Logic, What Logic?

This week I attended the Elizabeth Johnson lecture at the Institute of Physics, given by Mary Curnock Cook, chief executive of UCAS.  Her topic was Gender Maps in Higher Education, and it explored the differences between male and female progression through the various key stages at school and subsequent admission to university. There were a lot of statistics, fascinating comparisons of different parameters, and because I didn’t take copious notes I can’t reproduce them all. Her starting point was the BIS report which demonstrated a significant difference in the ‘graduate premium’ – crudely the difference between lifetime income between those with degrees and those without – for men and women. This premium amounts to the tune of £121K for men and £82K for women. I think many of us in the audience were surprised the graduate premium is as small as it apparently is, only £2-3k per year, but the difference by gender is stark. When she had read those figures she had started questioning what lay beneath. There is of course the possibility that it is nothing less than women on average being more likely to take some time out for families, perhaps working a few years part time, and so falling behind, but there may be other more worrying causes for concern. So she started delving into the wealth of statistics both publicly available and within the 50 years of information the UCCA/UCAS files hold on admissions.

There have frequently been stories in the news about girls doing better at most subjects at most exams throughout schooling, but I certainly hadn’t known that when it came to university admissions men were slightly more likely to be made an offer than women, for a given ‘tariff’ from A levels etc. At first sight that sounds rather shocking, but there may be a perfectly innocent reason: men are more likely to be doing A levels in the sciences, certainly in the physical sciences, and more likely to be applying for a STEM degree course where offers are on average slightly lower than for arts and humanities courses (though the data also shows that maths and physics A levels are ‘tougher’ than say English and psychology).  So, it is always difficult to say one is comparing like with like.  The Royal Society has done a lot of analysis of A level choices, based on data gathered for their fourth State of the Nation report  published earlier this year. The choices boys and girls make are strikingly different at A levels around the 3 sciences, with far more boys opting for physics (singly or in combination with other sciences) and biology on its own being most popular with girls. It could be argued that the reason boys get lower offers is that they are systematically making pragmatic choices to optimise their admission to university. This argument has been put forward in the context of the David Lammy attack on Oxbridge regarding Afro-Caribbean admissions, or lack thereof, namely that black children may be applying for the most competitive courses such as medicine, whereas private schools know that if their (possibly not-top-of-the-class) pupils want to get into these universities without a real vocational pull, they should be encouraged to apply for ‘softer’, less competitive options such as Classics. Of course if you genuinely want to be a doctor such a course of action would be fairly futile.

So, to return to Mary Curnock Cook’s talk, her final point was that for the first time  median pay differences in the current 22-29 age group actually show women earning slightly more than men, something also shown in the  recent CMI survey of junior executives. Perhaps the gender pay gap really will vanish in the years ahead as that cohort of young women, whose academic credentials at school now exceed the males, progress through the system.  She furnished us with much food for thought, and far too many statistics to be digested in one sitting. But nevertheless, as she was the first to admit, it could only be a superficial skating over of the analyses that potentially could be done with the rich data in principle available for consideration.

The questions that followed the talk made clear just how many further issues were potentially of interest to the audience, and what sort of questions need to be explored if we are really to understand not only the graduate premium and any gender differences, but also how schooling, socio-economic and ethnic background come into play. One question regarding why women weren’t yet showing up amongst many prize winners she in essence left to me to answer: the answer I gave is very simple but worth remembering. Awards committees can only award prizes to those nominated and people must remember to do this. It is too easy to ‘forget the women’, so this is a plea to all Heads of Department or other senior researchers reading this article to remember to do so. Nevertheless, the numbers of women receiving prizes is increasing (this year’s recently announced IOP prizes only seem to feature 2 women amongst the 16 prize-winners; the Royal Society is doing rather better in 2011 with 28% of its winners being women, even excluding the Rosalind Franklin Prize, up from 12% in 2010).

However, there was one question that left her – and I would guess most of the rest of the audience – fairly gobsmacked. Maybe collectively we misunderstood what the questioner implied when he asked something along the lines of ‘had there been any research to study whether women wrote less logically than men?’ It was a rather rambling question but the implication was that he thought women progressed less well because they were incapable of logic. Mary answered very diplomatically, and a representative from WES (if I recall correctly) pointed out research showed that exams, for instance, were marked more harshly if they had a woman’s name attached than if they had a man’s. But I doubt the questioner felt he had been answered, whereas the rest of us probably felt he didn’t deserve an answer.