Wednesday, February 10, 2010

IQ and Social Predictions

A new study reported in the New York Times today finds a link between IQ and Heart Disease.  The story then pretty much runs with a causation, not correlation interpretation.  Many commenters quickly fall in behind the narrative, assuming it confirms their own suspicions about IQ and success.  One assumes they are aware of the disparities in IQ scores among racial and socio-economic groups.  If one buys into the notion that these differences are caused by low IQ, and not the other way around, then this story certainly makes sense.

The pattern with stories like this is often the same: the reporter presents a skeleton of the study, minus the finer details of how it was performed (controls, variables, etc.).  The authors of the studies themselves are rarely interviewed, and generally in little depth.  The bigoted reader is then free to fit the story into their preconceived narrative, based as it is on false premises.  It is not political correctness which dismisses the bell curve model, but good old fashioned science. 

The idea that poor and minority people are genetically inferior is an excellent way of conveniently explaining away social inequities, thus liberating us from any sense of obligation to correct them.  The more difficult yet morally and intellectually honest approach begins with an examination of the ways in which privilege and disadvantage have worked against the American dream of freedom, and how they continue to today. 

The first casualty will likely one's sense that they are who they are today because of sheer self-determination.  Statistically, a few outliers succeed against the odds, but overwhelmingly it is factors such as one's parent's education level, income and family health (mental & physical) that are by far the most meaningful determinants.

The debate over what IQ scores measure and what they do not is complex.  For a good primer I encourage everyone to take a look at Stephanie Zvan's "Readings on IQ and Intelligence".

1 comment: