Role of statistical methods in qualitative control of analyses

It seems an evidence nowadays that a statistical approach is quite necessary in modern analytical chemistry both for ata processing and design of experiments,in spite of:technical progress, which has enabled full automation :of the whole analysis-sequences, differences are often observed between repeatability and reproducibility which are accurately defined. Any rig; rigorous reproducibility study requires. It:implementation: of a collaborative study and the the general technique used for data evaluation is the analyses of variance (ANOVA)I In its,one-way version the influence of the laboratory :factor can be studied: globally and the Variance:due to the between-laboratories source of-variation estimated if it-is found significantly different from zero. In order to minimize any dispersive effect, it is necessary lo identify:among the sources of error involved, Which ones have a significant: influence: ANOVA is again the best method of investigator. AS in quantitative determinations the true value is unknown, the challenge is to. give the best estimation through a confidence interval. In data exploitation, he problem arise of the discrimination between random errors and systematic errors:and their propagation. The greatest difficulty in accuracy studies is encountered When no standards nor certified reference materials are available; In: this case, only indirect methods can be carried out.