sorry if there is no detectable h2 for EA from the sib regression, then what does the within family ancestry and between family ancestry correlation even mean here. if you found h2(like in height) then I can decompose that into direct genetic effects(within family) and between family(possible environment and genetic correlation) but a percent of zero is.. zero? i'm confused. then why is there a correlation between ancestry and EA at all? even if you are picking up on environmental noise, once you control for the right moderators you should not see any correlation between ancestry and EA in Mexico then. and what is modeling jointly here? shouldn't the between family eat up all the variance then? if there is just environmental correlation, then there should be no within family signal, but the wording here does not say that. i read this a long time ago i don't remember precisely sorry if it is addressed