I don't think you can control for genetics, at least not directly. There is no intelligence genetic marker as far as I know which only leaves testing the IQ/intelligence of the whole family tree as the only option. However getting access to ancestors might not be feasible. On the other hand having a large pool of people is good enough since the distribution of intelligence is usually uniform and not restricted to things like class or parental income.
Here's one way to control for genetics - compare orphans who were adopted early (Pre 1yo) and then see what the correlation between their IQ and both the parents and adoptive parents IQ. (Or use income as a proxy for intelligence; it might be hard to get the dead parents to take an IQ test)
> I don't think you can control for genetics, at least not directly.
That's my point, really. My conclusion is that these studies are interesting, but the body of evidence isn't scientifically rigorous enough that it should affect policy decisions.
