Evaluation of the order of hierarchical structures for the calculation of method uncertainty using nested and other...

Several methods have been used to calculate method uncertainties using data from a previous publication. The major effort was to study the effect of changing the primary factor of the nested (hierarchical) design from concentration to matrix and then analyst. In the measurement of fluoroquinolones in honey, it was observed that when matrix was the primary factor the uncertainty values were generally slightly larger than when concentration or analyst were the primary factors. However, the differences are small, indicating the robustness of the nested design calculations. Two previously published methods of correcting for uncertainty due to intermediate precision were also examined and showed very little difference, confirming the robustness of the correction. A comparison was also made with previous non-nested designs, which tended to give lower calculated error values.