. This result suggests that the performance difference between males and females across education is similar. Females have a consistent advantage across education levels.

Discussion

The data analysis shows we can accept the first hypothesis and reject the null. Females outperform males in both social science and science but the difference is greater in social science. This shows that the level of female advantage is not consistent across all subjects and means more research needs to be completed in this area to explain why this may be.

The second hypothesis was rejected and the null accepted. The difference in female and male academic performance across the education levels was not significantly different, suggesting that female advantage is consistent from elementary to undergraduate. This leads to further questions due women still facing inequality in the workplace in terms of pay even though they have an advantage academically.

However, limitations affect the validity of these results. The sample sizes for each of the results varied, for example, one was 97 and another 857. This affects the ability to accurately compare these results as they are not proportional. The data was from a variety of countries which may have impacted on the results, some people have argued a few minority groups of males e.g. Black or Hispanic skew the results as these students do proportionally worse than other ethnic groups and nationalities (Warner, 2006).

Conclusion:

Female achievement is consistent across educational levels from elementary to undergraduate, however, even though females outperform males in all subjects the degree to which differs depending on the subject.

References

Voyer, D., & Voyer, S. D. (2014). Gender differences in scholastic achievement: A meta-analysis. Psychological Bulletin, 140(4), 1174-1204.

Warner, J. (2006). Opinion | What Boy Crisis? online Nytimes.com. Available at: http://www.nytimes.com/2006/07/03/opinion/03warner.html Accessed 4 Nov. 2017.

Adams, R. (2015). Girls lack self-confidence in maths and science problems, study finds. online the Guardian. Available at: https://www.theguardian.com/education/2015/mar/05/girls-lack-self-confidence-maths-science-oecd-school-engineering