In this paper the quantitative distribution of the quality feature "calcium" in mixed feed for laying hens is dealt with. The magnitudes of the partial error complexes "measuring technique", "materials", "man" and "machine" which affect the total error have been determined using methods of mathematical statistics. They are discussed in terms of the total error. In the example under investigation, the contribution (presented in the form of the variance) of the measuring technique to the total error is almost 5%; that of the raw materials, nearly 10%. Amounting to more than 80%, the partial error complex "machine/man" exerts a dominant effect on total variance.

No abstract

No abstract

a b a Zentralstelle für Futtermittelprüfung und Fütterung Halle-Lettin b Amt für Standardisierung, Meßwesen und Warenprüfung Fachabteilung Nahrungsgüter Fachgebiet Körnerfrüehte und Futtermittel , Halle

Model investigations into error dimension and error distribution with regard to piglet rearing feed show that the sampling regulations for mixed feed as laid down in TGL 29080/01 of the GDR basically guarantee the representativeness of the samples as to the assessment of lots. Based on the total scatter and the confidence intervals, the minimum and maximum contents laid down in the standards of quality can be observed. In comparison with pellets, the total scatter particularly for raw ash and the minerals Ca, Na, Mn and Cu is bigger when the mixed feed is provided in the form of coarse meal, which is due to high partial errors in sampling and the prepartion of sampling (partition). The quota of analysis errors in the total error is significant for parameters with relatively low content values (crude fibre, crude fat) and vitamin A in contrast to dry matter, crude protein, raw ash, bulk elements and starch so that double or triple determinations (for vitamin A) are appropriate and to be taken into consideration. Results with regard to vitamins and trace elements make a first survey possible.