Fig. 1

Schematic depiction of the superiority, equivalence, and non-inferiority designs. The x-axis represents the true population effect size (\(\delta\)), where c is the standardized equivalence margin in case of the equivalence test and the standardized non-inferiority margin in case of the non-inferiority test. Gray regions mark the null hypotheses and white regions the alternative hypotheses. The region with the diagonal black lines is not used for the one-sided superiority design. Note that the diagram assumes that high values on the measure of interest represent superior or non-inferior values and that a one-sided test is used for the superiority design