Are We Getting Smarter? Rising IQ Scores In the Twenty-First Century

hessian sends this excerpt from The New Republic: "[A] person who scored 100 a century ago would score 70 today; a person who tested as average a century ago would today be declared mentally retarded. This bizarre finding — christened the 'Flynn effect' by Richard Herrnstein and Charles Murray in The Bell Curve — has since snowballed so much supporting evidence that in 2007 Malcolm Gladwell declared in The New Yorker that 'the Flynn effect has moved from theory to fact.' But researchers still cannot agree on why scores are going up. Are we are simply getting better at taking tests? Are the tests themselves a poor measure of intelligence? Or do rising IQ scores really mean we are getting smarter? In spite of his new book's title, Flynn does not suggest a simple yes or no to this last question. It turns out that the greatest gains have taken place in subtests that measure abstract reasoning and pattern recognition, while subtests that depend more on previous knowledge show the lowest score increases. This imbalance may not reflect an increase in general intelligence, Flynn argues, but a shift in particular habits of mind. The question is not, why are we getting smarter, but the much less catchy, why are we getting better at abstract reasoning and little else?"

100 IQ is relative to average

[ls671]

IQ is relative, so even if people were getting smarter on average, they should not score higher in IQ tests :

http://en.wikipedia.org/wiki/Intelligence_quotient#Mental_age_vs._modern_method [wikipedia.org]

When an IQ test is constructed, a standardization sample representative of the general population takes the test. The median result is defined to be equivalent to 100 IQ points. In almost all modern tests, a standard deviation of the results is defined to be equivalent to 15 IQ points. When a subject takes an IQ test, the result is ranked compared to the results of the standardization sample and the subject is given an IQ score equal to those with the same test result in the standardization sample.

You missed something

[gr8_phk]

FTFA: "[A] person who scored 100 a century ago would score 70 today". So this means the scale has been adjusted and what we call average today would have been quite smart a century ago. You can't measure absolute IQ of a society, but you can do comparative studies of different societies or the same society at different times in its history.

Re:Other hypotheses- parasite load and nutrition

[JoshuaZ]

The Flynn effect is cross-cultural though, not just the United States, but essentially global. That's not consistent with the US increased emphasis on test taking. See for example http://www.sciencedirect.com/science/article/pii/S0160289604000522 [sciencedirect.com], http://www.sciencedirect.com/science/article/pii/S0191886905000711 [sciencedirect.com] for Norway and Australia as two example countries. Moreover, if this were caused by increased emphasis on test-taking you'd expect to see the entire bell-curve move up whereas most of the movement is on the lower end. Moreover, if testing were what mattered then the US military would have seen a decreased usefulness in IQ testing as an estimate for whether people will make good soldiers, and yet they haven't seen any decline in usefulness of the ASVAB http://en.wikipedia.org/wiki/Armed_Services_Vocational_Aptitude_Battery [wikipedia.org].