Abstract
A candidate material for the production of certified reference materials (CRM) should be sufficiently homogenous and stable. General requirements and more specific guidance for assessment of homogeneity, stability, characterization and uncertainty evaluation are provided in ISO Standard 17034 and ISO Guide 35. This article provides a detailed case study of homogeneity and stability assessment. Two statistical methods has been selected to analyze experimental data. One of them is linear regression analysis used to test statistical significance of the linear trend. The other method is two-way nested analysis of variance used to estimate a combined uncertainty consisting of the between-unit and within-unit heterogeneity contributions.


