In areas of inter-laboratory quality control, method comparisons, assay validation and individual bioequivalence, etc., the agreement between observations and target (reference) values is of interest. The mean of the squared difference between observations and target values (MSD) is a good measure of the total deviation. A new user-friendly statistic, the total deviation index (TDI(1-p)), is introduced that translates the MSD into an index that can be directly compared to a predetermined criterion. The TDI(1-p) describes a boundary such that a majority, 100(1-p) per cent, of the observations are within the boundary (measurement unit and/or per cent) from their target values. Statistical inference using the sample counter part (estimate) is presented. A Monte Carlo experiment with 5000 runs was performed to confirm the estimate's validity. Applications in laboratory performance and validation, as well as individual bioequivalence, are presented.
Copyright 2000 John Wiley & Sons, Ltd.