Russell's problem is that his paper was inundated with extremely low p-values and extremely low explanatory power (highest is a 14% correlation w/ ACT). Gould was correct that the Army Beta was a poor test. He was wrong in its implications for IQ as a concept on the whole.
@Russwarne summarizes his recent publication on the Army tests used in World War 1, and S J Gould's discussion of them.
There is only really one left-wing approach to eugenics: state fund it. Elites will still use it more, but not because of cost.
https://quillette.com/2019/03/19/the-mismeasurements-of-stephen-jay-gould/ …pic.twitter.com/HfNiaeALMQ
-
-
-
The low correlations of the current administration has to due mainly with the floor effects caused by Flynn effect, no? See my reanalysis: http://rpubs.com/EmilOWK/Warne_army_beta2019 …
-
I thought the authors chalked it up to Gould's students simply being of a higher caliber than Harvard's more current students (which I might be inclined to be believe given the news). They said they tried to control for Flynn, but let me take a look at this. I'm very interested.
-
Look at the item analysis plot. Almost all the items are most informative for -3 z subjects, meaning they suffer from extreme floor effects. Warne et al should try administering this test battery to like 6th graders.pic.twitter.com/dzoJnedIFl
-
Yes, I agree. Perhaps the test is not poorly designed, but is very poorly administered in its context. This is a test designed for non-verbals being given to Harvard students...it may be harder to find a more asymmetrical gradient regarding test reliability.
End of conversation
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.