Abstract
Stevenson has claimed to resolve the renormalization-prescription ambiguity inherent in a truncated perturbative expansion by a principle of minimal sensitivity. This is shown to define an αs such that its coefficients in the perturbative expansion are in the ratio of the corresponding coefficients of the β function, regardless of the physical quantity calculated. Stevenson's principle is then only relevant when the expansion of the β function for the optimal αs makes sense and is then negligibly different from the commonly assumed criterion of fastest apparent convergence, with no greater claim to "optimality."