The ideal way to do heritability studies is by looking at adopted twins. Basically, what they have found is that if you take a Japanese kid whose parents had an IQ of 107, and put them in a white family whose parents had an IQ of 100, the best guess of his adult IQ will be 107, not 100. (family conditions generally have zero predictive power on IQ - the non-heritable aspects of IQ seem to us to be 'random' [most likely developmental events in the womb, or singular, unique events in environment that are individualized and don't effect multiple individuals])

The results are replicable throughout decades and hated by 99% of social 'scientists'.

It is true that you can train for IQ tests, though the extent to which will differ (everyone can get on Mensa is likely untrue). The ideal is to be taking a test that has not been specifically trained for, which at a population level, is generally the case, but can make the individual score unreliable. That IQ is generally measuring something real and something useful is indicated by its high correlation with life outcomes AND brain size.

The ideal way to do heritability studies is by looking at adopted twins. Basically, what they have found is that if you take a Japanese kid whose parents had an IQ of 107, and put them in a white family whose parents had an IQ of 100, the best guess of his adult IQ will be 107, not 100. (family conditions generally have zero predictive power on IQ - the non-heritable aspects of IQ seem to us to be 'random' [most likely developmental events in the womb, or singular, unique events in environment that are individualized and don't effect multiple individuals])

The results are replicable throughout decades and hated by 99% of social 'scientists'.

It is true that you can train for IQ tests, though the extent to which (everyone can get on Mensa) is likely untrue. The ideal is to be taking a test that has not been specifically trained for, which at a population level, is generally the case, but can make the individual score unreliable. That IQ is generally measuring something real and something useful is indicated by its high correlation with life outcomes AND brain size.