Pink and Blue…
Or, Yes, We Now CAN Tell The Sex of a Person By Imaging Their Brain
Before recent developments in neuroimaging, I would have said that there was no way that we could determine the sex / gender of a person looking only at their brain. In fact, I DID say exactly that. But now, I don’t believe that that is an accurate statement, at least not wholely accurate, because a recent paper/letter has shown that with increased resolution and computer power we can determine the sex of a person that a particular brain resides in just from an analysis of the 3D image of their brains to 93% accuracy. The mere fact that this can be done shows that the human brain is in fact highly sexually dimorphic, because if we were to simply guess, we would only be right 50% of the time.
The number, 93%, sounded suspiciously familiar to me. That’s about the number of people who are not LGBT in the population. Given that we also know that LGB people are likely to have sexually dimorphic features that as a population, are shifted towards that of the opposite sex, I’m proposing an hypothesis and a prediction. If this analysis were redone excluding all known LGBT people, that the mathematical regression would result in greater predictive strength. It would not reach 100% because there would still be those who due to social desirability bias would fail to disclose their sexual orientation and thus still be included in the heterosexual study group. Increasing the accuracy in that instance will add evidence to a quip that I have made before, that humans don’t have male and female brains so much as androphilic and gynephilic brains.
There’s an important point that is missed by people with discussing the issue of whether the human brain is sexually dimorphic or not; The size and shape of any specific feature of the brain is to an extent only a very crude estimate of the number of neurons and the connection density of that region. It does not tell us the functional differences, if any, that that difference represents. As Cordelia Fine has pointed out, these differences, though they clearly exist, does not tell us what, if any, the differences may be in men’s and women’s minds. Only additional research will help us determine these.
But still, anyone who still says that it is not possible to determine the sex of a human brain hasn’t been keeping up with the science.
(Addendum 3/4/2017: I did a bit of calculation and found much to my amazement, that to “guess” the sex of the brain to 93% accuracy means that the effect size (Cohen’s d) would, if it were a single dimorphic feature, be a whopping 3.0 !!!! That’s an over the top value. Thus, as we get better imaging tools to see the fine details, we are learning that the human brain, in terms of multivariate statistics of multiple measurements at all points of the brain, is in fact extremely sexually dimorphic. The problem is that no one area is all that dimorphic, but in aggregate, they are quite dimorphic. That is to say, if one area is slightly dimorphic, giving a small statistical clue as to the sex of the individual, and a second area is also slightly dimorphic, giving a small clue as the sex of the individual, the two can be used together to give a medium sized clue to increase the accuracy… and with many many areas, each additively pointing towards one sex or the other, the accuracy gets quite good.)
Chekrouda, et al., “Patterns in the human brain mosaic discriminate
males from females” http://www.pnas.org/content/113/14/E1968.full.pdf