Schizophrenia and autism (as diagnosed on subjective criteria by professionals) are known to be polygenic and genetically heterogeneous, so that is a great hurtle to begin with. To try to categorize subjects on the basis of a photo or video would also be very subjective and uncontrolled. Instead of having a single assessor, you could have a panel of "experts", but then how would you deal with disagreements between the panel members?
The best way to do a study would be with objective measurements. Let's say that you had a candidate gene for N (non-sanguinity as a dominant allele). You could hypothesize that sanguine individuals would flush readily in response to a vasodilator -- say a measured amount of ingested alcohol (ml/kg body weight) -- as compared to non-sanguine individuals. You would devise a test in which the flushing is measured photometrically. You could eventually have two groups: the "extreme flushers" and the "extreme non-flushers", and you would then measure the prevalence of your candidate allele in each group. The study would be conducted by a statistician and designed to minimize bias.
« Back to index