krichard12

Look at what happens to student test scores (top panel) and home values (bottom panel) when school bonds are authorized (red line).

No, You Can’t Design Your Baby—And Trying Would Be a Terrible Idea A genomics firm saying they can help parents with “genetic optimization” of their embryos is tone-deaf Silicon Valley marketing trampling over legitimate science. Parents should be wary

Excellent takedown of Nucleus Embryo by Arthur Caplan and James Tabery. The problem with Nucleus isn't eugenics. It is rescued from its own eugenic intentions by the fact that it won't work. It is bogus science.

""So, to be clear, using the statistical basis of factor analysis we now conclude “There is not a g factor, there never was a g factor, and we can demonstrate this with anyone’s data”"

(page 16)

krichard12.com/docs/McArdle...

I might be wrong about the last point

The current state of the art is using item level data directly. Sadly it depends on the chosen IRT model, and power is still low. Also, it cannot detect AxC if the estimated C is zero, which is ofc a problem since AxC acts very much like A
ris.utwente.nl/ws/portalfil...

I looked through all my bookmarks, sadly I failed to find it. However I'm 99% sure the statement was made by Mcardle or Horn.
On the bright side I managed to recover this lost article my Mcardle, where he has to admit that Horn was right all along, g is not real.

krichard12.com/docs/McArdle...

Yeah Guttman is extremely underrated. Peter Molenaar also proved the equalivence between network models and common factors models in the early 2000s or late nineties. Network analysis is not that new, it just finally gained some traction

Did you know that one of the most influential g-men has written that Gould was correct about factor analysis? I'll dig up the quote someday

Installing an AC a few years ago was one of the best decisions of my life. It's good enough for heating in winter too, I never have to use gas for heating anymore, I'm pretty sure it's a net positive overall

In both the article and twitter reply, he claims to have fit both an independent pathway model(high c) and a common pathway model(low c, high a)

In the paid article he cites a different article(that's why he messed up the publication date).
But of course this article also notes that it is not identified.
link.springer.com/article/10.1...

Hereditarians claim that these are not g-laoded gains, whatever that means. Schooling doesn't inrcrease intelligence because the gains aren't on g(except for the studies that find the opposite result.). Most theories agree that g is just a statistical thing, it doesn't exist literally.

A few additional interesting points about Jordan Lasker. One, he appled to work as a graduate student in my lab years ago:

bsky.app/profile/ent3...

Took me a while, but here is my response to Michel's challenge regarding their new GWAS of the five factor model. I trust it is sufficiently grumpy....

This is so true, altho it is very weird. Like how can the human brain just start to understand memorized text without conscious effort to do so.

The moment the tiktok community found out that the shark one mocked God, its popularity went down and the tone of the comments changed which is another good sign that kids are not manipulated into hate *that* easily.

Only the shark and crocodil one afaik(but they are among the most popular). The rest are either random words that gpt made up, or are fairly innocent. Some are not even Italian, but only a few know. This implies that thankfully they don't translate them, so are not affected that much.

semPlot is better
sachaepskamp.com/semPlot/exam...

krichard12.substack.com/p/misestimat...

I wonder whether he kept his lies on the internet and IRL the same. He claimed being related to 4 chess grandmasters, being fluent in German, holding a degree in Chemistry, being a 6'3 bodybuilder among other things.

They go by the "they, them" pronouns FYI

I can't find the post, but he said something like "even the anti-g dutch psychometricians agree if you push them enough"

Crem used to claim on Reddit that "he was actually in contact with (dutch) researchers, and they agree with him but they won't admit it in public"

Turns out that "being in contact" meant that he sent out an email to some of them. And they don't agree with him. And he is horrible at using R.

Also, the PGS was made for predicting this phenotype. All the other features in the model weren't.

yes, but haven't read yet

zzz.bwh.harvard.edu/library/purc...
No published strong test for the EEA or existence of gxe exist, but there are simulations about how gxe affect ACE estimates.

What practical model can one use to estimate GxC from family data? Seems to me that all of the C variance goes to A due to GxC in twin and ETFD models.

The author is on Bluesky too @sillyolyou.bsky.social

stop it Get some help poster

If you find yourself talking about heritability as quantifying nature vs nurture

Even fricking TikTok has a higher average intelligence than twitter at this point. While things like this go semi-viral on TikTok, they mostly get ratioed by literal teenagers.

Meanwhile race realists on twitter are confusing themselves with colored eigenvector plots...