The influence of uncertainties of measurements in laboratory performance evaluation using an intercomparison program of radionuclide assays in environmental samples

To show the influence of measurement uncertainties in performance evaluation of laboratories, data from 42 comparison runs were evaluated using two statistical criteria. The normalized standard deviation, D, used by US EPA, that mainly takes into account the accuracy, and the normalized deviation, E, that includes the individual laboratory uncertainty used for performance evaluation in the key-comparisons by BIPM. The results show that data evaluated by the different criteria give a significant deviation of laboratory performance in each radionuclide assay when we analyse a large quantity of data.