Alison Wolf

Male-female differences will influence success in certain degrees. We should look deeper than statistics to examine bias

Back in the late 1980s, I nearly fell out permanently with a close friend and co-author over "sociobiology", the study of the biological basis of social behaviour. She found the idea of genetically based differences between male and female behaviour profoundly offensive: anyone who wrote about these ideas favourably was colluding in absolving society from all responsibility for injustice. I accused her of wanting to practise censorship, as well as ignoring fascinating and important research with profound implications. We calmed down and made up, and today such ideas are mainstream. But I nonetheless find myself surprised, 15 years on, by just how important gender is in our lives, and by the scale of male-female differences throughout education.

Women now form a substantial majority of university students in this country: 58 per cent of all UK-domiciled undergraduates and 55 per cent of postgraduates. This is typical of the developed world. A glance at the latest figures from the Organisation for Economic Cooperation and Development shows men to be still a small majority in eastern Europe, Korea and Turkey - but elsewhere women consistently make up more than half, and often close to 60 per cent, of the university student body.

Women perform better, too, on average. At university, this manifests itself more at the bottom than at the top of the attainment range: averaged out nationally, British women get almost exactly the same percentage of firsts as men, but only half the percentage of thirds and pass degrees. At school level, the differences are more striking. Girls have been outperforming boys, in terms of average numbers and grades of GCSE passes, for a good many years, but are now doing the same at A level (and not just in arts subjects). And with the growth of testing, we are getting comparable gaps reported at ages seven, 11 and 14.

A fascinating new book by Simon Baron-Cohen makes this more, not less, intriguing. The Essential Difference sets out the accumulating evidence for certain specific and significant differences between the (average) male and the (average) female brain. Males, Baron-Cohen argues, tend to be better at, and more interested in, systematising: that is, in understanding the rules that link inputs, operations and outputs. This shows up in laboratory tests, and it is also why small boys are far more likely than small girls to collect football cards, or to know everything there is to be known about dinosaurs, characters in war games or particular types of band. Girls, from early babyhood, are better at empathising with others: meaning that they can tell how other people feel about things, care about this, and pour their energies into complicated best-friend and gossipy relationships.

The evidence is all very convincing. But why, in that case, are girls doing so well academically? You would have thought that academic study was quintessentially a systematising pursuit: in its quest for the general and abstract, in the rigid mark schemes that decide success in school examinations, in the love affair of all university subjects with the scientific model. So what else is going on?

Differences between male and female brains also raise questions about the meaning of "bias". University degrees vary enormously in their demands and objectives, even within a subject. You can pick two maths degrees, never mind two French, politics or business ones, and find virtually no content overlap at all. Isn't it likely that, in the course of developing such diversity, we will also create courses that are systematically more suited to either males or females? Provided that there are clear objective rules and procedures for selection, marking and grading, should we bother about whether men's and women's results are identically distributed? Perhaps we should instead expect that they frequently won't be.

In recent years, Cambridge University historians have been the British academics most visibly occupied with gender bias. The trigger was the much lower proportion of firsts going to women. Resulting inquiries have certainly clarified what makes for academic success in this particular degree course. If, as seems to be the case, many faculty members feel this distorts the qualities that the course is meant to develop, then there is indeed an issue. But equally, male-female differences make it statistically almost certain that some degrees, somewhere, will produce stark differences in outcomes. Simple statistics are a poor way to examine bias. But then that is true for many of higher education's central concerns - which hasn't stopped statistical monitoring growing even faster than women's academic prowess.

Alison Wolf is professor of education at the Institute of Education, London.