Azara Blog: Sex Differences in Mind

Blog home page | Blog archive

Google   Bookmark and Share

Date published: 2005/01/21

The first lecture of the Darwin Lecture Series 2005 was by Simon Baron-Cohen, about genetic differences between men and women. This is a notoriously thorny issue, as witness the recent brouhaha created by Larry Summers, the president of Harvard.

The main problem with the Darwin lectures is that they are broad but not deep, so you know that you are only getting a superficial glance at what is always going to be a complex story.

Baron-Cohen's main work is on autism. This condition afflicts males much more than females (he quoted 4 to 1 for autism in general, and 9 to 1 for the more narrowly defined Asperger's Syndrome). So comparisons were made between males and females in general and then people (males only?) with autism.

There was introduced the concept of the "empathy quotient" (EQ) (which is supposed to be higher for women because they are allegedly more people-oriented) and the "systemizing quotient" (SQ) (which is supposed to be higher for men because they are allegedly more system-oriented). Needless to say, defining precisely what these mean and measuring them without allowing the influence of "nurture" (rather than "nature") to creep in, are the main problems. As soon as you are more than N months old (for some small N) your parents and society are already treating you differently depending on whether you are male or female, and removing this effect is difficult.

He mentioned several tests. One was the "eye matching" test (which is supposed to be related to your EQ), where you are shown a small slice of a human face with only the eyes showing, and are given four words that describe emotional states (e.g. "depressed") and you must guess which represents the "true" emotion of the person. The results were as follows (maximum score 25). (And here and below the autism category might be only the more narrowly defined Asperger's Syndrome.) (And the sample size was not stated.)

average stdev
male 19.5 2.6
female 22.1 2.0
autistic 16.6 2.9

So women are better at recognising the emotion of someone and autistic people (men?) are worse than men in general. It was claimed all these differences were significant (in a statistical sense). It is interesting that the difference between men and women is approximately the same as one standard deviation between men by themselves and between women by themselves. Thus although women score higher than men on average there is a reasonable number of men who score higher than a reasonable number of women.

Similarly there was the "finding the targets" test (which is supposed to be related to your SQ), where you were given a geometric outline and had to find it in a somewhat elaborate design. Here the scores were (in seconds for a certain number of examples?):

average stdev
male 46.2 20.5
female 66.7 36.7
autistic 32.2 27.0

with similar conclusions to the "eye matching" test except that here the men do better and the autistic people best of all.

He also mentioned other another EQ test where you answer questions like "I really enjoy caring for people" (women score higher) and a similar SQ test with questions like "When I listen to a piece of music I always notice the way it's structured" (men score higher). Both of these can have bias, for example one question he mentioned for the SQ test was about feeling confident about doing electrical wiring work in the home, and obviously if you are more experienced with that kind of thing (which quite likely today more men are) then you would score higher.

Then he mentioned some tests on children, including a "Faux Pas" test (reading a pretend conversation and figuring out if a social "faux pas" was made) and on children less than age five girls scored higher. Of course by age five kids have already been heavily propagandized.

So he mentioned a test on babies age 24 hours. Definitely no social bias there. They filmed babies for ten minutes with (a picture of?) a mobile overhead and a picture of a face, and counted the number of times the babies looked at each. The following are the percentages of boys and girls who looked more at the mobile than the face, more at the face than the mobile, and (approximately?) equal both objects.

mobile face equal
boys 43 25 32
girls 17 36 47

So even at this age there appears to be an EQ/SQ difference between the sexes.

He then said that they were looking at pre-natal measurements, and that they had found a (statistically significant) correlation between foetal testosterone levels and later measures of EQ and SQ at ages 12 to 48 months. Of course male fetuses on average have more testosterone than female fetuses. It's unlikely the story is just as simple as that, though.

All material not included from other sources is copyright For further information or questions email: info [at] cambridge2000 [dot] com (replace "[at]" with "@" and "[dot]" with ".").