4 ms·

I have several objections to the paper and to some of the responses here; in fact, it's really got me hot under the collar. It is not clear that this study can

I have several objections to the paper and to some of the responses here; in fact, it's really got me hot under the collar. It is not clear that this study can support such sweeping claims. Further, due to the fast-moving nature of this forum and the high velocity of misinformation, particularly among certain elements of the online media (https://www.xkcd.com/882/ https://www.xkcd.com/882/), I feel that I must take the time to reply now rather than waiting till this evening if I want anyone to read this reply, and so I must attempt to digest a very dry, verbose work of dubious quality as well a substantial related literature ASAP, and I resent like hell having to waste some prime work hours to do it. Therefore, I apologize in advance if I seem a bit short-tempered, unfair, uneven, or less-than-rigorous in this response. First, the paper itself. The buried lede is certainly the fact that previous, purely genetic studies based on bulk sequencing alone could only find a 5% variation in poly-genetic scores in the metrics for years of educational attainment ("EduYears") selected by Smith-Wooley et al (refered to from here on as the "authors"). Other papers cited by the authors find values above and below this number, but not exceeding %20. The papers broadest, proudest claim, that around %50 of various sorts of educational attainment (with correlations of 0.27–0.76 distributed across various "university success" variables) is determined genetically, is derived entirely from their twin study. Twin studies, at their core, analyse differences between sets of fraternal and identical twins. Fraternal twins share around %50 of their DNA on average, like most siblings. Identical twins share nearly 100%. Since both members of a twin set are assumed to experience an equal environment (the Equal Environment Assumption, or EEA), then differential phenotypes exhibited by fraternal twins and not identical twins can be explained entirely by genetic differences, or so it is claimed. However, the validity of twin studies isn't settled science, and ought not be held in the same regards as, say, a computer-verified proof of the four-color theorem or the discovery of the Higgs Boson. In particular, there exists a large body of work questioning the validity of the EEA in particular(e.g. https://scholar.harvard.edu/files/morris-singer/files/2008.pdf https://scholar.harvard.edu/files/morris-singer/files/2008.p...). There are questions about whether twin studies capture the actual (rather than assumed) genetic differences between different types of twins, issues regarding potentially-heritable epigenetic markers, and the differences between additive vs Mendelian and epistatic inheritances. These problems, in addition to other sources of error, when left unaddressed and compounded over several layers of statistical analysis, severely weaken any hypothesis built upon the assumption of their absence. With twin studies, these issues seem to be especially apparent surrounding studies that focus on the heritability of mental disorders. You don't have to believe me, though. As always, you should do your own research. Many of these problems can be remediated with additional research. For example, researchers can use questionnaires for the twin sets' families in order to ferret out unexpected violations of the EEA and unusual genetic circumstances. However this study seems to have been going after low-hanging fruit, using mostly archived data to reach its conclusions. Regarding the responses, I see a field of slain straw-men as far as the eye can see, and a I see a collossal tower of ick being built on a foundation of shifting sand. It is disingenuous in the extreme to equate Enlightenment ideals ("All men are created equal") with some sort bizzare desire for a Harrison Bergeron knock-off fantasy. Literally no one wants that. Further, the idea that people should be uncomfortable with genetic reductionist thinking seems to me to be common sense in light of the horrors of the first half of the twentieth century. Throw in the casual attacks on public education at all levels, and implications that we should bow down to a Calvinist tyranny of genetics," and I really have to wonder what is going on here. Now if you'll excuse me, I have a beautiful little ubermenchlet who just awoke from a nap to play with.

