Samples were sent to seven laboratories in five countries in a
blind inter-laboratory proficiency study, each using the described
GC method. Table 3 shows reproducibility (RSDR) of 12.84%,
reducing to 8.91% with three Grubbs outliers.
Duplicates performed as part of the proficiency study allowed
the estimation of additional repeatability (RSDr) data. An average of
2.92% (HorRat ¼ 0.46) was observed with no Cochran outliers and a
range of 1.12e6.65% (HorRat ¼ 0.21e1.04), concurring with
repeatability data reported in Table 1.