ISSN print edition: 0366-6352 ISSN electronic edition: 1336-9075 Registr. No.: MK SR 9/7 Published monthly

Computer-Assisted Data Treatment in Analytical Chemometrics. 3. Data Transformation

M. Meloun and J. Militký

Department of Analytical Chemistry, Faculty of Chemical Technology, University Pardubice, CZ-532 10 Pardubice

Abstract: In trace analysis an exploratory data analysis (EDA) often finds that the sample distribution is systematically skewed or does not prove a sample homogeneity. Under such circumstances the original data should often be transformed. The power simple transformation and the Box-Cox transformation improves a sample symmetry and also makes stabilization of variance. The Hines-Hines selection graph and the plot of logarithm of the maximum likelihood function enables to find an optimum transformation parameter. Procedure of data transformation in the univariate data analysis is illustrated on quantitative determination of copper traces in kaolin raw.

Full paper in Portable Document Format: 483a164.pdf

Chemical Papers 48 (3) 164–169 (1994)