I.Q. Test Paradox Resolved

K-12 Testing

In 1987, James Flynn showed that scores on IQ tests for the current generation were 9 to 20 points higher than scores of the previous generation. The average African American in the 1980s scored as well as whites did a generation before. The gains were least on IQ tests that most resembled basic school work and greatest on those measuring more abstract problem-solving activities.


The “Flynn effect” contradicted research by Arthur Jensen and others on kinship data (siblings raised together and apart). Jensen had promoted a model in which genetic differences explained 75% of the IQ test score variance between individuals, and he became notorious for arguing African genetic inferiority. But genetic endowments of “intelligence” could not have changed enough in a generation to cause the average person today to have an IQ equal to that of someone at the 80th percentile only a generation ago.


While Flynn’s work raised serious questions about IQ tests, researchers have been unable to construct an alternate model to explain the findings that Jensen attributed to genetics. Critics of IQ testing had pointed out flaws in the tests, questioned what they actually measured, uncovered fraudulent evidence and spurious reasoning among test proponents such as Cyril Burt and The Bell Curve authors Herrnstein and Murray (see Examiner, Winter 1996-97), and pointed to evolutionary biology for evidence that “races” among humans essentially do not exist.


Now, William Dickens and Flynn have proposed a model that resolves the paradox between Jensen’s kinship model and the “Flynn effect.” Essentially, “the reciprocal causation between IQ and environment” masks the actual effects of the environment, making gains and differences that are really rooted in the environment appear to be caused by genetics.


The authors use the marked improvement in basketball playing since the advent of television as an example. Those whose genes make them a bit better at basketball (tall and agile) are apt to practice more, play with other good players, and thereby keep improving. On a social scale, the advent of television exposed would-be basketball players across the U.S. to increasingly better examples of good play to emulate. “The rise in average performance feeds back into a new challenge for each individual,” Dickens and Flynn argue, and “every rise in individual performance raises the group average, which forces everyone to raise their individual performance a notch,” and so on.


The authors suggest that since the gains are highest on the most abstract tests, changes after World War II in work and leisure placed a greater social emphasis on more cognitively demanding activities which triggered a “social multiplier” that has resulted in the higher test scores. Their model shows that gains of 20 points in IQ in a generation are not implausible. Though they do not address racial gaps in detail, the authors point out that these social activities are unequally distributed by race and class, which would explain the persistence of IQ test score gaps.


• Psychological Review, 2001, V. 108, N. 2; http://www.apa.org/journal/rev/rev1082346.html; summary at http://www.brookings.edu/views/articles/Dickens/2001_psychreview.htm.
• See Flynn, J.R. (1987). “Massive gains in 14 nations: What IQ tests really measure.” Psychological Bulletin, 101, 171-191