Precision Parameters of Standard Methods of Analysis for Dairy Products

Abstract

The available collaborative studies for standard methods of analysis for various constituents of milk and milk products were examined in an attempt to assign specific repeatability and reproducibility precision parameters to these methods. The different collaborative assays for the primary constituents (moisture/solids, fat, protein), the nutritionally important elements (calcium, sodium, potassium, phosphorus), and miscellaneous analytes/physical constants (ash, lactose, salt, freezing point) produced different estimates of the precision parameters for the same method. A suitable summary of the precision estimates from collaborative studies is given by the reproducibility relative standard deviation, RSDg, which is relatively constant within a product and permits comparisons across products. An estimate of the variation of RSDR for an analyte from a number of collaborative studies is presented in terms of the median and 90% interval (the range of the centermost 90% of values). These estimates are only informative when a substantial number of independent studies are available for pooling the independent estimates to form a distribution of RSDR values. The RSDR for the determination of the primary constituents of milk and milk products is characterized by a median RSDR of 1% and a 90% interval of 0.3-3%, with RSDR estimates occasionally occurring below 0.3% and above 4%. These overall estimates appear to be independent of analyte, matrix, and method and apply to concentrations of primary constituents that range from about 2 to 80%. The repeatability relative standard deviation, RSDr, is unstable, although it tends to converge to about 0.5-0.7 X RSDR. Too few collaborative assays are available to characterize RSDR for the determination of certain other constituents (acidity, ash, lactose, salt, and the nutritionally important elements) unless RSDR values for different analytes, methods, and matrixes are pooled on the basis of similar analyte concentrations. When pooled, the RSDR values are generally better than predicted from the Horwitz equation, RSDR (%) = 2 exp (1-0.5 log10C), where C is the concentration expressed as a decimal fraction; all but one of 661 RSDR values are within the upper empirical limit of twice this curve.