bobcostas55 8y ago
About the EEA, scroll down to "Assumption 2" here: https://onlinelibrary.wiley.com/doi/full/10.1111/1745-9125.12049 https://onlinelibrary.wiley.com/doi/full/10.1111/1745-9125.1... PGS variance explained and twin study heritability estimates can't really be compared, the former is just a floor which will be raised with larger sample sizes.
- unknownkadath 8y ago
  This article, and the Barnes and Simon article it is critiquing, are not exactly casual reading, friend. Are you the author of this article? Could you break it down a bit for an interested outsider? In particular, if the EEA is not so important, what is the value of including fraternal twins at all, compared to, say, other close siblings? And doesn't it raise questions if a mathematical model gives the same results regardless of whether or not the assumptions used in its development are followed? Also, what do you think of Barnes and Simon's reply: https://www.gwern.net/docs/genetics/heritable/2015-burt.pdf https://www.gwern.net/docs/genetics/heritable/2015-burt.pdf I understand if you don't have time to answer in detail, all of the relevant paper is probably pushing 100 pages of dense inside baseball at this point. I merely wish for the folks with, ahem, deep convictions about heritability up-thread to understand that this paper isn't as unassailable as its press releases are making it out to be.
Valencian 8y ago
Thanks for your response, I kept checking this article/thread over and over again during the last day or two because I am precisely afraid of a world were our lives are "mandated" at a 100% before we're even born -I don't claim that genetics don't matter, I've always thought (even before I understood more and more the scientific literature on the subjects) that it's roughly 50% 50% in environment vs genetics, but the fact that "genetics" can be used to justify our political order and worldwide not for 100% of humanity but for only a certain percentage of it, say the rich or whatever, like it's happening here in HN en masse and like you correctly said that " Throw in the casual attacks on public education at all levels, and implications that we should bow down to a Calvinist tyranny of genetics," and I really have to wonder what is going on here.". I'd be delighted to read an expanded response in regard to this subject by you :) Edit: I like to treat this particular subject as dealing with the development of nuclear bomb. You cannot just say "oh, I am just researching this for the sake of science" when you clearly know that once you have that nuclear bomb and can put it to use, you can change the whole world.

bobcostas55 8y ago

About the EEA, scroll down to "Assumption 2" here: https://onlinelibrary.wiley.com/doi/full/10.1111/1745-9125.12049 https://onlinelibrary.wiley.com/doi/full/10.1111/1745-9125.1... PGS variance explained and twin study heritability estimates can't really be compared, the former is just a floor which will be raised with larger sample sizes.

unknownkadath 8y ago
This article, and the Barnes and Simon article it is critiquing, are not exactly casual reading, friend. Are you the author of this article? Could you break it down a bit for an interested outsider? In particular, if the EEA is not so important, what is the value of including fraternal twins at all, compared to, say, other close siblings? And doesn't it raise questions if a mathematical model gives the same results regardless of whether or not the assumptions used in its development are followed? Also, what do you think of Barnes and Simon's reply: https://www.gwern.net/docs/genetics/heritable/2015-burt.pdf https://www.gwern.net/docs/genetics/heritable/2015-burt.pdf I understand if you don't have time to answer in detail, all of the relevant paper is probably pushing 100 pages of dense inside baseball at this point. I merely wish for the folks with, ahem, deep convictions about heritability up-thread to understand that this paper isn't as unassailable as its press releases are making it out to be.

Valencian 8y ago

Thanks for your response, I kept checking this article/thread over and over again during the last day or two because I am precisely afraid of a world were our lives are "mandated" at a 100% before we're even born -I don't claim that genetics don't matter, I've always thought (even before I understood more and more the scientific literature on the subjects) that it's roughly 50% 50% in environment vs genetics, but the fact that "genetics" can be used to justify our political order and worldwide not for 100% of humanity but for only a certain percentage of it, say the rich or whatever, like it's happening here in HN en masse and like you correctly said that " Throw in the casual attacks on public education at all levels, and implications that we should bow down to a Calvinist tyranny of genetics," and I really have to wonder what is going on here.". I'd be delighted to read an expanded response in regard to this subject by you :) Edit: I like to treat this particular subject as dealing with the development of nuclear bomb. You cannot just say "oh, I am just researching this for the sake of science" when you clearly know that once you have that nuclear bomb and can put it to use, you can change the whole world.