Sample size calculation for 3 arm 1:1:1 trial

[Kelvin Wan]

Hi,

I am trying to do a power calculation for a 1:1:1 randomized control trial. The outcome is a continuous biomarker.

I want to detect a difference of 0.5 between the 3 treatment groups, the common standard deviation within the group is 0.8.

May I have some advice on what the syntax in Stata would be? I have tried reading the manual on pss but I cannot find the syntax.

Thank you very much for your help!

[Rich Goldstein]

this sounds like a one-way anova so look at

help power oneway

so your command would look something like"

power oneway mean1 mean2 mean3, options

where you put in whatever options you want for your analysis (do you want n for a certain power or power for a certain n or ...?)