Rdos blog

Similiarily as the physical traits, it should be possible to discriminate Aspies from neurotypicals based on brain anatomy. There is a recent study on this subject here: http://www.jneurosci.org/cgi/content/full/30/32/10612.

The popular press makes it sound like psychiatry has finally found the brain defect that will enable us to diagnose ASDs on physical  traits, and in the end identify the defect and dispose of it. This is hardly what the study found.

The study found that if 20 people with ASDs and 20 neurotypicals where mass-surveyed for every possible anatomical difference, they were able to place them in the correct category with very high confidence.

The method used looks an awfully lot like the factor-analysis method that Aspie-quiz used to obtain weight factors for being Aspie and neurotypical, and which made it possible to classify people without a maunal scoring algorithm. The researchers have basically replicated the Aspie-quiz method, albeit with a data-set of 40 where Aspie-quiz used 150,000.

The results seems similar as well. They found that no single difference can provide good discrimination between groups, but instead there is a need to use many different meassures. These were related to volume and  geometry, and existed in many different areas of the brain  and in both hemispheres  (even if one were more relevant than the other). This is what Aspie-quiz found as well. There was a need for about 150 diverse personality-traits in order to make good judgements for people’s Aspie-NT profile.

The study also found that ADHD was less well defined, which is expected when ADHD is mostly related to a single dimension in the Aspie profile (Aspie social).

This kind of distribution of many different brain differences being related to ASCs is what the Neanderthal theory expects us to see. The authors provide no real theory for how their findings could be explained, but the Neanderthal theory explains it very well.

It is necesary to point out that the small sample size and large amount of differences tested makes this a data-mining project with many possible problems. In the worse scenario, a new sample of 20+20 individuals might fail totally in placing participants in the correct groups if the differences found are only particular to the subjects used.