Equivalence Testing

In some studies, a researcher wishes to establish that two treatments are not different from one another. Commonly, one treatment is the “gold standard” and another is an alternate treatment that is cheaper, safer or less invasive. It is not enough to conduct an experiment that fails to reject the null hypothesis that the two treatments are equal- this is easy to achieve with a low-powered study.

Instead, a researcher must perform an equivalence test. Here, the null hypothesis is that the difference in treatment is *larger* than delta, or “margin of tolerable difference”. This value is chosen ahead of time by the researcher based on practical implications specific to the research question. The alternative hypothesis is that the difference is less than delta. In this case, you need a sufficiently powerful study to reject the null in favor of the alternative hypothesis to demonstrate that two treatments can be considered “equivalent.”

This workshop will introduce equivalence testing, including non-superiority and non-inferiority testing. We will cover the TOST procedure (“two one-sided tests”) and contrast it to the conclusions that can be drawn from more standard null hypothesis testing.