Prospecting research: knowing when to stop

An important aim of many surveys is to undertake prospecting research: the search for new possibilities and an understanding of the diversity in a population. This paper develops a method — the extrapolation of resampled possibilities (ERP) — for predicting how much new information will be revealed by extending a sample. This is useful for deciding whether it is likely to be worthwhile sampling more cases, bearing in mind the costs and the benefits. The method avoids making any assumptions about the nature of the underlying population, apart from the information implicit in the existing sample.