Statistical Evaluation of Spectrographic Methods

Abstract

By using some of the excellent background material available from the literature, the tool of the statistically designed experiment is applied in the development and evaluation of spectrographic methods A preliminary design is carried through to determine the major factors affecting a method's total variability Further designs then study these gross variables in greater detail Eventually a reliable method, of known precision and bias, is developed, and the method is ready for routine use During the routine use, a quality control program is carried on to maintain a check on the results and yield a realistic estimate of variability over a period of time The paper gives a bibliography of books and articles that the author has found particularly useful