rupp-certified saladbar wrote:
Depends on the effect size you're looking for, etc. I think it's safe to say 12 (really???) is not adequate for this or any other RCT design.
There are dozens of calculators available for this, or you can do it by hand if it helps make the concepts more clear for you.
veg wrote:
So what effect size would you be looking for in a PED, and what sensitivity to detect such an effect would you require (i.e. power) in order to agree that an RCT to find a performance enhancing effect is adequately powered?
You tell me. What effect size would be relevant to you? 5%? 1%? 0.1%? Because I think even that could make a difference in long distance events, and there is no possibility you can reliably detect that with this tiny sample size.
Really, the authors should have a statement about exactly this, as well as their power/sample size calculations, but don't seem to. Pretty fundamental stuff.