4 ms·

I would be very cautious citing that paper. They draw some very strong conclusions based on unsupported assumptions and evidence of dubious significance from po

I would be very cautious citing that paper. They draw some very strong conclusions based on unsupported assumptions and evidence of dubious significance from populations with unusual cultural features. > Therefore, we conclude that both variance and [variance ratio] in mathematics performance vary greatly among countries. Confirming our earlier finding ([21]), we also conclude that [variance ratio] is reproducibly essentially unity for some [predominantly muslim or low gender equality] countries. These findings are inconsistent with the greater male variability hypothesis.

kaitai 8y ago
What's so unusual about Muslim countries? or countries with high inequity? They seem to collectively have a pretty big share of human population. Don't fall into the WEIRD trap (Western, Educated, Industrialized, Rich, and Democratic).
- hackinthebochs 8y ago
  Because these countries also have very unusual patterns regarding education. Some examples mentioned in the study were non-normal distributions of scores, and a large percentage of religion-focused schooling. If we're trying to draw conclusions about innate differences operationalized as distribution test scores, any kind of biased schooling undermines the ability to draw conclusions from that population.
  - kaitai 8y ago
    But schooling in the US is biased in all sorts of ways, and trying to draw conclusions from that is being encouraged in this discussion.
    - hackinthebochs 8y ago
      True, but not all bias is created equal. You have to consider the nature of the potential bias and how it might affect measured outcomes. It's much less clear how bias might unnaturally increase variance than reduce it or alter the mean. But my point was mainly to argue against the conclusion that a lack of observed variance differences in some countries demonstrated the hypothesis false. I agree that the observed variance differences in western countries doesn't prove it true either.
nabla9 8y ago
> unusual cultural features. If culture affects the outcome, it waters down the premise of the variability difference theory. And what is unusual about Czech Republic, where VR and gender gap were near unity and zero?
- hackinthebochs 8y ago
  >it waters down the premise Couple of points here. A culture biased along gender lines may bias the outcome thus obscure the effect of genetics on gender differences. Also, no one in their right mind claims that environment has no effect, and so measuring some environmental influence doesn't undermine the greater male variability hypothesis. >And what is unusual about Czech Republic I wouldn't hazard a guess, but the "paradoxical effect" of gender equality negatively correlating with female interest in STEM might have some hard-to-tease-out influence here.

kaitai 8y ago

What's so unusual about Muslim countries? or countries with high inequity? They seem to collectively have a pretty big share of human population. Don't fall into the WEIRD trap (Western, Educated, Industrialized, Rich, and Democratic).

hackinthebochs 8y ago
Because these countries also have very unusual patterns regarding education. Some examples mentioned in the study were non-normal distributions of scores, and a large percentage of religion-focused schooling. If we're trying to draw conclusions about innate differences operationalized as distribution test scores, any kind of biased schooling undermines the ability to draw conclusions from that population.
- kaitai 8y ago
  But schooling in the US is biased in all sorts of ways, and trying to draw conclusions from that is being encouraged in this discussion.
  - hackinthebochs 8y ago
    True, but not all bias is created equal. You have to consider the nature of the potential bias and how it might affect measured outcomes. It's much less clear how bias might unnaturally increase variance than reduce it or alter the mean. But my point was mainly to argue against the conclusion that a lack of observed variance differences in some countries demonstrated the hypothesis false. I agree that the observed variance differences in western countries doesn't prove it true either.

nabla9 8y ago

> unusual cultural features. If culture affects the outcome, it waters down the premise of the variability difference theory. And what is unusual about Czech Republic, where VR and gender gap were near unity and zero?

hackinthebochs 8y ago
>it waters down the premise Couple of points here. A culture biased along gender lines may bias the outcome thus obscure the effect of genetics on gender differences. Also, no one in their right mind claims that environment has no effect, and so measuring some environmental influence doesn't undermine the greater male variability hypothesis. >And what is unusual about Czech Republic I wouldn't hazard a guess, but the "paradoxical effect" of gender equality negatively correlating with female interest in STEM might have some hard-to-tease-out influence here.