Introduction to Statistical Methods for Outlier Detection and Sample Homogeneity Assessment of Reference Materials and Proficiency Test Items

Abstract

Due to the homogeneity of the product or sample, it will affect whether it meets the scope of application and purpose. For example, the reference materials(RM) produced by the reference material producer(RMP), and the proficiency test items selected by the proficiency testing provider(PTP), in order to ensure the reference materials or proficiency test items have consistent characteristics or comparability, they should be proved to have certain homogeneity. However, before performing homogeneity assessment, it is necessary to measure the characteristic parameters of the reference materials or proficiency test items to obtain a sufficient number of measured values for data analysis, but there may be outliers in the measured values that may affect data analysis and interpretation of the results. Therefore, this article will refer to ASTM E178-16a:2016[1], ISO 5725-2:1994[2], ISO 13528:2015[3], etc., to introduce several outlier detection and homogeneity assessment methods, supplemented by case studies. Finally, this article will remind the precautions for the use of the method, so that readers can choose the appropriate method for use in the actual analysis.