Modern analytical chemistry involves more and more the use of statistical, mathematical and computer tools, both during the methods validation and, successively, during the internal quality control (IQC). Although these activities are far-back mandatory for the official laboratories which are accredited to ISO/IEC 17025 international standard, in practice analysts are not always sufficiently familiar with the statistical process control (SPC) theory. Furthermore, especially for multiresidue methods, the implementation of an ICQ using single univariate Shewhart charts for each quality characteristic (analyte) of the process can be very misleading [1]. This happens because in practice these quality characteristics are correlated. For this reason and to improve the effectiveness of IQC procedures, multi-analyte methods should also be controlled by multivariate methods that consider the joint distribution of variables. From this perspective, the Hotelling T2 chart represents an extension of the usual Shewhart charts to the multivariate case, and it is the most popular tool used for monitoring multivariate processes. So far there have been few known applications of multivariate statistical control in analytical chemistry, probably due to the lack of adequate skills among chemists. The aim of this work is, therefore, to present a user friendly add-on package, called “izsqcc”, for the R statistical software.
Multivariate Charts for Quality Control of Multiresidue Analytical Methods
SCRUCCA, Luca;
2010
Abstract
Modern analytical chemistry involves more and more the use of statistical, mathematical and computer tools, both during the methods validation and, successively, during the internal quality control (IQC). Although these activities are far-back mandatory for the official laboratories which are accredited to ISO/IEC 17025 international standard, in practice analysts are not always sufficiently familiar with the statistical process control (SPC) theory. Furthermore, especially for multiresidue methods, the implementation of an ICQ using single univariate Shewhart charts for each quality characteristic (analyte) of the process can be very misleading [1]. This happens because in practice these quality characteristics are correlated. For this reason and to improve the effectiveness of IQC procedures, multi-analyte methods should also be controlled by multivariate methods that consider the joint distribution of variables. From this perspective, the Hotelling T2 chart represents an extension of the usual Shewhart charts to the multivariate case, and it is the most popular tool used for monitoring multivariate processes. So far there have been few known applications of multivariate statistical control in analytical chemistry, probably due to the lack of adequate skills among chemists. The aim of this work is, therefore, to present a user friendly add-on package, called “izsqcc”, for the R statistical software.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
