## Blind 2Stage and an elephant [Two-Stage / GS Designs]

Dear Helmut!

Thank you for pointing on the article! I've tried to draw a graph similar to Figure 1 of Golkowski et al. by using power.tsd.ssr. It uses CV instead of N of the second size, but the tendency is pretty similar: low sample size of the first stage can significantly enlarge the TIE.

» Simulations… Might be tricky cause we have to think about unequal group sizes and/or variances.

Is it possible to do it with Power2Stage or some modifications? (As I understand power.tsd.p deals with unblind scheme?) The co-authors of the aforementioned article have also an article dedicated to the parlallel groups but relatively to non-inferiority trials...

"Being in minority, even a minority of one, did not make you mad"