Logo IMG
HOME > PAST ISSUE > Article Detail


Rising Scores on Intelligence Tests

Test scores are certainly going up all over the world, but whether intelligence itself has risen remains controversial

Ulric Neisser

Increase or Artifact?

These gains are far too rapid to result from genetic changes. There evidently are substantial environmental influences on g, even if we do not clearly understand them at the present time. Moreover, the sheer size of the gains undermines the very concept of "general intelligence." To see why, consider that individuals with IQ scores over 130 are typically regarded as "very superior" and those with scores under 70 as "intellectually deficient." In any normal distribution of scores with a mean of 100 and a standard deviation of 15, about 2.25 percent of the population can be expected to fall into each of those categories. This was indeed roughly the case for the first Stanford-Binet standardization sample in 1932. An ongoing rise of 0.3 IQ points per year means, however, that if a representative sample of the American children of 1997 were to take that 1932 test, their average IQ would come out to be around 120. This would put about one-quarter of the 1997 population above the 130 cutoff for "very superior"—10 times as many as in 1932. Does that seem plausible?

If we go by more recent norms instead, we arrive at an equally bizarre conclusion. Judging the American children of 1932 by today's standards—considering the scores they would have obtained if they had somehow taken a test normalized this year—we find that their average IQ would have been only about 80! Hardly any of them would have scored "very superior," but nearly one-quarter would have appeared to be "deficient." Taking the tests at face value, we face one of two possible conclusions: Either America is now a nation of shining intellects, or it was then a nation of dolts.

If we focus instead on the most g-loaded tests, the gains seem even more preposterous. The mean Raven IQ in the Netherlands rose by 21 points between 1952 and 1982; extrapolating backward, there has probably been something like a 35-point increase since the 1930s. Dutch 19-year-olds today are getting scores that would have been more than two standard deviations above the mean in their grandfathers' time. The size of these gains boggles the mind. If they do not reflect some trivial artifact, we (and especially the Dutch!) must be living in a truly remarkable age of genius. As Flynn puts it, the data imply that dozens of nations should now be in the midst of "a cultural renaissance too great to be overlooked." Because that does not seem to be happening, Flynn concludes that the tests do not measure intelligence but only a minor sort of "abstract problem-solving ability" with little practical significance.

Flynn's extreme position can help us organize the range of hypotheses that have been proposed to explain the rise in scores. Some hypotheses—including increases in test-taking sophistication and in the motivation to score well—are consistent with the view that there has been no real rise in intelligence at all. Other possibilities—including the impact of worldwide improvements in health and nutrition—conflict with Flynn's conclusion, suggesting that intelligence has really gone up. Several more subtle hypotheses—that the gains have been produced by changes in schooling, in child-rearing practices or by more general aspects of culture—lie in between these two extremes. Let us consider these possibilities one at a time.

comments powered by Disqus


Subscribe to American Scientist