There were some extensive studies conducted by a sociology professor at the University of Western Ontario (UWO) in London Canada, in the early 90’s. I don’t remember the researchers name but apparently he did find differences in various intelligence tests performed on people with different ethnic backgrounds. He actually appeared on the Heraldo Rivera show to talk about his findings.
He was severely criticised by his peers and I also think he was censured by the university. I was a student there at the time; I never actually read his papers but many students demonstrated against him for doing such a study. I think it was in an era of staunch political correctness.
Cheers -> Dante
This is a fascinating article from Atlantic Monthly about the biology/genetics of the human family. It says that there is more genetic diversity amongst a group of chimps living on one hillside than there is between populations of humans all over the world. We are really very closely related, and skin color/hair texture does not translate to differences in things like brainpower.
As I mentioned earlier, I never read his study. I was too busy failing my regular classes. So, I dont know what the professor’s criteria was, what kind of data set he used or if he even used the scientific method.
As a side note, I do seem to remember that the professor correlated penis size to intelligence levels; If you can believe that!!! BTW, penis size was found to be inversely related to intelligence (big surprise there…lol).
Inferno, the professor to whom you refer is Rushton. Do a search of this Forum (you don’t even need to include Great Debates) to find a few other opinions of him along with further examples of his strange beliefs.
I don’t recall whether it was Collounsbury or edwino who pointed out in the last few weeks that Rushton has set up (in his mind) an “equalizing” scale of intelligence vs sex. You can have a lot of sexuality and poor brain power (Rushton’s view of Africans) or a lot of brain power, but little sexuality (Rushton’s view of Asians), or you can be well-balanced, having the “right” amount of each trait (guess who Rushton assigned to that niche?).
[/hijack – this not directly related to AWC’s OP ]
At present in the US there is a measurable difference between blacks, whites and asians on IQ tests (no, I don’t know how they define race).
On the other hand there is a lot of disagreement over exact figures, and:
Average western IQs in '81 ranged from 110 in Holland to 95 in France, (the US was 100). Siblings differ a lot in IQ - on average, by 12 points, compared with 17 for random strangers.
One of the interesting facts about IQ is the steady world-wide rise in test scores, often called the Flynn effect.* In Holland** mean IQs have increased more than 15 points–a full standard deviation–in the last fifty years, and the rate of gain may be increasing.
(Also from above site)
In other words, to answer the OP, (and talking about IQ rather than intelligence) there is evidence that differences between races are not significant, but there is no definitive answer, due to the many problems with such a study (as mentioned by others).
(*After Flynn’s studies in '84 and '87)
(**Other studies have found the same effect elsewhere)
Um, if you evaluate the extent to which 2 groups are different, you are implicitly evaluating whether they are similar. If you get an 800 on your SATs and I get a 500, debating whether those numbers are statistically different is the same as debating whether they are statistically equal. If they are not statistically different, that implies they are statistically equal. **
[/QUOTE]
This is a mistake. You could conceivably disprove all the evidence in favor of a difference and be left with no evidence of any sort. This is not the same as having positive evidence of no difference. You seem to have arrived at your idea by switching from “evidence” to “debating” and “evaluating”, but the OP speaks of evidence. Generally, it is conceded that the evidence in favor of racial differences can be attributed to other factors, but this is not the same thing as asserting that it must be attributed to other factors.
Ok, here’s my thinking. Estimates of a statistical relationship between two variables (say particulate concentrations and rainfall) are attached to measures of accuracy of those estimates. Political polls, for example, are often said to be accurate to +/- 3%. Call that interval a “confidence interval”.
Now, if increasing particulate concentrations by say 1 unit leads to 1 unit increase in rainfall +/- 3 units, it’s said that there’s no statistical relationship, (because the confidence interval contains zero.)
Question for Izzy: What would you consider positive evidence for no difference? Statistically, we can only establish a lack of evidence for any difference. That is, we can only establish a confidence interval that contains positive values, negative values and zero. But perhaps there is a nonstatistical approach which I am missing.
My understanding is that we can’t; if you want to go beyond that, you end up among those who believe that intelligence is a rough label applied to a rather diverse set of skills, and that this whole exercise is confused. I think it might be clearer if we looked at how facility at trigonometry varied with the usual demographic groupings. At least in that context, nobody would confuse learned with innate ability.
Somewhat. The line about flowbark belongs there, but after the comment about shortie. (The “um, if you” etc. is from flowbark).
Sorry for putting you through this trouble.
Actually you are missing the statistical approach. A confidence interval can be established which would establish that the difference is negligible or non-existent. (If you tested a large enough number from two groups, the chance that there is a meaningful difference between them which has not manifested itself becomes exceedingly small.)
In Collounsbury’s absence, I’ll just throw out the link to the very comprehensive page on race and genetics provided by e. neubauer, which includes links to the vast array of race threads from the SDMB.
Agreed. There could be a confidence interval whose boundary corresponded to a difference that was trivial or exceedingly small. In fact, this could occur even if the mean estimate was statistically different from zero.
So in that sense, there could exist evidence that the (self-defined) races are roughly equal in intelligence.
Finally, I must note that if you want to establish an association between a demographic category and a test score, it is important to control for other factors (eg education level, family income, parental education, etc.) It is almost always the case that controlling for such factors reduces and at times eliminates any single measured correlation or differential.
Just when Grienspace finally gave up on this argument, along came Autumn Wind Chick to keep things riled up. That said, I hope Grienspace is doing well…
Yeah, I meant to put that in. The data is rather fuzzy, but it’s somewhere around 108:100:90 for asians:whites:blacks IQs in the US. There’s some stuff on e. neubauer’s page as well, though I haven’t been able to read it all. I can’t remember finding anything about whether the white:asian difference is also shrinking (though I’d like to think it is:))