Audit and assessment of sampling systems

Many sampling systems (procedures, equipment, practices) are “incorrect” in the Gy sense. Correct sampling means that all constituent elements of the lot have an equal opportunity of being in the sample, and the integrity of the sample is preserved both during and after sampling [G. Pierre, Sampling for Analytical Purposes, Wiley, West Sussex, England, 1998, p. 28]. Correct sampling is a generalization to bulk materials (solids, liquids, gases) of statistical random sampling. Sampling procedures may be vague, and sampling practices may not follow correct procedures. Sampling equipment may not perform well in practice even though it is designed correctly. Operators (those taking the samples) may not be properly trained in correct sampling. Without correctness, sampling bias is introduced and sampling variation is increased, sometimes substantially, beyond the unavoidable statistical sampling variation. Since samples are only as good as the sampling systems that generate them, incorrect sampling will remain undetected without an examination and evaluation of the sampling systems. Steps are given for conducting an audit and assessment of procedures, equipment, and practices. Several examples are included for illustration.