One Degree wrote:Wow! I am surprised it even got published. The media will have a field day trashing it. Too many people will be unwilling to accept it even if it is true. It definitely needs serious review.
The media tends to ignore publications like this, but if they are reported it obviously must be done by stressing the negative historical context and how dangerous this information can be. The only non-science article I could find does exactly that: Mindful of eugenics’ dark history, researchers are reexamining the genetics of social mobility
As you say, and also mentioned in the article, many are unwilling to even entertain the thought and education in particular is a stronghold of attributing everything almost exclusively to nurture.
mikema63 wrote:I'm somewhat skeptical. Educational attainment is such a complex social thing that's influenced by so many personality traits and social stuff that I find the idea of trying to do genetic analysis of it to be rather...ambitious in scope. More ambitious than trying to research intelligence broadly even.
I'd like to see studies on the more basic personality traits that coorespond to educational achievement on a genetic level.
With respect to twin studies the claim is quite in line with previous research. There is a tool here
which is based on a recent effort to provide access to more than 2,700 twin studies that have been carried out during the last 50 years (link to paper explaining the details
). If you select ICF/ICD10 Subchapter on the left and Education on the right dropdown field, it'll show you this:
The most interesting graph is that on the right which shows the average of the relative contribution of heritability (h2_all) and shared environment (c2_all) as reported by 31 twin studies that have looked at educational attainment.
Results for higher-level cognitive functions, which include various IQ test batteries and are based on 660 studies, are similar:
For temperament and personality functions h2 is lower but still substantial:
That said, identifying the genes to validate these results has turned out to be a formidable challenge. Under the current assumption that thousands of genes individually contribute very small effects (<1%), we would need samples of millions of genotyped people to detect all of these genes, and samples of this size are currently not available. Further, most studies that try to detect gene associations look at common SNPs (present in >1% or 2% of the population), yet not only are the vast majority of SNPs less common, but SNPs are just one type of genetic variation. It's therefore perhaps not surprising that reported heritability based on identified genes tends to be quite underwhelming, especially when compared with the estimates derived from twin studies. It's no different in the study mentioned in the OP
Second, we used a new GPS that explains twice as much variance in years of education as the GPS (Rietveld et al., 2013) used in the previous studies. This new GPS, derived from a 2016 GWA analysis of years of education in a sample of nearly 300,000 individuals, explains 3.9% of the variance in years of education in independent samples (Okbay et al., 2016). Finally, in addition to conducting DNA analyses, we used twin data to obtain the first well-powered estimate of the degree to which intergenerational educational attainment is heritable.
Twin ACE analyses
Figure 1 shows the proportion of variance in intergenerational educational attainment that was estimated to be accounted for by genetic and environmental factors. The graphs indicate a substantial genetic influence, as approximately half of the phenotypic variance in liability was attributed to inherited DNA differences in both the upward-mobility analysis and the downwardmobility analysis. The influence of shared environmental factors (i.e., factors that contribute to similarities between twins growing up in the same home) was almost as large, accounting for approximately 40% of the variance. Nonshared environmental factors (i.e., factors that do not contribute to twin similarity) explained less than 11% of the variance in both analyses.
So more than 40% of the heritability still needs to be accounted for by actual genes. Whole genome sequencing, which makes it possible to look at other variants and pretty much the whole genome, is now possible and affordable. There is currently a project underway, involving several universities, that will conduct WGS for the five major psychiatric illnesses. And WGS will quite likely propel us forward also in terms of behaviour and traits.
Wellsy wrote:Think this is the paper
The Genetics of Success: How SingleNucleotide Polymorphisms Associated With Educational Attainment Relate to Life-Course Development
Thanks, Wellsy, but that is a different paper.
Dave wrote:50% is probably an underestimate given that IQ appears to be over 80% heritable.
And Gregory Clark in his work has found a level of social mobility of around one-fifth going back a thousand years in nearly every society he's studied bar India (which has clos to zero social mobility ).
Not that actual evidence will change the left's view on this subject. They are religious fanatics.
80% is at the high end of IQ estimates commonly cited and correlation with education is somewhere around 0.7 if I remember right.
The field of education will be the hardest nut to crack if twin studies are confirmed in the future. But even if opposition to this type of research and thinking increases, advances coming from medicine will not be so easy to ignore. And regardless of how successful people in the west will be in delaying it, others don't have the same kind of hang-ups.