Loading…

Some pitfalls in the uncertainty evaluation of isotope delta reference materials

The Guide to the Expression of Uncertainty in Measurement (GUM) provides a framework for evaluating analytical data and characterizing their dispersion in a consistent manner. This is of eminent importance in the case of reference materials and their recommended values that are used for calibration...

Full description

Saved in:
Bibliographic Details
Published in:Accreditation and quality assurance 2023-06, Vol.28 (3), p.101-114
Main Author: Gröning, Manfred
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The Guide to the Expression of Uncertainty in Measurement (GUM) provides a framework for evaluating analytical data and characterizing their dispersion in a consistent manner. This is of eminent importance in the case of reference materials and their recommended values that are used for calibration of further measurements. The proper propagation of uncertainties for those data is essential. Guidance is provided in the GUM on how to calculate the combined standard uncertainty for a mean value or central value based on multiple individual measurements including their calibration uncertainty. However, according to published data, the guidance provided by GUM is not always properly applied in practice. Several published studies show calculated uncertainties much lower than those of input quantities. This may be caused by improper handling of the propagation of uncertainty components, thereby breaking the traceability chain for reported values. A simple check method using conventional statistical means is proposed to detect and to avoid such mistakes related to calibration uncertainties. It is intended to help to ensure a proper uncertainty propagation, to maintain a system of reliable and traceable reference materials. A spreadsheet is provided for the implementation of such a check. Example calculations for published data are presented.
ISSN:0949-1775
1432-0517
DOI:10.1007/s00769-022-01527-6