The reliability of chemical analysis is discussed in a general way, stress being laid on the necessity of using the methods of mathematical statistics in the design as well as in the evaluation of experiments (factorial design; analysis of variance; randomisation in time; control charts), carried out to develop analytical methods and procedures, in order to obtain reliable results in the application of the methods in practical analysis.
Get full access to this article
View all access options for this article.
References
1.
COHEN — Rationelle Metallanalyse, p. 147. Basles, 1948. Chemische Analysen-Methoden (Schieds-Analysen), Aluminium Zentrale, Berlin, p. 35. C. I. M. O.; non published procedure.
2.
The greater part of these methods have been borrowed from the book of HAYWOOD and WOOD: Metallurgical Analysis with the Spekker Absorptiometer, London, 1944.
3.
The coefficient of variation is defined by the relation—coeff. of variation = 100 (standard deviation)/average
4.
HackerW.ZimmermannA.RachmannH.; Z. Anal. Chem.129 (1949) 104.
5.
DaviesO. L. — Statistical Methods in Research and Production, p. 211, London1948.
