Limitations of the Application of the Horwitz Equation

The basic assumptions of the Horwitz equation are revisited using the example of mycotoxin assays. Prediction intervals from the Horwitz equation often span one order of magnitude and including this variation in the calculation of the Horrat values would lead to a range of values including those exceeding 2, which is often used as criterion to assess interlaboratory comparisons, questioning the suitability of the Horrat value for this purpose. In addition, available analytical data shows significant improvement in reliability over time, which casts serious doubts on the applicability of the Horwitz equation for current analytical methods. The use of the Horwitz equation in the analytical laboratory is discussed, and it is concluded that it is not suitable for estimating uncertainties as required by ISO 17025. The Horwitz equation can be a valuable summary of historical data of analytical performance. However, it should not be used as a performance criterion due to: the shortcomings of the basic model, the uncertainty in the values determined using the equation, and its incompatibility with accepted methods for the determination of measurement uncertainty as required by ISO 17025. Instead of using the Horwitz equation, a proper identification of all components of uncertainty of measurement and reasonable estimation as stipulated by the “Guide to the Expression of Uncertainty in Measurements” (GUM) is recommended.

LINSINGER Thomas;  JOSEPHS Ralf; 

2006-12-05

ELSEVIER SCIENCE LONDON

JRC33981