Framework for designing sampling programs

A general framework for designing sampling programs is described. As part of the sampling program the problem of concern, or reason for sampling, needs to be clearly stated and objectives specified. The development of a conceptual model will assist the clarification of objectives and the choice of indicators to be sampled.Objectives can then be stated as testable hypotheses and decisions made about the samallest difference/changes that are to be detected/observed by the sampling.To allow the collection of representative samples, and the statistical analysis of data to be collected, the potential sources of variability in the data must be considered. Site, selection, frequency and replication must account for the expected variability.Before field collection of samples occurs, the sample collection device needs to be tested as to its efficiency to collect a representative sample. It also will usually be necessary to consider how samples are to be preserved to inhibit biological and chemical change. All sample programs require a quality assurance program to identify, measure and control errors.As well as the above the cost-effectiveness of the program should be evaluated in terms of maximizing the information obtained/cost.