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. 1. Exploratory Analysis of Univariate Data

M. Meloun and J. Militký

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

Abstract: The first step of univariate data analysis called an exploratory data analysis (EDA) isolates certain basic statistical features and patterns of data. EDA is based on the general assumptions as a continuity and differentiability of underlying density. For visualization of data the quantile plot, the dot and jittered dot diagrams, and the box-and-whisker plot are proposed. Peculiarities of sample distribution are investigated by the midsum plot, the symmetry plot, the curtosis plot, and the quantile-box plot. Construction of sample probability density function is carried out by the kernel estimation and the histogram. The quantile-quantile plot serves for comparison of sample distribution with selected theoretical ones. EDA is illustrated on a quantitative chemical analysis of phosphorus content in blood.

Full paper in Portable Document Format: 483a151.pdf

Chemical Papers 48 (3) 151–157 (1994)