On the collaborative trial in sampling

Abstract

The feasibility of the collaborative trial for validating a sampling protocol is demonstrated. An example in environmental sampling was convenient for the study but the method could be applied in any situation where replicate samples could be taken. A number of samplers each independently take duplicate samples in a random fashion. The samples are then analysed chemically under repeatability conditions and then resultant data analysed statistically by nested analysis of variance. Results can be compared with criteria based on considerations of fitness-for-purpose, and uncertainties associated with sampling can be estimated.