Chuck

Some guy

[Chuck @acoolcoolhippo]

@JoannaTeg @ELVUgo @KaylaKatin @nathancofnas @Miss_Snuffy Jamaica participated in the previous wave of PISA, allowing a direct comparison with the UK and USA. The students scored about 1 standard deviation lower, putting them about the same amount ahead of African countries. On regional tests such as CXC Barbados outperforms Jamaica.

[Chuck @acoolcoolhippo]

@JoannaTeg @ELVUgo @KaylaKatin @nathancofnas @Miss_Snuffy Caribbean countries such as Barbados do well.

[Chuck @acoolcoolhippo]

@EzeugwuMartin @nathancofnas @Miss_Snuffy Kenya is an example of a country that, embracing cognitive-capitalism, has invested heavily in education and has markedly increased student academic outcomes. In the last round of SACMEQ regional testing Kenyan scored around 6/10 of a standard above other African countries.

[Chuck @acoolcoolhippo]

@SashaGusevPosts @the21stgman As Bruce Weir already comented "Hudson (1992) and Nei (1986) [Gst] are the same for populations in Hardy-Weinberg equilibrium". Recall, he was surprised you would argue this and so requested that I write.

[Chuck @acoolcoolhippo]

@kstrna62 @PsychRabble @6851cf3c @MeltedJonSnow @yungmulababylol This was essentially the 19th-century argument for why human races -- in the sense of lineages or ancestry groups -- were, taxonomically speaking, varieties and not species. Wallace summarized the argument as:

[Chuck @acoolcoolhippo]

@kstrna62 @PsychRabble @6851cf3c @MeltedJonSnow @yungmulababylol One typical understanding of "race" in Nature History was precisely ancesty group -- as Mr. Darwin noted:

[Chuck @acoolcoolhippo]

@kstrna62 @PsychRabble @6851cf3c @MeltedJonSnow @yungmulababylol 'Race' (or 'lineage' in the original French) & 'ancestry' sound synonymous to me. Is the argument that we can't make 'valid' ancestry-based classes like the typical biogeographic ancestry groups used in genetic epidem -- that we can only talk about ancestry, not ancestry groups?

[Chuck @acoolcoolhippo]

@AlexTISYoung @SashaGusevPosts (4) So importance of results is they reinforce finding of very low SR h² for EA, raising doubt that the trait is substantially "sociogenetic", especially in less-developed countries, which mainly challenges inter-individual hereditarianism for EA & only by extension between.

[Chuck @acoolcoolhippo]

@AlexTISYoung @SashaGusevPosts (3) Regardless, education in this sample can't serve as a genetic proxy for cognitive ability among sibs, since SR h² ≈ 0 — so the Jensen-like hypothesis couldn't be tested this way.

[Chuck @acoolcoolhippo]

@the21stgman @SashaGusevPosts Not a big deal, but I was waiting for your reply before finalizing the qeios preprint.

[Chuck @acoolcoolhippo]

@the21stgman @SashaGusevPosts Speaking of which, Sasha — did you concede that I was correct about the neutral expectation? I didn't see a reply.
Would imply that assertions that there are only trivial genetic differences across major ancestry groups carry a burden of genetic evidence.

[Chuck @acoolcoolhippo]

@grok @craterbrainlet @abysmalpudding @ZoomerHistorian @JFGariepy @grok Would you agree that PCA graphs show multivariate genetic similarity? But it's a mistake to cite them as contradicting the meta-population relative coefficient of relatedness, which measures excess IBD relative to the metapopulation baseline?

[Chuck @acoolcoolhippo]

@grok @craterbrainlet @abysmalpudding @ZoomerHistorian @JFGariepy @grok So you agree?: In the Wrightian subdivided-population framework (metapopulation-relative kinship), the r is higher between parent and random co-ethnic kin than parent and mixed-race child when (approx.) FST > 0.20—as shown here: https://t.co/JKmrX8RrUn

[Chuck @acoolcoolhippo]

@grok @craterbrainlet @abysmalpudding @ZoomerHistorian @JFGariepy "@grok No need to invent a split between 'coefficient of relatedness' and 'pedigree/Mendelian relatedness'—they're the same in kin selection contexts. Pedigree r=0.5 is the local/subpopulation framing. The FST-adjusted version is just a metapopulation-relative reframing.

[Chuck @acoolcoolhippo]

@craterbrainlet @abysmalpudding @grok @ZoomerHistorian @JFGariepy @grok, clarify why the coefficient of relatedness is not 'always 50%' when dealing with subdivided populations (Wright 1965; Rousset 2002; Harpening, 2002).

[Chuck @acoolcoolhippo]

@ApolloRegime @shi_huang5 @skdh @nrken19 @johnhawks @PIE_Animals @Sulkalmakh @juliet_turner6 Key point often overlooked: relatedness is 'relative to' a reference population. Wright's framework conditions relatedness on the metapopulation average (not ancestral pop), resulting in zero average relatedness to a random metapopulation person, aligning with intuition.

[Chuck @acoolcoolhippo]

@ApolloRegime @shi_huang5 @skdh @nrken19 @johnhawks @PIE_Animals @Sulkalmakh @juliet_turner6 Note: In Wright's framework, relatedness is defined as the excess probability of IBD relative to the metapop average; under this definition, the FST threshold drops to .20. This excess-similarity view seems to better reflect Gariépy's original intent. See: https://t.co/JKmrX8RrUn