« Godwin's Law & Iraq | Main | A Tax Increase by Any Other Name »

February 22, 2003

Why IQ is a Moving Target

CalPundit brought up an issue that liberals and progressives do tend to avoid, which is how to discuss intelligence. Kevin wants to separate out useful scientific research on measuring intelligence from the racism of folks like Charles Murray and the Bell Curve misusing that science.

I actually tend to disbelieve there is such a singular thing as intelligence - g in Kevin's phrasing- but think it is rather made up of a constellation of skills and traits that each may have some inherited aspects, but the effectiveness of their combination in any one person is so affected by their environment as to make any focus on heredity almost irrelevant.

And I get that from one of the most striking results of intelligence research, the so-called Flynn Effect, named for researcher James Flynn. Here's the deal-- IQ increases each generation AMONG THE SAME POPULATION. A generation ago, all groups on average got lower IQ scores.

How can that be, you ask? You might have heard that scoring a 100 is average intelligence today and was always the average score.

But test-makers continually "restandardize" tests to LOWER the scores of people today. If you compared the raw scores on correct questions answered, people today would appear to have far higher intelligence than people in past generations.

This from Scientific American:

He discovered that certain IQ tests--specifically, the Stanford-Binet and Wechsler series--had new and old versions and that both were sometimes given to the same group of people. In the case of one of the Wechsler tests, for instance, the two versions had been given to the same set of children. The children did much better on the 1949 test than they did on the 1974 one. Everywhere Flynn looked, he noticed that groups performed much more intelligently on older tests. Americans had gained about 13.8 IQ points in 46 years, Flynn reported in 1984.
These results hold across countries. Look at this graph on the rise in raw IQ test scores over the last sixty years in different countries using the same test and not "restandardizing" the scores.

Conservative black columnist Thomas Sowell has highlighted the Flynn effect to note how "restandardization" created an illusion that blacks were not advancing:

Black Americans' test score results in 1995 would have given them an average IQ just over 100 in 1945. Only the repeated renorming of IQ tests upward created the illusion that blacks had made no progress, but were stuck at an IQ of 85.
I agree with both Kevin and Sowell that liberals should not be afraid of serious research on intelligence.

But the results, as James Flynn argues, is to make the whole idea of inate intelligence suspect as a scientific concept. If scores have increased so much in recent generations, any idea of some genetically stable "intelligence" looks relatively farfetched.

Some reading:
Full Scientific American story on Flynn.
From Intelligence Theory and Testing

See this additional post on Intelligence: Genes Cause Environment

Posted by Nathan at February 22, 2003 11:28 AM

Trackback Pings

TrackBack URL for this entry:


Post a comment

Remember Me?

(you may use HTML tags for style)