Leaving Cert calculated grades open to danger of gender bias

Non-anonymous marking may expose department to legal action for prejudice

The Department of Education and Skills assumes that teachers will act professionally when estimating the percentage mark that their students might have received in the 2020 Leaving Cert, and in the ranking of their students relative to one another.

They have been asked to take into account records of their students’ performance including assignments, projects, mock exams (“with caveats”), previous results and any other relevant information. Individual teachers’ calculated scores are to be aligned with others teaching the same subject, with oversight by the school principal.

When the academic experts did not know whether the applicant was a boy or a girl, the proportion of girls who were successful increased

This process involves the abandonment of the Leaving Cert’s anonymous marking system and teachers have been asked to be alert to possible bias that might stem from students’ classroom behaviour, or their knowledge about a student’s socio-economic or family background.

There is no reference to gender bias – as reflected in the differential evaluation of boys and girls work – in the department’s guidelines but international evidence suggests it is highly probable this will occur.

In an experimental study involving two identical CVs, one with a boy’s name and the other with a girl’s name, both men and women assessed the one with a boy’s name more positively than the identical one with a girl’s name.

The Irish Research Council found that obscuring the gender of the applicant made a difference to success rates for girls in applications for post-doctoral funding in science, technology, engineering and mathematics. These applications were all assessed by an expert panels of academics.

Distorted judgment

In 2013 when the gender of the applicant was obvious to the assessors, girls made up 43 per cent of the total number of applicants but only 35 per cent of those which were successful. In 2017, after the gender of the applicant was obscured, the proportion of girls who got awards increased: rising from 35 per cent in 2013 to 57 per cent in 2017.

This suggests that the judgment of the experts was influenced by the gender of the applicant – and that the girls’ applications were judged more harshly than those of the boys. When the academic experts did not know whether the applicant was a boy or a girl, the proportion of girls who were successful increased.

In 2014, the Equality Tribunal ruled that Micheline Sheehy Skeffington had been discriminated against by NUI Galway in the 2008 promotion competition on the basis of her gender. NUI Galway has since promoted and compensated five other women lecturers who argued they were discriminated against on gender grounds.

Sweden is one of the most highly ranked countries in terms of gender equality. Observers who were put on research funding boards there noticed that doubts about applicants’ intellectual autonomy were much more likely to be raised in the case of applications from women than from men. This had consequences – with the men being more likely to be successful than the women.

There is evidence from another Swedish study that the least productive male applicants were judged to be as competent as the most productive female applicants. Women had to be 2.5 times more productive than the average man to get the same competence score. These assessments were made by professionals.

Devaluation of women

The devaluation of women is not peculiar to academia. The New York Philharmonic Orchestra recognised, following legal action, that women were not being hired to play because of a bias against them. They started to hold auditions with musicians playing behind a screen – and in their bare feet so that their tread did not reveal their gender. This had an immediate effect on the proportion of women hired to play in the orchestra.

It is not clear what steps will be taken by the department to eliminate gender bias  likely to arise in calculated grades

The Department of Education and Skills will use a standardisation process involving historical data to ensure that the calculated grades reflect a common national standard and the actual marks allocated by teachers will be adjusted to bring them in line with the expected distribution of grades for the school.

It is not clear what steps will be taken by the department to eliminate the gender bias which is likely to arise in the case of calculated grades: a bias which was eliminated by the Leaving Cert process of anonymous marking.

It is also not clear how the rank order of students can be adjusted in the case of mixed schools if the boys are systematically ranked more highly than the girls.

At the very least, the department’s statistical model needs to bench mark the performance of boys and girls in each school relative to each other in 2020 against previous years; and to ensure that the gender profile in each subject nationally reflects the previous years’ gender profile.

If this does not happen, then it is very possible that the parents of daughters could successfully take legal action against the department.