Sorry, was being vague about my use of the word "environmental". It would be interesting to find twins with differing IQs whose difference was attributable to educational environment, not random environmental physical causes. If a child gets brain cancer or falls and suffers brain damage, I guess it's still environmental but those situations wouldn't help. I guess one would have to reasonably rule out any important in-womb effects, too, for the results to be very meaningful.
Since even under the most pro-genetics studies the IQ correlation is only as high as it is, I'd expect a fair number of instances of substantial differences where major physical damage or some growth aberration wasn't the cause. Out of those people, there are probably some from really subpar educational environments or who suffered during development from poor nutrition-- and though it might be hard to pick those people out exactly, excluding on the basis of low parental income might be a decent rough substitute as such studies seem to often include such info, especially if we want to err on the side of caution. I'd personally be surprised if there were no one left at all after all this culling, but of course the existing studies are unlikely to contain much useful info for figuring out the causes.
I think one-point measured IQ differences simply don't matter, and I don't think that's what Cricket2 was asking about (passing a numeric threshold) but rather a more substantial change. This is the same thing that always bugs me about firm adherence to the Ruf Levels, definitions of HG/EG/PG tied to specific IQ measurement thresholds, etc. I think we understand each other well enough at this point not to quibble about that, though.
I've never read "Outliers", but it's on a list saved somewhere. Will have to follow up soon.
