Is Correlation equal to causation?

[Anonymous No.15414863]

>The following genes are known to increase intelligence
>The following genes are known to decrease intelligence
How would you even test for something like this to discover that the connection is related? Seems like starting out with a theory and working in reverse to confirm it.

[Anonymous No.15414866]

correlation isn't causation, but it is consensus

[Anonymous No.15415458]

>>15414863
>Is Correlation equal to causation?
If and only if it supports /pol/tard ideas.

[Anonymous No.15416082]

>>15414863
You need an experiment. I want to know whether these genes (X) affect intelligence (Y). Then I need some kind of exogenous variation that allows me to isolate it from other shit that affects intelligence.
I don't know which identification strategy would work best here, maybe you could look at twins? Some twins are identical, some are not, so can compare them or something.

That's how people usually jump from correlation to causation, by exploiting an exogenous variation in the explanatory variable.

[Anonymous No.15416104]

>>15416082
>You need an experiment. I want to know whether these genes (X) affect intelligence (Y). Then I need some kind of exogenous variation that allows me to isolate it from other shit that affects intelligence.
Researchers have already been doing that for quite a while and the resultant controversy over nonreproducibility from GWAS due to correlations disappearing when assessing causality has even lead whole journals to greatly upping their standards. https://en.wikipedia.org/wiki/Missing_heritability_problem

Candidate SNPs almost always are false positives with a very low replication rate. Individual SNP effect sizes tend to, if replicable, be fractional of fractional percentages. Which is why OP's image is completely useless and uninformative.

https://www.nature.com/articles/s41588-018
-0147-3
>A joint (multi-phenotype) analysis of educational attainment and three related cognitive phenotypes generates polygenic scores that explain 11–13% of the variance in educational attainment and 7–10% of the variance in cognitive performance.
>we identify 10independent genome-wide-significant SNPs and estimate a SNP heritability of around 0.3% in both men and women
This did not really improve in the subsequent paper https://www.nature.com/articles/s41588-022
-01016-z
> In our updated X-chromosome GWAS, we increase the number of approximately uncorrelated genome-wide-significant SNPs from 10 to 57. Our dominance GWAS identifies no genome-wide-significant SNPs. Moreover, with high confidence, we can rule out the existence of any common SNPs whose dominance effects explain more than a negligible fraction of the variance in EA.
>Power calculations indicate that, at genome-wide significance, we had 80% power to detect dominance effects with an R2 of 0.0015% (Supplementary Note). Such effect sizes would be over an order of magnitude smaller than the largest additive effects (R2 ≅ 0.04%).
tl;dr OP is full of shit.