## Power in case of a 3-arm BE study [Power / Sample Size]

Dear Louis,

» For a 3 arm (Ref, Test1, Test2) study what is the power adjustment necessary in order to get an overall power of 90%. Let's say by using a Bonferroni method. Can it be done directly in PowerTost? This is in order to find the 'right' sample size.

Have a look at this post and links there in.
And use the search button to find more discussions about that topic.

The need of an alpha adjustment (f.i. by Bonferroni method) is discussed in this post. It depends on your overall hypothesis, as described in the mentioned discussions.

If such an alpha-adjustment is necessary it can be done quite easily in PowerTOST. Just change the alpha argument in the function sampleN.TOST() as needed.

Regards,

Detlew