[waveman2]

The studies seem to be confusing causation and correlation. They seem to assuming that there must be an environmental cause for the differences in people's IQs. This may seem like a stupid assumption to make, but a lot of people in the social sciences take this as a given, even though the weight of evidence is strongly against it.

Perhaps I have missed something, but I would expect that, if they did not have ideological blinkers on they would be saying things like:

"We are taking particular care to separate environmental and genetic factors here by including adopted children, [identical] twins reared separately, other children raised by other than biological relatives".

But I see none of this at all here.

You may be interested to read "The Nurture Assumption" for a detailed analysis of the effect of parents on their children. The basic conclusion is that if you do not seriously abuse your children you don't affect them much one way or the other. Steve Pinker's book "The Blank Slate" is also relevant.

[bayesianhorse]

Today everyone can and usually does criticize "cause vs effect" whenever a study doesn't fit into his world view.

The discussion usually is omitted in journalistic articles anyway, so it's easy to say for readers that the scientists just "have no clue".

For me it looks like this research is looking for interventions. They are not just trying to understand intellectual development, but what can be done to improve early education. They succeeded at that. They didn't succeed at satisfying romantic and simplistic understandings about a genetic basis for intelligence.

[ronaldx]

> twins reared separately

This is the gold-standard for testing genetic association, yes, but it is increasingly considered unethical. And is not perfect - twins are still likely to be placed in similar homes (typically both in foster homes, after all), which likely results in an overestimate of the genetic effect.

Modern statistics is about handling the data that is available: present-day researchers cannot rely on a statistically perfect study which would have to be generational and eugenic in scope.