Unwarranted Faith in the IQ Test

In this well-constructed review of those who have tried to link IQ to racial or ethnic differences, David Weigel, writing on Slate, details the downfall of Heritage Foundation statistician and scholar Jason Richwine. Dr. Richwine had written an analysis of the predicted outcomes of immigration reform, suggesting that the reform would be a drag on our economy. This analysis was met with little controversy (though plenty of criticism); instead, it was Dr. Richwine’s past that got him in trouble. Specifically, Dr. Richwine wrote a thesis while at Harvard entitled “IQ and Immigration Policy” that suggested Hispanic immigrants should be denied entry to the United States more frequently than other groups because Hispanic people had lower average IQs. (In the following paragraphs, I will talk about Dr. Richwine but I am referring specially to his past argument. I have no information about his current beliefs and if he has changed his position on IQ and racial and ethnic differences.)

What’s the problem here? First and foremost, there is political backlash. These criticisms come from both liberals and conservatives. The nature of the criticism may be different. For example, a liberal might suggest that making the claim is inherently racist because racial differences are not biological but rather are socially constructed. A conservative might argue that IQ is not a good measure of economic productivity and therefore is not a reasonable basis for immigration policy decisions. But both these arguments stem from a desire to condemn politically-controversial statements.

In this case, however, politics is best put aside. Instead, it is more fruitful to pay attention to the IQ test itself and then to probe deeper and question our own often blind faith in testing as a means of assessing individuals. In short, the IQ test falls short not for a political reason but instead because it is a flawed instrument.

The IQ test is (and must be) structured to fit various groups. There are IQ tests for children that contain different questions and question-types than the IQ tests for adults. There are IQ tests for people with high reading ability and for people who are illiterate. There are IQ tests that are translated and culturally adjusted so that IQ can be compared between countries. This variance in the testing instrument suggests that individuals may perform better or worse depending on which test they take. These differences may explain any incidental differences between different groups. And this is the best explanation for any difference between racial or ethnic groups.

Beyond this, what additional explanation for variance is left? That is, what can Dr. Richwine offer by way of theoretical mechanism for IQ differences between racial and ethnic groups? He is left with just one explanation: genetic differences that produce a predisposition toward intelligence. Here again, the political backlash to such a claim is powerful, but Dr. Richwine is proposing a testable hypothesis. Unfortunately for him, the evidence comes back in complete and utter opposition. There are no identifiable genetic differences between racial and ethnic groups related to intelligence. In short, Dr. Richwine has made an observation (IQ differences) but has failed entirely to explain that difference. If he had successfully explained it, then he would know that ethnic or racial group differences are mere surface level indicators of flaws in the testing process and an inability to make a test that is appropriately catered to all. He might also find that the measure itself has been heavily criticized and is hardly a value that remains stable across life, as it is intended to do. (This too has to do with testing issues – how can you make a test that produces the same score for the same individual at five years old and at 55?)

The most reasonable conclusion in this case is that IQ itself is a flawed measure. But it is treated by Dr. Richwine as a perfect indicator of a constellation of factors: intelligence, certainly, but also ability, dedication, work ethic, and self-reliance. This flawed thinking doesn’t render Dr. Richwine’s conclusions wrong; instead, it makes them moot. The conclusions are akin to saying a student’s low SAT score should keep him from riding a roller coaster. The purpose of the IQ test is so far removed from the immigration debate that its use in such a debate violates all foundational principles of the idea of testing itself. Asking the question about IQ and racial or ethnic differneces isn’t the biggest problem here; instead, Dr. Richwine’s biggest failing is a misuse and shallow analysis of data.

Leave a Reply

Your email address will not be published. Required fields are marked *


9 − = three

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>