(This is yes-and for your point, Doc.)
For those who are interested, ref. Stephen J. Gould’s The Mismeasure of Man, or any WAIS-IV IQ test applied to a neurodivergent person.
The executive summary is: the whole concept of an IQ is a statistical fiction, and the idea that it reflects some actual quality of “intelligence” which has been labelled g is seven kinds of nonsense resulting from getting it entirely backwards.
An IQ test is not a test, it’s a whole bunch of tests, testing different things. There are tests for linguistic memory, numeric memory, spatial orientation, pattern recognition, general knowledge, vocabulary, processing speed, comprehension, problem solving, all sorts of things. The trick then is that the results of all those tests are mathematically mapped onto a scale, where the results of the general population form a normal distribution bell curve, where the mean is 100 and standard deviation is 15. This means that you would expect roughly two thirds of the population to have an IQ between 85 and 115. That’s literally how it’s defined.
The thing is, each of these subtests is individually scaled, then the results all averaged up for the final number. The idea is that if you score at a certain point on one test, you are likely to score more or less the same for the others. That’s how it tends to look for Neurotypical people.
For a neurodivergent person, though, that goes out the window. I, for example, score very differently for spatial orientation and vocabulary than I do for working memory. At a certain point, if the subtest scores vary too far from each other, then the pattern is of far more significance than the average.
