ABSTRACT
In clinical research and development, the method of hypothesis testing is usually employed to test for the treatment effect of a test treatment under investigation. The purpose is not only to detect whether there is a treatment effect that is of clinical importance but also to assure that the detected treatment effect (or observed difference) is of statistical meaning (i.e., it has achieved statistical significance) in the sense that it is not by chance alone and is reproducible under the same experimental conditions. Commonly employed hypothesis testing includes testing for equality, testing for noninferiority, testing for superiority, and testing for equivalence/similarity. Under a given, valid study design, sample sizes required for achieving the study objectives are different depending upon the hypotheses (i.e., equality, non-inferiority, superiority, and equivalence/similarity hypotheses) to be tested. In the following, whenever equivalence will be discussed, the similarity of two products will be equally considered.
