Excellent points! In light of this distinction, how might behavioral genetics studies inform interventions? Say a large scale twin study finds a negligible shared environmental effect, but a substantial genetic effect. Can we take nothing applicable from these findings?
-
@KirkegaardEmil Do you have any thoughts on this discussion? Am I totally misguided here? I am finding it hard to fathom (perhaps because it is discouraging) that we can learn nothing applicable from behavioral genetics research.
It's a common angle you will get from, say, Turkheimer and generally people who approach these things from a philosophical angle instead of practical-empirical. Environmentality (1 - H²) is a kind of crude index of the malleability of a trait, sure.
-
-
You can make predictions of likely efficacy of interventions based on variance analysis. If it targets environmental variation already known to not affect the trait, of course, intervention won't work (e.g. parenting reading to enhance adult offspring intelligence).
-
But if there's no or little variation in some environmental context and your intervention would produce a large increase here, then it could conceivable alter the trait of interest. Usual choice of example is PKU, which is caused by combination of genetic defect + specific diet.
-
In that case, almost no variation in the diet for relevant property (all diets contained phenylalanine. Intervention produced a large chance in that environment for the relevant population, and thus was not excluded by general high heritability of intelligence.
-
But if the intervention proposal is dietary advice for parents to alter BMI of kids, yeah, there's lots of variation in that already and it's not linked to actual BMI of kids (approx. no shared environment), so that intervention has very bleak prospects. Try something else first.
End of conversation
New conversation -
-
-
This is utterly false, even if 1-H2 represented environmental contributions (in humans it's really just a gross error term) there is no ground to say those components cause any part of the trait and zero support for any extrapolations on interventions. See https://www.jstor.org/stable/pdf/2529584.pdf?casa_token=uLg7Q8zFI1sAAAAA:Epa9P95HR8sqcyzwbGP9uHSgSahskVE0NbWMo2tw8sQwsWZ2WmChtZyg0WDg4ri_j2SLkhZ53EGsAd_j6jmxP_2edWm6tiwpFJAydGBOEdNlgobwNglD …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Curious on your explanation for why Turkheimer’s (
@ent3c) approach is, according to you, less empirical than others working in the field. One of the points he makes is that the CNS interposes btwn genotype & phenotype for complex behavioral traits. That’s an empirical position -
Going back to the original post, the biggest problem is that it isn't only plausible targets (schizophrenia, low IQ) that are heritable. So is divorce and how hot you like your tea. So the argument boils down to a bland assertion that we should look at the "biology" of everything
- 3 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.