Sampling in surveys with reduced populations: a simplified method for the water, sanitation, and hygiene sector
Making decisions efficiently and equitably requires up-to-date and reliable information. In the water, sanitation, and hygiene sector, devolving decision-making to local governments is increasingly promoted to stimulate local development. However, too few data are available at a level of disaggregation that is appropriate to allow decisions to be made about local-level allocations or for monitoring equitable outcomes across communities. Collecting robust data through cost-effective methodologies is therefore a key element of planning and programming, and for this correct sampling methods are of primary importance. Although different sampling strategies are currently being used to support national-level interventions, none have been optimized for data collection at the local level with only small populations, and standard approaches are usually overly expensive and time consuming. Against this background, we used simplified linear piecewise approximations in this paper to calculate the sample size for proportions in terms of given precision, confidence levels, and population size. To support the use of this proposed approach by practitioners in the field, easy-to-use pre-calculated tables have been included. For sampling, easy-to-follow practical guidelines for household selection and transect walk planning are also provided. Further, six rural communities in Honduras are presented as an initial case study, with total populations ranging from 11 to 44 households. The results illustrate the validity and applicability of this approach for sampling design and sample size determination.