Sir, – Your report on the impact of perceptions on the mathematical performance of girls ("Girls' maths ability underestimated due to stereotypes, study finds", News, October 9th) would indicate a serious issue with the effect of gender bias in the education of our young people.
Fortuitously, in this instance we have a real-world “experiment” that serves to stress test to an extent some of the findings you have detailed. That experiment is the recent exercise in Leaving Cert predicted grades, which reflects a considerable sample of teacher assessments for some 60,000 students. Your report notes that the outcomes of that exercise “transcend’ the findings of the study”. It would appear they do not so much transcend, as fundamentally contradict elements of the study!
The report from the National Standardisation Group to the Independent Steering Committee and the Programme Board provides an insight into how teachers view the performance and capability of their students. The report details this on a gender basis for the three core subjects, mathematics, Irish and English . As noted in that report, research points to the existence of student gender effects in teacher estimates of student performance in a range of contexts. For the Irish context, in higher-level mathematics, males historically tend to do better on average than females; however, the teacher-based estimates are noted to reverse this trend.
For higher-level mathematics at grade H1, we see that teachers estimated that 10.4 per cent of girls and 12.9 per cent of boys should receive these. This is relative to an average of 3.6 per cent and 8.1 per cent in the previous three years for these groups, giving a grade uplift at this level of 289 per cent for girls and 159 per cent for boys based on teacher perception. When the model is applied this drops somewhat to a, still remarkable 219 per cent for girls and 110 per cent for boys.
Based on this it is difficult to conclude that teachers are disposed unfavourably to their female students in terms of rating their mathematical ability, at least in the context of the Leaving Certificate outcome.
Indeed the more appropriate hypothesis would appear to be that their male students are adversely affected in this manner, with girls receiving a considerable uplift over boys when their teacher outcome is taken in favour of the historical trends.
It might be speculated that this is simply a case of redressing an underlying inequality in favour of the traditionally weaker group. However, if this were the case, we would expect similar trends in the reported results for English and Irish, where boys historically underperform girls. However, this is not what happened. For English, the teacher estimated H1 grade percentage for girls shows a 236 per cent grade inflation versus the three-year average, while boys saw “only” a 184 per cent adjustment. The situation in Irish is a little better, with relatively even H1 grade inflation, versus the three-year average, of 227 per cent versus 241 per cent.
Given this it is difficult to reconcile the real-world outcomes for this cohort of 60,000 students with the argument put forward in your news report. Indeed it would be desirable to see further research that considered the outcomes of the predicted grades approach, more fully considering the possibility of bias against key segments.
Bias. whether on the basis of gender or other factors, should be understood and addressed, regardless of whom that bias serves to impact. – Yours, etc,
ALBERT WINSTON,
Glasnevin,
Dublin 11.