Bioequivalence and Bioavailability Forum

Dear mittyri!

Many thanks! I've been using PowerTOST for quite long time but haven't noticed the simulation functions. I must've been

You are right, the between subject variability is effectively zero in my original code. And I also asked chatGPT, it did provide quite detailed answer. It seems that AI is getting better every day.

Your mention of chatGPT reminds me that we still haven't been able to reproduce the FDA's unscaled ABE using the following SAS code, which was also summarised in one of Helmut's post. Maybe we can ask chatGPT to come up with R code to reproduce the result from SAS? I'll play it when I got some time and report back here.
PROC MIXED;                                 
  CLASSES SEQ SUBJ PER TRT;                   
  MODEL Y = SEQ PER TRT/ DDFM = SATTERTH;     
  RANDOM TRT/TYPE = FA0(2) SUB = SUBJ G;     
  REPEATED/GRP=TRT SUB = SUBJ;               
  ESTIMATE 'T vs. R' TRT 1 -1/CL ALPHA = 0.1;