A procedure for estimating fundamental variability [version 1; peer review: awaiting peer review]

Background: The question whether a given set of test items can be considered “identical” is often addressed in terms of the homogeneity of the test material from which said items were taken. However, for some types of matrices – in particular, for matrices consisting of minute separate particles, only some of which carry the analyte under consideration – even in the case of homogenous test material, an irreducible source of variability between test items may remain: the fundamental variability. Methods: In this paper, the concept of fundamental variability is explained, and procedures for reducing and characterizing it are described. Results: The proposed procedure allows both the detection and quantification of the fundamental variability. In addition, it is shown how fundamental variability cannot be identified in homogeneity testing according to Annex B.1.1 in ISO 13528. Conclusions: The fundamental variability can be brought under control by increasing sample size and by finer grinding of samples. Both of these measures are tantamount to increasing the average number of carrier particles per subsample. The procedure proposed in this paper for characterizing the fundamental variability requires a high number of individual analyses. It therefore makes practical sense to link the proposed procedure with suitable model assumptions in order to then arrive at an estimate of fundamental variability in routine applications with much less effort. The case of heterogenous samples lies beyond the scope of ISO 5725. The question whether irreducible sample heterogeneity should be considered an actual component of an analytical method’s precision has perhaps not yet received sufficient attention. One possible approach would be to distinguish method precision per se (laboratory bias and repeatability precision) from matrix heterogeneity, and to provide information regarding both. Certainly, if the fundamental variability is non-negligible, then it must be accounted for in any evaluation of measurement uncertainty.