Efficient estimation of interlaboratory and in-house reproducibility standard deviation in factorial validation studies
In case that fewer laboratories are available than required in order to ensure reliable precision estimates, the method validation study approach that is put forward in ISO 5725-2 is not applicable. Nested designs (as described in ISO 5725-3) can represent an alternative approach, yet they present important drawbacks, too: For example, a large number of measurements must be carried out and the available number of factor levels may be insufficient. An alternative approach consists in implementing orthogonal factorial designs. Both, the number of required laboratories and the number of required measurements inside each laboratory are considerably reduced. Moreover, the breakdown of the variability into component factor variances allows a differentiated and efficient estimation of the precision data.