Looking back at this article (full text here), I notice something I didn't write about in my earlier (as yet unpublished) post about it --- one of the things its authors were testing for, looking to see if it had any relation to how much testosterone they found in their study participants' mothers' amniotic fluid, was IQ (measured using the Wechsler Abbreviated Scale of Intelligence, or WASI).
More importantly, the researchers also tracked their subjects' scores on one particular subscale of the Wechsler test --- Block Design --- which has been shown over and over again to be something autistic people are very, very good at.
Not only are we better, overall, than non-autistic people are at this task, we also score much higher on it than we do on any other subscale of a Wechsler IQ test.
Our scores, when you plot them across all the different subscales, tend to show this characteristic pattern, of a valley on the (verbal) Comprehension subscale and a very high peak on the Block Design subscale:
(Image adapted from Figure 1 in Dawson et al., 2007; each square along the line represents a different subscale of the Wechsler Intelligence Scale for Children - Third Edition)
Here is another image, from a much older study, showing the difference between the distribution across the different subscales of a Wechsler test of autistic and non-autistic test-takers' scores.(Figure 5, in Shah and Frith, 1993; both solid lines represent groups of autistic children --- the line connecting unfilled circles represents autistic children with high overall IQs, while the line connecting darkened circles represents autistic children with low overall IQs. Dotted lines represent various non-autistic groups: children with intellectual disability but not autism, children without disabilities, and teenagers without disabilities)
The peaked shape is present in both the low-overall-IQ and high-overall-IQ autistic groups, while both the low- and high-IQ controls' scores are more or less the same across all the subscales.
So, if more testosterone in the womb ---> autistic-like brains, you would expect 1) boys to score higher than girls on this task, 2) boys to show greater disparity between Block Design scores and other scores, 3) both absolute and relative Block Design scores to correlate with fetal testosterone within each sex.
As raw scores for each subscale of the WASI were not given in Auyeung et al., 2009, I can't tell anything about whether their subjects' Block Design scores were particularly high relative to the other subscales.
However, I can tell you that there were no sex differences on Block Design --- boys' average score was 19.14 to girls' 16.26, but the standard deviation for both sexes is probably somewhere between 9.5 and 10*, making the difference between the sexes a little less than one-third of a standard deviation, and thus not statistically significant.
Also, within each sex, amniotic testosterone had no relationship to any of the IQ measures the researchers reported --- not full-scale IQ, not verbal IQ, not performance IQ and not the Block Design subscale of performance IQ.
So, this cognitive ability characteristic of autism doesn't seem to vary with sex or with androgenicity.
*Standard deviations are only given for boys and girls separately; for boys, it is 9.41; for girls, 10.08. There are 43 boys and 31 girls in this study population.
Auyeung, B., Baron-Cohen, S., Ashwin, E., Knickmeyer, R., Taylor, K., & Hackett, G. (2009). Fetal testosterone and autistic traits British Journal of Psychology, 100 (1), 1-22 DOI: 10.1348/000712608X311731
Dawson, M., Soulieres, I., Ann Gernsbacher, M., & Mottron, L. (2007). The Level and Nature of Autistic Intelligence Psychological Science, 18 (8), 657-662 DOI: 10.1111/j.1467-9280.2007.01954.x
Shah, A., & Frith, U. (1993). Why Do Autistic Individuals Show Superior Performance on the Block Design Task? Journal of Child Psychology and Psychiatry, 34 (8), 1351-1364 DOI: 10.1111/j.1469-7610.1993.tb02095.x