Race gaps in brain size in the very young -- figures here given the basic pattern but there's a lot more evidence. http://emilkirkegaard.dk/en/?p=7121 pic.twitter.com/Ea0R037Bri
-
This seems weak to me, esp. data on people of Asian ancestry in U.S. Also, even estimated cranial capacity has only R~=0.3-0.4 to IQ if same age, and R=0.46 for capacity at infancy vs at age 7. Noisy. https://www.quora.com/What-is-the-relationship-between-cranial-capacity-and-IQ-in-modern-humans … https://abc102.wordpress.com/2011/02/17/brain-size-and-correlates-with-iq/ … https://www.sciencedirect.com/science/article/pii/S0160289697900040 …
Asians are smaller, so one needs to adjust for that, not done above. The individual level correlation is not so important for comparing groups. Confusion of levels. See alsohttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC2668913/ …
-
-
Definitely agree with your second point and wanted to include it, but ran out of space. This issue comes up quite often in e.g. quant finance also. Regarding first sentence, though, very interesting that this does not appear to affect IQ across sexes.
-
Encephalization quotients using various divisors differ systematically between men and women, but group mean IQ appears identical in every reliable study I have seen on the subject.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2668913/figure/fig2/ …
-
This is the explanation from the cited paper, but (a) the claimed late-adolescent IQ difference only appears in some studies and not others, (b) encephalization quotients already differ before then, (c) purported domain-specific explanation rather muddled: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2668913/ …pic.twitter.com/kHoSRnggKs
-
It's a 9 year old review by now. I'm not so sure about the null sex difference. See recent findings here. http://mankindquarterly.org/archive/issue/58-1 … Remember that test makers have often tried to build tests so they show no gaps. This gerrymandering of items can be done to some extent.
-
Fair points. Given known processes of CNS maturation extend to at least age 25, while relatively unbiased samples become quite difficult to obtain beyond roughly age 16-18, it is at least conceivable. Timing of male late growth spurts does come awfully close to that limit.
-
It does indeed. With regards to item gerrymandering, see this study which showed that it was possible under constraints of some test properties (using NHST). http://www.sciencedirect.com/science/article/pii/S0160289616300423 …
-
Ah yes, for crystallized intelligence tests especially, it seems almost impossible by definition to avoid the ability to create systematic bias in one direction or another. Calibrating such a test for sex equality would be trivial.
-
There is a way. You choose the items after what predicts multiple, and relevant outcomes as well as shows lack of DIF. So, start by making e.g. 250 items, then choose 50 best.
- 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.