Evaluation of risks in conformity assessment of multicomponent materials or objects

Francesca Pennecchi, Quality of Life Division, Istituto Nazionale di Ricerca Metrologica (INRIM), Torino, Italy (f.pennecchi@inrim.it)
Ilya Kuselman, Independent Consultant On Metrology, Modiin, Israel
Ricardo J. N. B. da Silva, Centro De Química Estrutural, Faculdade De Ciências Da Universidade De Lisboa, Lisboa, Portugal
Brynn Hibbert, School Of Chemistry, Unsw Sydney, Sydney, Australia

JCGM 106:2012 offers a Bayesian approach for evaluation of risks of false decisions in conformity assessment of a single item property taking into account its associated measurement uncertainty. The probability of accepting the item when it should have been rejected is named ‘consumer’s risk’, whereas the probability of falsely rejecting the item is the ‘producer’s risk’. When multicomponent materials such as medications, alloys or environmental compartments undergo conformity assessment, the approach based on a single property evaluation is not complete. Even when the assessment of each component of the material batch or lot is successful, the total probability of a false decision concerning the batch or lot as a whole (total consumer’s risk or total producer’s risk) might still be significant. In IUPAC Project 2016-007-1-500, Bayesian models for total risk evaluation are formulated for multicomponent materials, for independent and for correlated component concentrations of the material and corresponding measurement results (variables). In the former case, the total risks can be evaluated as appropriate combinations of the particular risks (i.e. those related to each particular component). In the latter case, modelling variables by multivariate prior probability densities and likelihood functions is required in order to obtain a corresponding multivariate posterior distribution from which the risks can be calculated. Analytical results of considered examples are obtained in the R programming environment. User-friendly MS-Excel programs are also developed for quantification of the risks. A comprehensive IUPAC/CITAC Guide on the topic is under development.

---

Short Biography of Presenting Author

Dr. Francesca Romana Pennecchi, PhD in “Metrology: measurement science and technique”, is a researcher at the Istituto Nazionale di Ricerca Metrologica (INRIM, Italy). Her activity is  focused on the development of mathematical and statistical methods for evaluating measurement uncertainty, analysing interlaboratory comparisons data and modelling risk of false decisions in conformity  assessment. She is committee member of  CITAC, Eurachem/CITAC WG on Measurement Uncertainty and Traceability, ISO/TC 69/SC 6/WG 7 on Statistical methods to support measurement uncertainty evaluation, and participated into several IUPAC, EMRP, EMPIR Projects. She is co-author of about 60 papers published on JCR journals, with h-index 12.