Stability criteria for the outcomes of statistical tests to assess drug effectiveness with a single study.

Conservative sample size estimation in nonparametrics.

Adapting by calibration the sample size of a phase III trial on the basis of phase II data.