In the original paper we write: "One important choice a researcher must make in defining the equivalence margin is whether the margin should be defined on a raw or standardized scale."

We would like to correct this statement. Recall that hypotheses are statements about parameters and not about the observed data. In other words, hypotheses are strictly non-random. As such, it is incorrect to define the equivalence margin in terms of a function of the data (e.g., in terms of the estimated standard deviation, s_p). Instead, the correct choice is to define the parameter of interest to be on a standardized scale. While in practice, the difference between these two approaches may seem minuscule, it should nevertheless be acknowledged.