Quote:
Originally Posted by TomCowley
The biological pathway from genes to innate intelligence would basically require that basically all blacks have a bunch of something that's effectively dominant stupid genes in all/almost all intelligence-related functions. There's no reason the biology should be expected to work like that and not like something closer to an average.
It can certainly look like it is genetic. Blacks prior to the 1860s were genetically predisposed to being slaves in the US. Their phenotype proves it.
Most geneticists would argue against that interpretation though, I imagine.
Quote:
It's easy to selectively sample 1 SD apart. Go test buildings with white professionals and their black janitors and you might get 30 points. It's easy to legitimately BE 15 points apart- I mean it's easy to imagine that any particular/potential bright child, if you were so disposed, could be/have been turned into anything from average to full ****** with enough bad nutrition, disease, lack of intellectual stimulation, and whacks upside the head. Find a sample that does something badly relative to an otherwise identical sample and there's 15 points without trying, and none of it's genetic. Whatever genetic issues there may be, even if it's a full 15 points in some cases, aren't going to stand out obviously from random variation and all of the other environmental crap going on. It would be blindingly obvious something big was going on if US blacks were genetically -50. It wouldn't be obvious at -10.
To make it easy on you, there typically isn't a selection bias. We can use statistical techniques to control for most of the variables you are bringing up and there aren't many researchers out there stupid enough to believe that a random sample from a population isn't a random sample from a population (we even have tests to show that the sampling was random).
What we can't control for (in the studies between countries) is the country the person is in and (in the within country studies) the effects of looking like someone of a particular racial group.
Quote:
I found a link with a bunch of Irish-English schoolkid studies showing the gap has been shrinking, and the author attributed it to more relative urbanization in Ireland (which I assume actually happened) and said that no biological (i.e. genetic) explanation makes sense. So who knows.
Good boy. You have discovered the Flynn effect. It occurs from time to time and place to place.
We also don't find the Irish-English IQ differences in the USA or in Australia. We used to find them in the US.
The point that you seem to correctly be arriving at is that these sorts of studies cannot be interpreted as showing a genetic effect (unless you believe that some strange gene-environment interaction is in effect).
The only way of figuring out whether the cause of the population differences is due to genetics is to find the genes specifically responsible. I don't know about you, but I don't care whether I mate with an Ashkenazi Jew or African American as much as whether I mate with someone who carries the gene for Tay-Sachs disease or sickle cell anemia. This is mostly because, despite my best efforts, I rarely get to mate with a population instead of an individual.*
Quote:
The problem with controlling for occupation too carefully is that if the occupation itself self-selects for a particular IQ range (and it's pretty obvious that they do), or even worse self-selects differently by race, then you've lost representative sampling and muted whatever difference there might be. I mean if you test only janitors, you'll get some B-W difference, but if 20% of the white population turned into supergeniuses (and sought employment accordingly), and you tested again, you'd get about the same difference because ~none of the supergeniuses would be janitors and you'd totally miss the huge difference in population averages. Or when your occupation is "voluntarily enlisted to go overseas in the army post-war", you basically lose the entire right tail and its effect on the population difference.
You are imagining the problem to be a bit bigger than it is. We don't just run one analysis on one set of data and call it a day.
*Technically, that isn't true. What we could do is do test-tube babies assigned to random wombs, peel their skin off at birth, do some plastic surgery a la Michael Jackson and raise them in standardized environments, but that study probably won't be done any time soon.