ABSTRACT
Analytic sample size formulas have been developed for a range of standard study designs and a variety of software packages for sample size and power analysis are available, both commercially and freely distributed. When closed-form expressions for sample size exist, calculations can be quickly performed and easily verified by the study reviewers. In practice, however, study designs can be quite sophisticated and may not be suitable for implementing any of the existing sample size formulas. Sample size and power analysis for stepped wedge designs is still a very active area of research. Simulations can be used to verify that a study designed using existing sample size formula would have adequate type 1 error and planned statistical power or target precision. There are several reasons why simulation studies can be an important and powerful tool for every practicing statistician: flexibility, intuitive approach, ability to answer other study-related questions and ease of implementation.
