REAPPRAISAL OF HYPOTHESIS-TESTING FOR METHOD VALIDATION - DETECTION OF SYSTEMATIC-ERROR BY COMPARING THE MEANS OF 2 METHODS OR OF 2 LABORATORIES

The usual way of testing the equivalence of two measurement procedures is analyzed, and possible alternatives are discussed. It is concluded that for method validation purposes, defining an interval of acceptable values and applying interval hypothesis tests is to be preferred over the usually applied point hypothesis tests. Statistically testing against acceptance limits has the advantage that the beta-error of the usual null hypothesis (there is no bias) can be limited. It is shown that significance testing at a fixed beta-error requires, however, reversal of the usual null and alternative hypotheses. Equations are presented for the case in which the acceptable bias is expressed as a ratio of the reference mean. Moreover, the principle of randomization tests is explained, and their possible advantages over the ordinary test statistics in method validation are discussed.