@_ChrisAlbertyn @richardaemsley @KCLBHI @Trial_Forge @MRCNIHRTMRP What do you mean by a cluster? Are the clusters randomised or are individuals randomised? If the trial is individually randomised this paper provides a good explanation of how to deal with
6. Comment: Adjust for clustering at a level above randomization Response: You don't need to do this because the errors can't be correlated if you randomized at a lower level. It's good news! https://t.co/kdsfZuZ5DQ
RT @Brennan_Kahan: This reaction is going straight on my CV. (and props for bringing back hot sauce) https://t.co/lCegb2sl9E
RT @Brennan_Kahan: This reaction is going straight on my CV. (and props for bringing back hot sauce) https://t.co/lCegb2sl9E