I’m reading James Flynn’s new book Does Your Family Make You Smarter? and I’m absolutely blown away by how subtle and creative his ideas are.  Unlike many environmentalists, who simply nitpick the research of IQ hereditists, Flynn actually makes his points with great statistical creativity.

In order to understand this book, you must first understand that according to the conventional wisdom, our IQs have three main causes:

  1. Genes: The effect of genes on IQ can be measured by correlating the IQs of identical twins raised apart (though the correlation should be reduced a bit for shared prenatal environment)

2.  Family environment:  The effect of family environment can be measured by    correlating the IQs of unrelated people raised in the same home.

3. Chance environment:  These are random environmental events that affect our individual IQs, but don’t effect our siblings raised in the same family.  For biological determinists like Arthur Jensen and I, these might be getting hit on the head with a golf ball or not getting enough oxygen in the womb.  For cultural determinists like Flynn, these might be having a school teacher who inspires you to learn.  The effect of chance environment on IQ can be measured by noting that there’s a higher IQ correlation between the same person tested twice, then there is between two identical twins raised in the same home.  In other words, you are more similar to you, then you are even to your identical twin raised in the exact same family, so there are obviously unique environmental effects that make even identical twins raised together, individuals.

Now according to Jensen’s (1998) summary of kinship and adoption studies, genes explain about 45% of the IQ variation in childhood, 65% of the variation in adolescence, and about 80% in later maturity.  Family environment explains about 35% of the IQ variation in childhood, and near zero by late adolescence.  Meanwhile chance environment explains about 25% at all ages.

Why does family environment go from 35% to zero percent?  The conventional wisdom is that people with high IQ genes do well in school and end up in stimulating adult environments, and these adult environments replace the effects of family environment.  Further, because their genes caused these adult environments, the effect shows up as genetic variance.

Flynn’s brilliant insight

I’m going to do my best to explain Flynn’s ground-breaking new method. Although so far I’ve only read the first few chapters,  I hope I’ve at least partly grasped the concept (it’s pretty subtle):

High IQ adults outperform high IQ children on tests, both because of age differences, but also because by adulthood, high IQ genes create high IQ environments, creating a multiplier effect.

By contrast average IQ adults score higher than average IQ children, ONLY because of age differences.  In a normal curve, by definition the average person is, on average, average in genes and environment.

Thus by subtracting the test performance gap between average IQ adults and average IQ kids, from the performance gap between high IQ adults and high IQ kids, you get a pure measure of the effect of non-chance environment on test performance.

What this brilliant method shows is that on tests that I have always known are quite culture reduced, like using colored blocks to form designs, the effect of non-chance environment vanishes by the early teens, while on tests that are obviously cultural (vocabulary), non-chance environment persists into adulthood.

To me, the beauty of this method is that two tests could have the exact same heritability, but one test could be heritable for indirect reasons (genes causing environments that improve test performance) while the other test is heritable because genes directly affect performance.

How ironic that Flynn, the man best known for debunking the notion of culture reduced tests via the Flynn effect, may have unintentionally proven they exist after all.