I see your 1994 paper and raise you a 2000 one by Raven himself: https://sci-hub.se/https://www.sciencedirect.com/science/article/abs/pii/S0010028599907351

"It would appear from the results summarized above that there has been, and still is, considerable—if far from perfect—similarity in the SPM norms obtained in different societies with a tradition of literacy at any given point in time. However, in common with the scores on other tests, and especially those measuring eductive ability through verbal or nonverbal items [see, for example, Bouvier (1969), Thorndike (1975, 1977), Garfinkel & Thorndike (1976), and the large number of published and unpublished studies brought together by Flynn (1984, 1987)], there has been a continuous increase in the scores at all levels of ability over time."

Now, it is true that I am look at this sideways, and looking at the differences in score over time, but i think it is fair to say that the culture of a place today is not the same as it was 50 years ago. In other words, those two time points represent different cultures.

We can clearly SEE either a bias between those cultures (assuming racial hetererogenity between the times) or some uncorrected for change in the tests, or something else. So either there is a cultural bias in the test, or some other factor is in play.

I think the ongoing revolution in qbit based computing should be paralleled by advances in the science of intelligence. All our old models have a leaning towards computational systems. Linear, maybe parallelised, processing. We've missed the whole side of intelligence that is ignored by western science. Intuition, Inspiration... the non computational solutions of problems that is now physically reproduced in quantum computing systems.

If we've missed that entirely in our IQ tests - AND WE HAVE (because we don't test for them), then it is likely that we are failing to measure differences in those processes across cultures.

I know that when I was working with raven's matricies, the tests stopped being valid, because I stopped thinking about them as I was solving them, and could just pick the right answers without working them out, as quickly as I could write the answers down. Much like most advanced mathematicians no longer consciously work through their complex equations in a linear fashion, but simply solve them.

So that's the place i'm coming from with my 'cultural differences', really. The idea that the tests miss something, and that failure is entirely due to the dominant cultural problem solving environment that created the tests. Its only applicable in certain circumstances.

I am explicitly NOT coming from this kind of place: https://www.sciencedirect.com/science/article/abs/pii/019188699290037P

And attempting to explain away racial IQ differences with undefined 'cultural bias' accusations.

So, sorry for any confusion. My interest is in the upper end of the intelligence range, not with the general population. I suppose i bothered to write what I did in the context of OP because... The american indian population had a very different approach to life, existing in a space with very different problem solving needs to the industrial west. They did very well, and so those as yet unmodelled systems of problem solving would not necessarily have been linked with the linear computational neural architecture that we test with RPMs I was just thinking about applying IQ tests to american Indians in their pre-contact state and so was examining the validity of standardised testing in that situation.

a bit like: https://www.apa.org/monitor/feb03/intelligence

Of course, these days you could just test an american indian who grew up in a western culture and get a great measure of that linear architecture.