Item

Abstract

Before the introduction of the International Association of Geoanalysts’ (IAG) Protocol (Kane et al. 2003), characterisation programmes for geological reference materials were often organised such that laboratories were invited to volunteer to contribute data without having to first demonstrate that the quality of their data was adequate for the purpose of certification. The specific issue of relevance is that reference material characterisation programmes require “state of the art” data quality when the resultant reference material will be used for calibration, to assess “state of the practice” laboratory performance, and/or to demonstrate the traceability of a laboratory’s measurements. Rather frequently, the resulting reference values were stated without uncertainties that are now a requirement for traceability purposes. Additionally, had uncertainties been developed, they would have been of similar magnitude to those achieved by typical laboratories or even larger, rather than significantly smaller, as is required in order to effectively transfer accuracy from certification laboratories to others (Uriano and Gravatt 1977).