Robustness and Corrections for Sample Size Adaptation Strategies Based on Effect Size Estimation

The robustness of the adaptation of the sample size for a phase III trial based on phase II data is studied -when phase III is lower than phase II effect size. Conservative sample size estimation strategies are compared. When the rate between phase III and phase II effect size is greater than 0.8, Calibrated Optimal Strategy provides acceptable results. A correction for balancing the structural bias is introduced, based on a postulation of the bias. When the postulated correction is right, or even smaller than necessary, COS works well. A higher than necessary correction should be avoided. Note: The following files were submitted by the author for peer review, but cannot be converted to PDF. You must view these files (e.g. movies) online. cos_sse_diff_scen3.tex URL: http://mc.manuscriptcentral.com/lssp E-mail: [email protected] Communications in Statistics Simulation and Computation