Heritability is simply the percentage of the variation in a trait (i.e. IQ) that is “explained” by genes. It is equivalent to the correlation of genetically “identical” (i.e. MZ twins) people raised apart. But critics argue that identical twins raised apart are not separated in random ways because adoptive homes are not representative of American homes in general.

However the following formula allows us to estimate heritability without using adopted people:

H^2 = 2[(correlation between MZ twins raised together) – (correlation between DZ twins raised together)]

Based on data from wikipedia, this formula can be solved:

H^2 = 2 (0.86 – 0.55)

H^2 = 2 (0.31)

H^2 = 0.62

Although a heritability of 0.62 proves that HBDers correctly asserted that most of the variation in IQ (within Western countries) is “explained” by genes, it’s significantly lower than the heritability estimates reported by scholar Thomas Bouchard’s study of identical twins raised apart. Does this mean that critics of adoption IQ studies were right? Not necessarily. Bouchard’s data is from adults, and the data on twins reared together might not be, and since heritability is known to rise precipitously with age, that matters. Further, according to scholar Arthur Jensen, the formula I used above underestimates heritability slightly because it does not take into account the assortative mating that occurs in the parents of twins (and everyone else).

Lastly, all twin estimates of heritability are underestimates because identical twins are not 100% genetically identical.

So 0.62 is a very conservative estimate of adult heritability, yet even still, it’s quite high. Taking the square root of 0.62 tells us that IQ correlates about 0.8 with genes. This is roughly the same as two different IQ tests correlate with one another. In other words, once the genetic variants for IQ get identified, a DNA test will predict your IQ (on a future test) as accurately as your past IQ scores would. And since the IQ-genotype correlation is said to be highly mediated by g (general intelligence), while IQ tests correlate largely because of non-g factors (verbal ability, spatial ability, etc), a DNA test will likely be a better measure of g than an IQ test is!

Furthermore, if parts of Einstein’s brains are still persevered, it will be possible to sample his DNA and estimate how he would have scored on an IQ test had he been alive today, though that may not necessarily tell us much about how smart he was in his own time because of potential gene-environment interactions that don’t transcend generations. It might even be possible to examine Neanderthal DNA to test physicist Steve Hsu’s theory that neanderthal would score IQ 70 if cloned and raised in today’s America (I think that’s way too low an estimate for a people who adapted to the ice age although Hsu cites evidence that Neanderthals had low IQ genetic variants).