August 6, 2019

I didn’t read The Book of Why, but I heard of it a lot on Twitter lately. As you may know, I’m a big fan of Stephen Senn’s work, and I keep following his posts here and there even if I’m no longer being involved in medical statistics. Here is another nice dicussion of Lord’s paradox, where the author explains why neglecting random effects may affect the conclusions drawn from a study. If you’re interested in causal modeling of pre-post study, take a further look at this recent paper: Causal Graphical Views of Fixed Effects and Random Effects Models.