Remix.run Logo
tptacek 5 days ago

There is in fact not good evidence for IQ being "highly linked to genetics".

exoverito 5 days ago | parent | next [-]

It still amazes me that people believe that intelligence has no link to genetics. Down's syndrome is not simply a social construct, Trisomy 21 is a genetic disorder from a third copy of chromosome 21, distorting gene dosages and resulting developmental pathways. Nearly all phenotypes are a mixture of genetics and environment, yet some traits have a much higher degree of heritability. For example, height has a heritability of 60 to 80%, once you control for sufficient nutrition it's almost entirely due to genes.

The most elegant proof of IQ being linked to genetics:

The same person taking an IQ test twice has mean correlation of 0.85 or above in their scores. Identical twins reared together: 0.86 Identical twins reared apart: 0.76 Biological parent and child: 0.42 Adoptive parent and child: 0.19 And of course, any two random people will have a correlation close to 0.

https://en.wikipedia.org/wiki/Heritability_of_IQ#Correlation...

If you do not believe this, then I would have to hypothesize you are succumbing to motivated reasoning out of a deeper value system placing equality above all other values. This is a well known pattern of belief amongst leftists, where they think humans are infinitely malleable blank slates and all inequalities can be rectified given enough social engineering. They deny objective group differences because they want a utopia where everyone is equal. This is clearly unrealistic, but furthermore it contradicts their value of diversity, where if people are diverse, then you would expect variation in all traits, intelligence included.

legacynl 5 days ago | parent | next [-]

> The most elegant proof of IQ being linked to genetics: > > The same person taking an IQ test twice has mean correlation of 0.85 or above in their scores. Identical twins reared together: 0.86 Identical twins reared apart: 0.76 Biological parent and child: 0.42 Adoptive parent and child: 0.19 And of course, any two random people will have a correlation close to 0.

How is this proof of IQ being linked to genetics if adoptive parents can have 1/2 of the correlation that identical twins have? I think this proves that genetics only has a minor part, and upbringing/environment seems to be the most important factor.

exoverito 18 hours ago | parent [-]

First of all, you misread the figures. The correlation between identical twins reared together is 0.85, while adoptive parent and child is 0.19.

Second of all, correlation coefficients do not scale linearly. A correlation of r=0.85 means r^2 explains 0.72 of the variance, while a correlation of 0.19 explains 0.03 of the variance. In other words, adoptive parents have 0.03/0.72=4% of the effect on IQ as being an identical twin does.

tptacek 5 days ago | parent | prev | next [-]

I'm not a blank-slateist or an anti-hereditarian, as someone else claimed earlier today. I understand that cognitive disabilities are often causally --- mechanistically understood --- genetic.

But leaving aside things like Trisomy 21, your evidence here is twin study heritability. Heritability is not genetic determinism; it's almost a category error to claim otherwise, since "heritability" is really just a way of framing the question of whether something is genetically determined --- you still have to answer the question! There's a whole big research field controversy about this, "missing heritability", exploring (in part) why molecular genetics results, especially when corrected for things like within-family bias, are returning such lower heritability estimates than classic twin studies.

I do not believe that any random child selected at birth has an equivalent potential to win a Fields Medal, given the optimal environment to do that in. But the "hard truth nobody wants to face", from the parent commenter, is subtextually about race --- and there the evidence is a wreck; extraordinarily unlikely to bear any fruit.

sksinx 5 days ago | parent [-]

> and there the evidence is a wreck; extraordinarily unlikely to bear any fruit

Is there any new reading here? I used to follow this stuff much more closely a decade ago, but came to the conclusion most scientists will go to great lengths to avoid saying some races (if they don’t barrage you with pedantry regarding what race is) are on average different in some axis than others. There were a few out there who were able to say the politically incorrect thing only because objective science was strongly in their favor, but they still had the full force of the consensus academia coming down on them.

I lost interest when, much like history, it became obvious the field was too political for any real truth to be found. Maybe in 100 years or so.

tptacek 5 days ago | parent [-]

This is very actively studied, and the idea that it's somehow suppressed in the academy is another pervasive Internet myth --- put it on the shelf alongside "every employer would use IQ tests to hire but they're illegal" (they most certainly are not).

I think what some people are noticing is that there aren't splashy results to confirm, like, The Bell Curve. Yeah, that's because The Bell Curve was really dumb; it's from the phlogiston era of this science.

sksinx 5 days ago | parent [-]

> it's from the phlogiston era of this science.

I’ll try again in 100 years :)

dmbche 5 days ago | parent | prev [-]

Intelligence is not a defined measurable trait

But keep fantasizing you're born in the best race in the world, lucky you

skellington 5 days ago | parent | prev [-]

You're right. It's the only human trait that has no genetic link. Convenient. That's why dogs and humans also have equal IQs.