Posting: # 19249
Is it possible to design the reference scaled approach following parallel study (single period)?
If yes, how to design the study, randomization and statistical procedures?
Posting: # 19250
» Is it possible to design the reference scaled approach following parallel study (single period)?
By definition reference-scaling is applicable if the within-subject variability of the reference is >30%. CVwR is not accessible in a parallel design (not even CVw…).
» If yes, how to design the study, randomization and statistical procedures?
Tóthfalusi and Endrényi proposed a method for studies in a parallel design.1 However, in a follow-up paper2 they didn’t mention parallel designs any more. Given the troubles we already are facing with replicate designs (inflation of the Type I Error, convergence issues in the partial replicate) I guess the chance of a rather exotic approach to be accepted by any agency is practically zero.
The quality of responses received is directly proportional to the quality of the question asked. ☼
Posting: # 19252
Thank You! HS