means
Equivalence test (TOST)
Test whether a two-group mean difference is small enough to be practically equivalent.
Requirements
- numeric outcome
- two independent groups
- a prespecified meaningful margin
Avoid when
- the margin was chosen after seeing results
- paired observations
Formula
H₀: Δ ≤ −δ or Δ ≥ δ; H₁: −δ < Δ < δ
Interpretation
Equivalence is supported only when both one-sided tests exclude effects beyond the lower and upper margins.
Worked example
Test whether two formulations differ by less than a prespecified clinically meaningful margin.
Open example datasets