Talk:Rs7816345
From SNPedia
Effect[edit]
- 0.85x breast size (smaller breast size)
Aren't the effect sizes from the GWAS being reported wrong here? A multiplier like '0.85x' implies that it's a multiplicative increase of breast size or perhaps some sort of odds ratio for a binary measurement like "large vs small breast size", but the GWAS does its analysis on cup sizes. Which is small both absolutely, since the beta is 0.15 (cups), and relatively, because the mean/SD of cups reported is 5(1.45) so the d is only 0.10. Whereas if it was multiplicative as implied, instead of increasing/decreasing by a tenth of a cup, it'd increase/decrease by almost a whole cup size (0.15*5=0.75), an effect 7.5x larger... --Gwern (talk) 19:39, 22 September 2018 (UTC)
