From Catherine Scott:

Re: Gender math gap erasable, studies suggest (May 30): The article on the hypothetical differences between human males and females in maths attainment is great example of the lengths to which even those who style themselves as followers of the scientific method will go to preserve their core beliefs. The existence of fundamental cognitive differences between men and women is one such core belief because it cuts straight to our identity. First and foremost we think of ourselves as men or women and anything that proposes that these categories are built on shifting sand is anxiety-provoking, so anxiety provoking that we’ll even clutch at evidence of sex differences in animals not even in the phylum as us as ‘proof’ of the naturalness and inescapability of apparent gender differences. (What about spatial differences in our nearer relatives, primates? I bet they wouldn’t exist, otherwise female monkeys etc would routinely fall out of trees and drive the species to extinction. ) That male cuttlefish are thought to have more in common with men that female humans shows the extent that gender is a fundamental and essential COGNITIVE category for male humans, (if not cuttle fish :-) ). And it shows the lengths to which people will go to amplify any small differences between men and women they can find.

Let’s look at the evidence as if it were experimental in origin and see what conclusions we reach: In a population of interest there are two naturally occurring groups, A and B. In one experimental condition, the two groups are subjected to the same intervention, X. In the second condition Y the two groups are subjected to two different interventions, such that group A experiences intervention Y and group B experiences intervention Z (not a good experimental design). Measurements on an outcome variable reveal that the two groups that experienced intervention X did not differ, however, the groups that experienced interventions Y and Z did differ, Now, on the basis of this what is the logical source of the measured differences? It has to be the interventions, NOT a pre-existing between group difference. Like it or not the only way to prove this not to be correct is to raise some girls like boys and some boys like girls. Not going to happen, so we have to rely on naturally occurring ‘experiments’. When girl do boy things that they perform like boys, as in the computer game example.

Boys are regarded as naturally less able at reading than girls but when reading is taught well (which it mostly isn’t in Anglophone cultures) this difference also disappears. Presumably the reading differences didn’t excite your correspondent’s interest because in our culture maths ability is regarded as the true measure of ‘braininess’ (which is why it s important to find ways to defend men’s ‘natural’ superiority’ at it).