Learning by design: lessons from a baseline study in the NAMWASH Small Towns Programme, Mozambique

This paper examines the design and application of a baseline study for a comprehensive water, sanitation, and hygiene (WASH) intervention in Mozambique. The study was developed to investigate the relationships among key parameters of interest both for comparison to post-implementation data and to contribute to planning the WASH intervention itself. We use this study to discuss key issues surrounding baseline studies. This includes providing guidelines for designing a WASH baseline survey, determining an appropriate sample size, and highlighting key considerations in analysing the survey data, such as incorporating the study design in statistical analyses, post-stratifying, and utilizing geospatial data. We also show how statistical analyses from a baseline survey can be used to inform subsequent surveys. For example, results from this study suggest that in future WASH studies, self-reporting by households should be supplemented by observational or population data to remove or quantify reporting bias, and care must be taken to reduce respondent fatigue.