The below chart shows how six diverse populations averaged on (1) polygenic education scores, (2) brain size (3) environmental quality (4) IQ tests and (5) Guestimated IQ if everyone lived in First World. The last column was guestimated by adding 13 IQ points to people reared in the Third World type environments (Lynn 2006) and half that to people living in 2nd World type environments though these classifications are admittedly subjective. For example non-white Caucasoids live in some oil rich Arab countries but also nutritionally challenged South Asia. China has a lot of malnourished people, but those in Lynn’s samples were likely from First World type cities. Native Americans often live in the First World, but often in reservations that lack clean drinking water.
The 13 point correction was based on Lynn’s 2006 analysis that sub-Saharan environments depress IQ by 13 points. Brain size was not corrected for environment because the cranial capacities Lynn reported were from people who likely died before Western countries began really outdistancing the global South. Even in Western countries, they should be considered 3rd World brain size because they were from before the post-WWII living standards yet long after the excellent nutrition of the Paleolithic when brain size was inflated to First World levels.
| education PGS (Piffer 2018) | Brain size (Lynn 2006) | Guestimated environment level of Lynn’s IQ samples | IQ estimates derived Lynn (2006) | Guestimated IQ if everyone fully lived in First World | |
| Oceania | -2.1 | 1225 | 3rd Worldish | 62 | 75 |
| Middle East & South Asia | -0.4 | 1293 | 2nd World | 84 | 91 |
| Europe | 0.8 | 1369 | 1st World | 99 | 99 |
| East Asia | 1.2 | 1416 | 1st World | 105 | 105 |
| America | -0.75 | 1366 | 2nd Worldish | 86 | 93 |
| Africa | -0.2 | 1280 | 3rd World | 67 | 80 |
The below chart shows the correlation of both brain size and education PGS with population IQ, both when IQ is tested in their home countries and then again if tested with benefit of First World environment. Note these are (group-level) ecological correlations (Jensen, 1998), for individuals these correlations would be less than half as strong.
| brain size | education PGS | |
| raw population IQ | 0.90 | 0.58 |
| Guestimated population IQ if everyone lived in First World | 0.95 | 0.85 |
In both cases it seems brain size better predicts the IQ of a people than PGS does. When we compare the two scatter plots we see the line of best fits the data points better when brain size is the predictor:
Until Davide Piffer’s gets access to more full coverage genomes and includes a larger sample of IQ related genes, it’s wise to supplement his PGS scores with brain size data, and perhaps the best predictor of a population’s potential IQ is a combination of both (r = 0.96) but we’ll need a lot more than just six populations to test this conclusively:
Pumpkin Person 