Fig. 12

Differences between precise frequentist hypothesis testing and equivalence testing: Standard NHST for a sharp point null hypothesis H0:θ=0 against its alternative H1:θ≠0 (top left) or, in general, of H0:θ=θ0 against H1:θ≠θ0 (top right); TOST procedure for testing H0:θ<δL orθ>δU (shown in red) against H1:δL≤θ≤δU (shown in blue) (bottom left) or H0:θ<θ0−δL orθ>θ0+δU (shown in red) against H1:θ0−δL≤θ≤θ0+δU (shown in blue) (bottom right)