@anne_vanH I think 'learning by doing' generally works best, keeps you motivated! Personally I prefer the tidyverse approach to R over the base R approach and am a big fan of the R 4 Data Science book for that: https://t.co/YYcoNW5HJP
Very interesting analysis in IPD of n>8000 randomized to SSRI or placebo, finding that item-level analyses change the story of whether the ADM>placebo advantage is moderated by baseline severity