I have a given experimental design for a clustered RCT, and I am using iterations of the code below to determine the MDE for my given design. I'd like to compare some treatment arms with different numbers of observations per cluster (e.g. 10 per cluster in one treatment arm and 20 per cluster in control arm). . Any advice on how to do this in stata?
Code:
sampsi `mean_`x'' `mean2_`z'_`x'_1', sd1(`sd_`x'') ratio(`gratio0') power(`power0') alpha(`alpha0') pre(1) post(`post0') r01(`autocor') r1(`autocor') method(ancova)
mysampclus, obsclus(`obspclus`z'') rho(`icc_`x'')