Read on Twitter
Jason Richwine wrote a whole PhD thesis advocating for IQ testing immigrants to keep dumb peasants out: https://delong.typepad.com/pdf-1.pdf The weak analysis he does to support his thesis uses Lynn & Vanhanen's National IQ dataset. In case you weren't aware, this "dataset" has problems..
1/ https://twitter.com/NewsfromScience/status/1326728951356657665
1/ https://twitter.com/NewsfromScience/status/1326728951356657665
Basically, Lynn & Vanhanen's National IQ dataset is a very elaborate work of scientific fraud. See https://psyarxiv.com/tzr8c/
https://sci-hub.se/https://www.sciencedirect.com/science/article/abs/pii/S0160289609001470
https://sci-hub.se/https://bpspsychub.onlinelibrary.wiley.com/doi/full/10.1348/135910707X174339
https://sci-hub.se/https://www.researchgate.net/publication/222668167_A_Systematic_Literature_Review_of_the_Average_IQ_of_Sub-Saharan_Africans
2/
https://sci-hub.se/https://www.sciencedirect.com/science/article/abs/pii/S0160289609001470
https://sci-hub.se/https://bpspsychub.onlinelibrary.wiley.com/doi/full/10.1348/135910707X174339
https://sci-hub.se/https://www.researchgate.net/publication/222668167_A_Systematic_Literature_Review_of_the_Average_IQ_of_Sub-Saharan_Africans
2/
Issues with the Lynn & Vanhanen dataset include, but are not limited to:
1) tiny sample sizes
2) unrepresentative samples of the population
( @RebeccaSear wrote a thread on this) https://twitter.com/RebeccaSear/status/1271547090221572096
(excerpt from https://psyarxiv.com/tzr8c/ )
3/
1) tiny sample sizes
2) unrepresentative samples of the population
( @RebeccaSear wrote a thread on this) https://twitter.com/RebeccaSear/status/1271547090221572096
(excerpt from https://psyarxiv.com/tzr8c/ )
3/
3) ambiguous inclusion criteria.
The criteria Lynn & Vanhanen use to include/exclude IQ studies is never given. Instead they cherry pick older lower quality studies that estimate lower IQ's for Africans (hmm, wonder why?)
(excerpt from https://sci-hub.se/https://www.researchgate.net/publication/222668167_A_Systematic_Literature_Review_of_the_Average_IQ_of_Sub-Saharan_Africans)
4/
The criteria Lynn & Vanhanen use to include/exclude IQ studies is never given. Instead they cherry pick older lower quality studies that estimate lower IQ's for Africans (hmm, wonder why?)
(excerpt from https://sci-hub.se/https://www.researchgate.net/publication/222668167_A_Systematic_Literature_Review_of_the_Average_IQ_of_Sub-Saharan_Africans)
4/
4) over half the national IQ's Lynn & Vanhanen report is simply made up (ahem, I mean *imputed*) https://en.wikipedia.org/wiki/IQ_and_the_Wealth_of_Nations
More on this later...
5/
More on this later...
5/
5) the dataset was so sloppily put together that even people who thought it was valuable found glaring errors
Maybe this gives you an idea of what kind of rigorous scholarly practices are on display in other aspects of this dataset.
(excerpt from https://sci-hub.se/10.1016/j.intell.2006.11.002)
6/
Maybe this gives you an idea of what kind of rigorous scholarly practices are on display in other aspects of this dataset.
(excerpt from https://sci-hub.se/10.1016/j.intell.2006.11.002)
6/
5) IQ estimates that were obtained decades apart are assumed to be comparable (because apparently national IQ never changes!)
6) IQ estimates from widely different age groups are assumed to be comparable (because apparently IQ doesn't change from childhood to adulthood!)
7/
6) IQ estimates from widely different age groups are assumed to be comparable (because apparently IQ doesn't change from childhood to adulthood!)
7/
7) IQ tests are assumed to "culture fair", with cross-cultural comparisons not being an issue.
Except they aren't "culture fair" (see @spiantado)
8/ https://twitter.com/spiantado/status/1275783954411347968
Except they aren't "culture fair" (see @spiantado)
8/ https://twitter.com/spiantado/status/1275783954411347968
7 continued)
For example, Lynn & Vanhanen take this "culture fair" assumption to most pseudoscientific extreme.
(excerpt from: https://www.nybooks.com/articles/1994/12/01/the-tainted-sources-of-the-bell-curve/)
9/
For example, Lynn & Vanhanen take this "culture fair" assumption to most pseudoscientific extreme.
(excerpt from: https://www.nybooks.com/articles/1994/12/01/the-tainted-sources-of-the-bell-curve/)
9/
I find it pretty amazing how a failson like Jason Richwine can get a PhD from Harvard by pretending to do science, and continue to fail upwards through right-wing think tanks all the way to a senior National Institute of Standards and Technology post. Truly inspirational
10/
10/
