I always thought the variations of continue series test (progressive matrices, number sequences, word A is to word B as word C is to ?? etc) are very culturally biased. You solve those best and easiest by sharing with the test maker the learning environment (and for visual ones, sharing visual environment), as well as sharing neural architecture. That lets you pick same choice as the test maker [edit: and do so easily and naturally]. And this research provides very good demonstration.
If IQ testing determined how culturally similar test-takers are to the test-makers, then you would expect that people the are the most similar culturally to score the highest. They don’t (for instance, East Asians score higher than White Americans). Therefore, IQ testing is probably not mostly about cultural similarity between test-makers and test-takers.
If IQ testing determined how culturally similar test-takers are to the test-makers, then you would expect that people the are the most similar culturally to score the highest. They don’t (for instance, East Asians score higher than White Americans). Therefore, IQ testing is probably not mostly about cultural similarity between test-makers and test-takers.