A long-term, envisioned, objective of the Customs laboratories (CLEN), is the acquisition of harmonization of tasks and laboratory practices and procedures. One such task was the enrichment of the CLEN methods with their performance characteristics (PChs).
The aim of this investigation was to define a valid statistical estimation procedure of the (PChs) (Repeatability and Reproducibility limits) from existing Proficiency Testings’ (PT) data, whose implementation could become the basis for all the other CLEN methods.
The procedure was based on the estimation of the in-between reproducibility, considered to represent the systematic error of all the variations between and within laboratory, promoting thus Harmonization.
There was no rejection of any data and hence, there was no need to follow any statistical rejection rules. The effectiveness and validity of the results was assessed and supported by the conformance to the AOAC minimum analytical requirements and the verification to the comparison with standard methods.
The estimation and evaluation of data for the Butyric Methyl Esters (BAME) and Total fat in biscuits and chocolates, were found to be for BAME at concentration range (1.7–4.0)%m/m, repeatability (r) 0.14 and Reproducibility (R) 0.38 and Total fat at concentration range (10–20)%, repeatability (r) 0.28 and Reproducibility (R) 0.84 and for Total fat at concentration range (20–30)%, repeatability (r) 0.48 and Reproducibility (R) 1.07,though, in the milk products of high protein content, for (BAME) and Total Fat below 1.5%, methods, gave not conforming results. It is suggested that implementation of this estimation procedure will be of great benefit for CLEN laboratories, not only for the method acceptance, but for further investigation in different matrices, aiming to amendments and hence Harmonization. It is proposed that this procedure is reliable and can be implemented to any CLEN method, provided that sufficient PT data are available.
